Download:
pdf |
pdfAttachment 4(c)
National HIV Surveillance System (NHSS)
OMB # 0920-0573
Supplemental Surveillance Activity 1: HIV Incidence Surveillance Technical Guidance
1
Technical Guidance for HIV
Surveillance Programs
HIV Incidence Surveillance
HIV Incidence and Case Surveillance Branch, Atlanta, Georgia
Cleared by HICSB Management 10/15/2012
Notes
Cleared by HICSB Management 10/15/2012
Table of Content
Content
1. Background
1.1 Introduction
2. Structural Requirements
2.1 Policies and Procedures
2.2 Definition of Reportable Information
2.2.1 STARHS Result
2.2.2 HIV Testing and Treatment History (TTH)
2.3 Staffing Needs
3. Process Standards
3.1 Collection of HIS Data Elements
3.1.1 TTH Data Elements
3.1.2 STARHS Result
3.2 Entering Data into Surveillance Databases
3.3 Creating the HIS Dataset and Transferring Data to CDC
3.4 Data Quality
3.5 Ensuring Security and Confidentiality
3.6 Analyzing and Disseminating Data
3.7 Training Staff
4. Outcome Standards
5. References
6. Appendices
6.1. Appendix 1 National Center for HIV/AIDS, Viral Hepatitis, STD, and TB Prevention’s
Non-research Determination for HIV Incidence Surveillance
6.2 Appendix 2 Training Resources for HIV Incidence Surveillance
6.3 Appendix 3 Test Used by STARHS Laboratory for Recency Classification
6.4 Appendix 4 Appendix 4.1 STARHS Specimen Guidance: Specimens Originating from
the Public Health Laboratory
Appendix 4.2 STARHS Specimen Guidance: Specimens Originating from a
Private/Commercial Laboratory and Sent to a Public Health Laboratory
Appendix 4.3 STARHS Specimen Guidance: Specimens Originating from a
Private/Commercial Laboratory and Sent Directly to the STARHS
Laboratory
Appendix 4.4 Guidelines for Preparing Specimens for Shipping and
Transporting to the STARHS Laboratory
6.5 Appendix 5 Guidance for Collection and Data Entry of HIV Incidence Surveillance
Information
6.6 Appendix 6 Local HIV Incidence Estimation Guide
6.7 Appendix 7 Completeness Report Documentation
6.8 Appendix 8 Error Report Documentation
Cleared by HICSB Management 10/15/2012
Page
Technical Guidance for HIV Surveillance Programs — Policies
and Procedures for HIV Incidence Surveillance
Background
Introduction
The Centers for Disease Control and Prevention (CDC), in conjunction with a number of state and
local health departments, conducts national HIV incidence surveillance (HIS), to generate annual
national, state, and local estimates of the number of new HIV infections, both diagnosed and
undiagnosed. HIV incidence reflects the leading edge of the epidemic, its trends, and its impact on
the public’s health. Using the stratified extrapolation approach1 (SEA), CDC calculated the first
national HIV incidence estimate based on a direct measure of recency of infection for 20062. In
2011, using a refinement of the SEA, estimates were updated for 2006 and calculated for 200720093. CDC expects to annually publish four year estimates based on the refined SEA in order to
provide updated data on trends in incidence as the data are available.
Since 2001, CDC has funded selected state and local health departments to conduct HIS; including
the collection of information on demographics, testing and treatment history (TTH), and results of
a serologic marker of recent HIV infection (i.e., the serologic testing algorithm for recent HIV
seroconversion [STARHS])4 which uses a test of recent infection to classify new diagnoses of HIV
infection as either recent or long-standing. In 2004, STARHS used a detuned version of a
commercially available HIV diagnostic test, the Vironostika HIV-1 EIA assay, which made the test
less sensitive. Because the Vironostika HIV-1 EIA had previously been approved for diagnostic
purposes by the United States Food and Drug Administration (FDA), the less sensitive version of
the test (Vironostika-LS) required individual consent. TTH and consent were collected during an
interview with persons seeking HIV testing (pre-test) or after receiving a new HIV diagnosis (posttest). In March 2005, the FDA allowed the use of a second assay for recency testing, the BED
HIV-1 Capture enzyme immunoassay (BED), for surveillance purposes only, which eliminated the
need for patient consent. In 2007, CDC reduced the number of data elements required for TTH and
expanded modes of collection of TTH data elements to include provider reports, medical record
review, other databases and the National HIV Monitoring & Evaluation System (NHM&E),
formerly Program Evaluation and Monitoring System (PEMS). Since the implementation of HIS
through much of 2011, HIS data was stored in a CDC-provided Microsoft Access database. In
2011, HIS data were imported into eHARS, a document-based data system used for storing all HIV
surveillance information and producing datasets for local analysis and transfer to CDC.
As an integrated component of the national HIV surveillance system, HIS incorporates into routine
case reporting the collection of TTH data and recency test results in the states and cities that
receive funding to conduct HIS.
The purpose of HIV incidence surveillance is to provide reliable and scientifically valid estimates
of the number of newly acquired HIV infections at the local, state, and national level. CDC’s
human subjects protection process has determined that HIS, like other public health surveillance
Cleared by HICSB Management 10/15/2012
activities, is not research (See Appendix 1- National Center for HIV, STD, and TB Prevention’s
Non-research Determination for HIV Incidence Surveillance).
The primary functions of HIS are to:
Describe demographic and behavioral characteristics of newly HIV infected populations
and subgroups
Identify emerging epidemics and monitor trends in transmission
Assist with national, state and local planning and evaluation activities
Monitor the outcomes of national, state and local level HIV prevention programs and
strategies
The tasks to achieve the functions of HIS include:
Collaborate with CDC, laboratories, HIV testing providers and affected communities to
further develop and ensure the capacity to conduct HIV incidence surveillance
Obtain HIV testing and treatment history information on all individuals newly diagnosed
with HIV infection reported to HIV Surveillance
Collect results from tests for recent HIV infection (e.g., STARHS or other methods as they
become available) necessary for the statistical estimation of HIV incidence. Currently,
results are obtained through the required submission of remnant blood specimens or dried
blood spot specimens from HIV diagnostic tests for recency testing using STARHS at a
CDC funded laboratory. In the future, the collection of dried blood spot samples may be
necessary
Submit data monthly or as specified by the CDC using eHARS software according to data
submission standards established by CDC
Conduct systematic evaluation of HIV incidence surveillance
Calculate and disseminate annual population-based HIV incidence estimates and promote
the use of HIV incidence data for prevention and health services planning
Ensure that the security and confidentiality procedures of the program are consistent with
the requirements delineated in the Data Security and Confidentiality Guidelines for HIV,
Viral Hepatitis, Sexually Transmitted Disease, and Tuberculosis Programs: Standards to
Facilitate Sharing and Use of Surveillance Data for Public Health Action
The prerequisites (structural requirements), best practices (process standards), and outcome
standards for HIS are described next.
Structural Requirements
HIV infection may be detected at various points along the spectrum of disease, and reportable
events range from reporting of HIV infection in otherwise asymptomatic persons to deaths among
people with HIV. Areas conducting HIS are required to incorporate the collection of TTH
variables and results from a test for recent infection from all newly diagnosed individuals aged 13
or older into HIV case reporting. TTH should be collected for all cases newly diagnosed with HIV
infection, however, those diagnosed with HIV infection, Stage 3 (AIDS) are excluded from
collecting results from a test for recent infection. Data should be collected in accordance with the
Technical Guidance for HIV Surveillance Programs, Vol. I: Policies and Procedures - Access to
Cleared by HICSB Management 10/15/2012
Source Data and Completeness of Reporting and the Instructions for Completing the Data
Collection Form. HIV surveillance case report data, in combination with HIS data elements, are
used to calculate population-based HIV incidence estimates.
Policies and Procedures
It is important to document HIS activities and to address HIS needs in the jurisdiction’s HIV
surveillance policies and procedures manual to establish standardization, continuously maintain
conformity of meaning for data elements, document changes over time, and develop training
programs. In addition to the information listed in the Technical Guidance for HIV Surveillance
Programs, Vol. I: Policies and Procedures, HIS specific policies and procedures should include
information related to:
Obtaining HIV testing and treatment history information on all individuals newly
diagnosed with HIV infection reported to HIV surveillance
Collecting results from tests for recent HIV infection. Currently, results are obtained
through submission of remnant specimens from HIV diagnostic tests for testing at a CDC
funded laboratory. Therefore, for all newly diagnosed HIV infections reported to HIV
surveillance, include procedures for:
o Establishing and maintaining collaboration and communication with public and
private HIV testing laboratories (within and outside the state) to secure remnant
specimens from the original diagnostic HIV test or another HIV related test
conducted within 3 months of the initial diagnosis
o Securing, transporting and tracking remnant specimens shipped to the designated
laboratory for recency testing from public and private HIV testing laboratories
(within and outside the state)
Developing a STARHS eligibility list to identify which diagnostic specimens represent
diagnoses of HIV infection not known to have progressed to AIDS and to inform the
appropriate laboratory of the need for shipping and/or testing using an approved test for
recent HIV infection
Training all HIV case reporters (e.g. HIV testing providers, HIV care providers, etc.) and
staff involved in TTH data collection (See Appendix 2 - Training Resources for HIV
Incidence Surveillance)
Managing data and conducting activities to ensure data quality
Conducting systematic evaluations of HIV incidence surveillance using outcome and
process standards and using evaluation results for program improvement
Calculating and disseminating annual population-based HIV incidence estimates and
promoting the use of HIV incidence data for prevention and health services planning
Using data collected for HIS (including estimates, results of tests for recent infection and
TTH data)
Ensuring security and confidentiality procedures of the program are consistent with the
requirements delineated in the Data Security and Confidentiality Guidelines for HIV, Viral
Hepatitis, Sexually Transmitted Disease, and Tuberculosis Programs: Standards to
Facilitate Sharing and Use of Surveillance Data for Public Health Action
Definition of Reportable Information
Cleared by HICSB Management 10/15/2012
STARHS Specimen Information and Results
STARHS is a two test algorithm conducted on remnant specimens from HIV tests performed as
part of a diagnostic algorithm or after confirmation of HIV diagnosis. The first test is used to
confirm the presence of HIV specific antibodies using a standard enzyme-linked immunoassay
(EIA) test and is followed by a test that determines recency of HIV infection (See Appendix 3 Test Used by STARHS Laboratory for Recency Classification).
HIS coordinators in funded areas collaborate with public and private/commercial laboratories to
locate, determine the disposition of, and ship remnant diagnostic blood specimens for testing using
STARHS ((See Appendix 4.1 – STARHS Specimen Guidance: Specimens Originating from the
Public Health Laboratory, Appendix 4.2 - STARHS Specimen Guidance: Specimens Originating
from a Private/Commercial Laboratory and Sent to a Public Health Laboratory, Appendix 4.3 STARHS Specimen Guidance: Specimens Originating from a Private/Commercial Laboratory and
Sent Directly to the STARHS Laboratory, Appendix 4.4 - Guidelines for Preparing Specimens for
Shipping and Transporting to the STARHS Laboratory).
Testing and Treatment History Data (TTH)
The primary purpose of gathering HIV testing and antiretroviral (ARV) use history is to
calculate a statistical weight corresponding to the probability that an individual would be tested
for HIV in the STARHS recency period. This weight is used in HIV incidence estimation.
TTH data elements are used to determine testing frequency, in order to classify cases as new
testers (i.e. people whose first HIV test was positive) or repeat testers (i.e. people testing HIVpositive after a previous negative HIV test). This distinction is important for incidence
estimation because the probability of being classified as recent is calculated separately for new
testers and repeat testers. In addition, critical TTH data elements are used to determine whether
ARV use might have affected the results of the recency test.
A standard set of HIV testing and ARV use history data elements needed for the HIV incidence
estimate has been developed (See Appendix 5 - Guidance for Collection and Data Entry of HIV
Incidence Surveillance Information). This information is necessary for cases newly reported to
HIV surveillance to estimate HIV incidence successfully.
Staffing Needs
HIS operation requires personnel with specific skills and dedicated time to integrate HIS
activities with case surveillance and other program activities (e.g. data collection, data
management, data quality, data de-duplication, dissemination, etc.) to build a seamless
surveillance system and maximize efficient use of resources. Generally, HIS staff should have:
An understanding of the general principles of HIV surveillance in the jurisdiction
Good communication skills
Strong leadership skills
Enthusiasm about disease reporting for public health purposes
An orientation towards meeting HIS process and outcome standards
Cleared by HICSB Management 10/15/2012
Ability to work closely with CDC, other states, local sites, private providers, and
laboratories
CDC recommends at least 0.5 full-time equivalent (FTE) be dedicated to the incidence
surveillance coordinator (ISC) position for monitoring and evaluating the activities of HIS and 0.5
FTE for data management. Other personnel assigned to HIS may vary depending on program
needs, prevalence of HIV, and available resources. Additional staff may include an epidemiologisttrainer (0.5 FTE), a laboratory liaison (0.5 FTE) and field staff for data abstraction. Finally, as HIV
incidence estimation is a highly technical activity, areas funded for HIS should ensure access to a
mathematical statistician. If a person with these skills is not available within the health department
consultation services should be procured. Recommended staffing roles and responsibilities of HIS
staff are described next.
Position
Incidence
Surveillance
Coordinator
(ISC)
Responsibilities
Epidemiologist
/Trainer
Serve as the primary point of contact for CDC on HIS
Provide overall management of the HIS program and determine
program goals and needs
Update HIS policies and procedures manual
Oversee processes for the collection of TTH and STARHS data
Identify which diagnostic specimens represent new diagnoses of
HIV infection not known to have progressed to AIDS
Communicate with HIV case reporters and data collectors as needed
regarding ascertainment and accuracy of HIS data elements
Coordinate the analysis and dissemination of HIV incidence data
and assess trends in new infections at the local/state level
Evaluate the performance of HIV incidence surveillance using
process and outcome standards
Manage any employee or other service contracts related to HIS
Participate in CDC site visits, trainings, and workshops
Maintain the security and confidentiality of HIS data
Serve as lead for training HIV testing providers and laboratories on
HIS, including development/modification of surveillance areaspecific training materials
Verify the adherence to TTH data collection standards and proper
submission of TTH data elements
Conduct quality control activities to address the accuracy of the data
collected
Provide feedback to HIV case reporters and data collectors on TTH
completeness and timeliness
Coordinate the collection of TTH data elements from HIV case
reporters and data collectors
Use process and outcome standard evaluation results for program
improvement
Provide consultation and technical assistance on TTH data
collection
Participate in CDC site visits, trainings, and workshops as
Cleared by HICSB Management 10/15/2012
Certifications/
Expertise
• SAS (SAS EG)
[Desired]
Training
Data collection
Data quality
SAS (SAS-EG)
Laboratory
Liaison
Data Manager
appropriate
Maintain the security and confidentiality of HIS data
Coordinate with public and private HIV testing laboratories (within
and outside the state) to secure remnant specimens from the original
diagnostic HIV test or other HIV related test conducted within 3
months of the initial diagnosis
Coordinate/communicate with public and private HIV testing
laboratories (within and outside the state) to arrange transport of
remnant specimens to the designated laboratory for recency testing
Maintain communications between public health, private, and
community laboratories, the STARHS laboratory and the ISC
Establish specimen tracking mechanisms
Serve as the subject matter expert on the preparation and shipping
of specimens to the STARHS laboratory including quality control
procedures
Oversee specimen identification and tracking
Provide feedback to laboratories on specimen quality and quantity
as well as frequency of specimen shipments
Participate in CDC site visits, trainings, and workshops as
appropriate
Maintain the security and confidentiality of HIS data
Oversee processes for data entry of incidence data elements
Conduct data quality assessments, dataset creation and transfers,
and data management programs
Maintain documentation of HIS data management processes in
policies and procedures manual
Participate in process and outcome standard evaluations
Serve as subject matter expert on HIV incidence data elements, data
set preparation for local and national incidence estimation, and daily
data management
Participate in CDC site visits, trainings, and workshops
Maintain the security and confidentiality of HIS data
• Air Transport
Association
(IATA)
Training/Speci
men
handling/aliquo
tting
experience
• Laboratory
database
experience
• Encryption
software
• SAS (SAS EG)
• Encryption
software
Process Standards
HIS involves the following processes:
Ongoing, collect HIV testing and treatment history information on all individuals aged 13
years or older newly diagnosed with HIV infection and reported to HIV surveillance
Ongoing, collect results from tests for recent HIV infection for all individuals 13 years and
older newly diagnosed with HIV, excluding those with HIV infection, Stage 3 (AIDS).
Obtain results through required submission of a remnant specimen from an HIV diagnostic
test to a CDC funded laboratory as needed
At least monthly, enter HIS data into the eHARS database (version 3.3 or later)
Monthly, transfer HIS data electronically to CDC using eHARS software according to data
submission standards established by CDC
Routinely, conduct data quality control activities including data error resolution and
monitoring monthly data completeness
Cleared by HICSB Management 10/15/2012
Calculate and disseminate annual population-based HIV incidence estimates and promote
the use of HIV incidence data for prevention and health services planning
At least annually, conduct a systematic evaluation of HIV incidence surveillance using
outcome and process standards and use evaluation results for program improvement
Ongoing, ensure that data handling procedures comply with Security and Confidentiality
Guidelines (See Data Security and Confidentiality Guidelines for HIV, Viral Hepatitis,
Sexually Transmitted Disease, and Tuberculosis Programs: Standards to Facilitate
Sharing and Use of Surveillance Data for Public Health Action and Model State Public
Health Privacy Act)
At least annually, provide training to HIV reporters and data collectors in HIV surveillance
and HIS methods
Collection of HIS Data Elements
As a general principle, TTH data should be collected for all individuals aged 13 years or older
newly reported to the HIV surveillance system by all providers of HIV testing. Results from tests
for recent HIV infection should be collected for all individuals 13 years and older newly diagnosed
with HIV, excluding those with HIV infection, Stage 3 (AIDS).
Cases Out of Jurisdiction
HIS funded jurisdictions should collect TTH data (including ARV use variables) and STARHS
results for all persons newly reported to HIV surveillance and work with other areas to exchange
needed data. Refer to the Technical Guidance for Surveillance Programs, Vol. I: Case Residency
for information about interstate/reciprocal notification of reportable diseases. For cases reported to
an HIS funded jurisdiction and residing in any HIS funded jurisdiction HIS areas are expected to
collect TTH data and ensure that a remnant of the diagnostic specimen is tested using STARHS for
all cases (not known to be AIDS – Stage 3). All TTH and STARHS data should be forwarded to
the state of residence at HIV diagnosis if the state is funded for HIS. In addition, ARV use data
typically collected through HIS are critical to Molecular HIV Surveillance (MHS; formerly
variant, atypical, and resistant HIV surveillance) data analyses; HIS areas should collaborate with
other jurisdictions to provide needed data.
STARHS Specimen Information and Results
Collecting results from tests for recent HIV infection for all individuals 13 years and older newly
diagnosed with HIV, excluding those with HIV infection, Stage 3 (AIDS) is required for HIS. For
all newly diagnosed HIV infections reported to HIV surveillance, funded health departments are
required to:
Identify and locate diagnostic specimens that represent diagnoses of HIV infection not
known to have progressed to AIDS at the time of HIV diagnosis
Secure remnant specimens from the original diagnostic HIV test or other HIV related test
conducted within 3 months of the initial diagnosis from public and private HIV testing
laboratories (within and outside the state).
Determine the disposition of specimens for recency testing using STARHS by checking the
quarterly specimen inventory list sent by the STARHS laboratory and by developing a
monthly STARHS eligibility list for both the public health laboratory and the STARHS
Cleared by HICSB Management 10/15/2012
laboratory
Establish mechanisms to track shipments as appropriate and receipt of results
Establish regular communication between the ISC, HIS staff, the public health laboratory,
private and commercial laboratories, and the STARHS laboratory
Results from tests for recent infection are obtained through locating and determining the
disposition of specimens, preparing STARHS eligibility lists, and transporting remnant HIVpositive specimens to the STARHS laboratory. This laboratory is a CDC funded centralized
laboratory responsible for receiving and storing specimens, conducting tests for recent infection
within STARHS, and communicating results of tests for recent infection to jurisdictions
conducting HIV incidence surveillance
Laboratories that perform routine diagnostic confirmatory/supplemental HIV testing and
immunological status tests such as CD4, or viral load counts should report results to the state/local
health department surveillance program per existing requirements. For the purposes of this
guidance, laboratories that are sources of remnant specimens for recency testing have been
classified into 3 types:
1) Commercial laboratories that process specimens from many states and/or jurisdictions
(included in this category are: Quest Diagnostics Inc., Laboratory Corporation of America
[LabCorp], ARUP Laboratories, , and Mayo Clinic)
2) Private laboratories, which include smaller private/university/hospital or medical center
laboratories that provide service primarily at the state or local level
3) State and/or local public health laboratories that work in collaboration with other arms of
the nation’s public health system, to provide clinical diagnostic testing, disease
surveillance, environmental and radiological testing, emergency response support, applied
research, laboratory training and other essential services to the communities they serve.
In each surveillance area, all laboratories that perform HIV diagnostic tests for residents of the
surveillance area should be identified from a review of local HIV surveillance data and state
laboratory licensing records. An updated directory of laboratory contacts at all reporting
laboratories should be maintained to facilitate communication in the event that reporting or
shipping of specimens is disrupted or that changes in policy or procedures need to be
communicated. The laboratories identified should be contacted to request that remnants of HIVpositive diagnostic specimens be made available for recency testing using STARHS. The reports
that these laboratories send to the state/local health department surveillance programs per the
established reporting requirements should include the specimen accession numbers.
Sending specimens from a laboratory that performs the initial screening HIV EIA test (originating
laboratory) to a second laboratory for confirmatory/supplemental HIV testing (reference
laboratory) is a common practice. A specimen may be assigned a different accession number at
each laboratory that receives it; therefore, it is important to ensure that the accession number that is
assigned by the laboratory that ships the specimen to the STARHS laboratory is transmitted to the
surveillance program as part of the laboratory reporting process.
Cleared by HICSB Management 10/15/2012
HIS staff, using routine HIV surveillance reporting procedures, determines whether the
specimen represents the person’s first reported positive HIV test result in the HIS area or was
from an HIV related test drawn within three months of the initial reported positive HIV test
result and is eligible for recency testing.
A specimen should be tested for recency of HIV infection if
the specimen represents the diagnostic specimen (the HIV-positive specimen that led, or
should have led an adolescent/adult HIV case to be reported to HIV surveillance) or
the diagnostic specimen is unavailable and the specimen was drawn within 3 months of the
diagnostic specimen and
the specimen was drawn for an HIV-related test (including viral load, polymerase chain
reaction [PCR] test, and CD4 count)
A specimen should not be tested for recency of HIV infection if:
the individual had a previous specimen that was tested for recency
the specimen was drawn more than 3 months after the diagnostic specimen
the individual is known to have HIV infection, Stage 3 (AIDS) at diagnosis
the individual has a positive HIV-1 viral load or qualitative RNA specimen, but a negative
HIV-1 antibody test result on or after the collection date of the positive HIV-1 viral load
the individual has a positive rapid test and an undetectable viral load and no other
diagnostic test results are available to confirm/supplement diagnosis
the individual has a HIV-2 antibody positive diagnosis or is positive for both HIV-1 and
HIV-2 antibodies (dual infection)
Specimens will be classified as HIV-positive at one of three types of laboratories previously
described. For each positive HIV test reported to an HIS jurisdiction the disposition of that
specimen as either eligible or ineligible to be tested for recency of HIV infection must be
determined. At the public health laboratory only specimens deemed eligible to be tested for
recency of HIV infection are shipped to the STARHS laboratory whether specimens were tested at
the public health laboratory or shipped there by another diagnostic laboratory. The disposition of
the specimens must be communicated by HIS staff to the public health laboratory (See Appendix
4.1 – STARHS Specimen Guidance: Specimens Originating from the Public Health Laboratory,
Appendix 4.2 - STARHS Specimen Guidance: Specimens Originating from a Private/Commercial
Laboratory and Sent to a Public Health Laboratory ).
Commercial and private laboratories may ship all remnant HIV-positive specimens directly to the
STARHS laboratory. The disposition of these specimens must be communicated by HIS staff to
the STARHS laboratory. Communication with the STARHS laboratory related to the disposition of
remnant specimens is accomplished using the STARHS Eligibility List (See Appendix 4.3 STARHS Specimen Guidance: Specimens Originating from a Private/Commercial Laboratory and
Sent Directly to the STARHS Laboratory). For instructions for the shipment of remnant HIVpositive diagnostic specimens please see Appendix 4.4 - Guidelines for Preparing Specimens for
Shipping and Transporting to the STARHS Laboratory.
Cleared by HICSB Management 10/15/2012
As future tests of recent HIV infection become available, programs must adapt to CDC guidance
for the collection of results from these tests.
Testing and Treatment History (TTH)
The general principle in TTH data collection is to obtain the most accurate information available to
characterize a person’s HIV testing and treatment history. Critical TTH data elements include
standard variables used for HIV incidence estimation1, 3 and HIS evaluation.
TTH data are reported by providers and surveillance staff using the area standard reporting
procedures or other procedures that meet the routine security and confidentiality guidelines for
HIV surveillance. These data:
Should be included for all individuals aged 13 years or older that are newly reported to the
HIV surveillance system by all individuals responsible for reporting diagnoses of HIV
infection
May be collected from multiple sources of information, including patient interview,
medical records, laboratory reports, and other databases including NHM&E
Should be collected following the tenets of document-based surveillance
May be collected when investigating a case, completing a case report form, or conducting
interviews as part of counseling and testing services, partner services or other means
TTH fields are included in the revised CDC adult case report form (ACRF) (See Appendix 5 Guidance for Collection and Data Entry of HIV Incidence Surveillance Information and Technical
Guidance for HIV Surveillance Programs, Vol. II: Instructions for Completing the Data Collection
Form). It is important to examine all information available and ensure that data collected most
accurately reflect the patient’s actual HIV testing history and antiretroviral use.
CDC has developed materials for use in training HIV reporters and data collectors to obtain HIV
testing and treatment history data. (See Appendix 2 - Training Resources for HIV Incidence
Surveillance)
Entering Data into Surveillance Databases
HIS data may be entered manually or imported into eHARS, but should follow the principles of
document-based surveillance. TTH data are entered into an eHARS TTH document; separate
eHARS documents are used to enter and store data from different forms, or from multiple sources,
or to update information. This important feature allows CDC and the project area to track the
relative proportions, and data quality and completeness of TTH documents collected through
various reporting methods (e.g., patient interview, medical record review, or passive case
reporting) for the purpose of program improvement. Data entry staff is expected to enter data in
eHARS as reported on the data collection form. If conflicts within a data collection form are
identified, the HIS program should have procedures in place to resolve the data conflicts. If the
conflict is the result of a typographical error in a single document, the error should be corrected in
that document.
STARHS eligibility lists prepared by HIS staff are sent either to the public health laboratory
(which will assign STARHS IDs to eligible specimens and ship them to the STARHS laboratory)
Cleared by HICSB Management 10/15/2012
or to the STARHS laboratory for specimens sent directly from private or commercial laboratories
(which ship specimens identified only by specimen ID or accession number). The STARHS
laboratory and the PHL will send back to the HIS staff a link file for eligible specimens matching
STARHS ID numbers assigned by the appropriate laboratory to corresponding specimen accession
numbers for specimens that were identified as eligible by HIS staff. Following testing for recent
infection, the STARHS laboratory will send results (identified only by STARHS ID) to HIS staff.
There are two approaches for entering specimen and STARHS information into eHARS based on
whether or not the jurisdiction uses an external data tracking system to manage STARHS specimen
data. The first approach applies to jurisdictions that manually enter or import STARHS specimen
data into eHARS before importing STARHS results. STARHS specimen required and
recommended information for eligible specimens should be entered into eHARS on a regular basis,
including STARHS ID, first, middle and last name of the case reported, specimen collection date,
STARHS test type, specimen type, relevant specimen IDs or accession numbers, and if necessary
information on the reasons specimens were not sent for testing for recent infection (e.g. quantity
not sufficient, etc.). It is important to note that before importing STARHS result files, the
STARHS specimen data, particularly the STARHS ID, collection date and test type must be
entered into eHARS as a Lab document. In this approach, when STARHS results are returned,
they will be matched to cases in eHARS using the previously entered/imported STARHS
ID. Some important STARHS specimen information including specimen collection date,
accession number, specimen ID and the case’s first, middle and last names will be extracted (using
the CDC provided SAS program) from the previously entered STARHS specimen lab document
and included in the STARHS result import file. After importing the STARHS result file, a new
Lab document containing the results and specimen collection date, accession number, specimen ID
and the case’s first, middle and last names will be added to each corresponding case. Thus, each
case will have at least two documents associated with a given recency testing event. For analysis,
the data from all Lab documents for a unique STARHS ID should be combined.
The second approach applies to jurisdictions that manage STARHS specimen information using a
locally designed data tracking system outside of eHARS. Approach 2 combines the locally-stored
specimen data (from the tracking database) with the STARHS results (from the STARHS lab)
outside of eHARS and allows the import of both into eHARS on a single document. The external
data tracking system should include the following required and recommended information:
STATENO, first name, middle name, last name, STARHS ID, laboratory specimen IDs or
accession numbers, specimen collection date, and specimen type. To combine STARHS results
with the specimen data, one needs to first extract the specimen data from the external tracking
system as a CSV file (see Table 1 which provides the structure of the CSV file along with variable
names, descriptions, types and acceptable values), then run the SAS import preparation program
provided by CDC. The SAS import preparation program combines the specimen data and
STARHS result data by STARHS ID and prepares an auto-import STARHS data file based on
STATENO. Importing the auto-import STARHS data file into eHARS will add a STARHS lab
document containing both specimen and STARHS result data for a given specimen to existing
eHARS cases. If the jurisdiction uses the CDC-provided SAS program, the eHARS auto-import
file will contain only specimen data for those specimens that have a matching STARHS ID in the
Cleared by HICSB Management 10/15/2012
STARHS result file. The specimen data without a matching STARHSID in the STARHS result
file are likely to be specimens that are waiting for STARHS results, and therefore will be matched
to future STARHS results. Most surveillance areas also maintain in their tracking system data on
specimens that are not sent for testing for recent infection. Information on the reason specimens
were not sent for testing for recent infection cannot be imported into eHARS, therefore data for
these specimens should be manually entered into eHARS, along with required or recommended
specimen information such as specimen collection date, STARHS test type, specimen type and
relevant specimen IDs. With Approach 2, surveillance staff can enter data into eHARS manually
when necessary.
Table 1: The STARHS specimen data CSV file structure with variable name, label, type, and
acceptable values. Note: It is REQUIRED to maintain the order of the variables; the header row is
not necessary.
STATENO
State No
FIRST
First
name
MIDDLE
Middle
name
LAST
Last
name
SSTARHSID
STARHS
specimen ID
Char(35)
Char(50)
Char(50)
Char(50)
Char(15)
SSTATEID
State Lab or
other
specimen ID
Char(50)
LSrceID
Source Lab
specimen ID
Char(50)
LDteObt
Specimen
collection
date
mm/dd/yyyy
LSpecTy
Specimen
Type
Char(10)
URN
BLD
OTH
SAL
UNK
A list of data elements necessary for HIS can be found in the Local HIV Incidence Estimation
Guide – Appendix 6. Information regarding HIS data entry and variables that are required or
recommended for eHARS HIS documents are designated in the Guidance for Collection and Data
Entry of HIV Incidence Surveillance Information - Appendix 5.
Creating the HIS Dataset and Transferring Data to CDC
eHARS is the only data entry and management system that state or local health departments should
use for HIS. HIS data collected from other sources, including the excel spreadsheet from the
STARHS laboratory that contain STARHS results identified by STARHS ID, National HIV
Monitoring and Evaluation system (NHME), formerly known as the Program Evaluation
Monitoring System (PEMS) or other local databases should be manually entered or imported into
eHARS.
Jurisdictions should update and export eHARS person-based and document-based
datasets and run the CDC-provided SAS MACRO (HIS Data Export Program – provided
by the CDC HIS Program Coordinator upon request) to generate an incidence SAS
dataset containing all cases reported by the jurisdiction with accompanying incidence data
(TTH and STARHS result). Until National Data Processing (NDP) of eHARS data is
implemented, jurisdictions should submit the data to CDC via secure electronic data
methods between the 1st and 15th of each month. After the implementation of NDP,
jurisdictions should submit the previously described dataset to CDC via secure electronic
data methods between the 1st and 15th of the first month of each quarter. Data transmitted
Cleared by HICSB Management 10/15/2012
to CDC must not include personal identifiers and must be encrypted and password
protected see the Data Security and Confidentiality Guidelines for HIV, Viral Hepatitis,
Sexually Transmitted Disease, and Tuberculosis Programs: Standards to Facilitate
Sharing and Use of Surveillance Data for Public Health Action.
Data Quality
CDC has developed programs that generate reports to be used to improve the quality of incidence
data. There are two data quality measures: the HIV incidence completeness and error reports.
The HIV incidence completeness report provides a snapshot of the proportion of HIV cases
diagnosed within a three year period that have required TTH data elements and STARHS result.
Completeness is based on all HIV cases diagnosed in a specific year. Percentages are based on
cases reported from each HIS area and residing in any area funded for HIS. The HIV incidence
completeness report serves as a data quality monitoring tool to measure HIS data completeness,
identify problems with data transfer and monitor progress toward meeting target performance
levels and data quality standards. High completeness will support the HIV incidence surveillance
programs to meet outcome standards. If issues are identified, CDC will provide technical
assistance to improve the completeness of data.
The HIV incidence error report monitors data entry errors and data inconsistencies between
variables. It is based on cases reported from an HIS area and residing in areas funded for HIS. The
validity and accuracy of the HIV incidence estimates are limited if the data contain errors or a
large proportion of information is missing or unknown. If errors and inconsistencies are checked
and fixed routinely, programs can detect problems with data entry and data collection, and improve
data quality.
Monthly completeness reports and quarterly error reports serve as data quality monitoring tools for
CDC epidemiologists and state/local HIS staff to use in understanding HIS data and in working
together to improve the validity and accuracy of the data. HIS sites investigate and determine the
resolution for errors based on the recommendations provided by CDC (See Appendix 7Completeness Report Documentation; Appendix 8 - Error Report Documentation).
Ensuring Security and Confidentiality
Data collected for HIV incidence surveillance are considered part of routine surveillance and
should be held to the standards of security and confidentiality for HIV surveillance outlined in the
Data Security and Confidentiality Guidelines for HIV, Viral Hepatitis, Sexually Transmitted
Disease, and Tuberculosis Programs: Standards to Facilitate Sharing and Use of Surveillance
Data for Public Health Action. Policies and procedures, based on the guidelines and local laws, are
already in place at state and local health departments and are used to secure hard copies and
electronic data to protect the confidentiality of persons reported as having HIV infection.
Specimens and associated information that are carried through US mail or carrier services should
be treated as case information and should follow all security and confidentiality requirements that
apply to physical and data security. Access by all staff to information in eHARS, HIV testing and
ARV use history, and STARHS data is governed by the same security and confidentiality
Cleared by HICSB Management 10/15/2012
requirements. STARHS results using the BED are intended for HIV surveillance purposes only to
support population studies and cannot be interpreted at the individual level for clinical decisions.
Analyzing and Disseminating HIV Incidence Estimates
Population-based HIV incidence is estimated using a statistical approach that is analogous to that
used to estimate a population total from a sample survey. Following the sample survey approach,
the number of new HIV infections in the population is estimated from a sample composed of all
newly diagnosed HIV infections in the selected period of time classified as “recent” using the BED
by assigning a sampling weight to each result in the sample. This weight is the inverse of the
estimated probability that a seroconverted person would be identified as having a recent infection
based on the test for recent infection. Data from persons who choose to have a confidential HIV
test and who test positive are used to estimate the incidence of HIV infection, both diagnosed and
undiagnosed nationally and in participating areas. The method used to generate the populationbased incidence estimate was initially described by Karon et al in Estimating HIV Incidence in the
United States from HIV/AIDS Surveillance Data and Biomarker HIV Test Results1
HIV prevention programs can use HIV incidence estimates for targeting resources, and monitoring
and evaluating prevention activities locally, regionally, and nationally. CDC has the primary
responsibility for analyzing interpreting, and disseminating national HIV incidence estimates using
data from HIS areas as appropriate. Results will be presented at conferences and published in peerreviewed journals. The number of representative authors from participating areas and CDC will be
determined for each presentation or paper. State or local HIS areas should consider developing
their own HIV incidence estimates following Appendix 6 - Local HIV Incidence Estimation Guide.
Training Staff
In accordance with the Data Security and Confidentiality Guidelines for HIV, Viral Hepatitis,
Sexually Transmitted Disease, and Tuberculosis Programs: Standards to Facilitate Sharing and
Use of Surveillance Data for Public Health Action, HIS staff must also receive training on
security and confidentiality procedures, and must sign a confidentiality statement upon being
hired and annually thereafter. Because HIS is a fully integrated component of HIV surveillance,
all HIS staff should receive training in the local policies and procedures for case surveillance
including:
Active and passive surveillance methods
Laboratory reporting mechanisms
Data management processes
Besides the general training mentioned above, the HIS laboratory liaison, data manager and ISC/
epidemiologist should receive specific job related trainings such as:
Laboratory Liaison:
Role of Public Health Laboratory in incidence surveillance
Specimen handling i.e. specimen numbering, storage and retention
Aliquotting and specimens shipment to the STARHS laboratory according to guidelines
Entering data and generating reports for specimen tracking and determining specimen
Cleared by HICSB Management 10/15/2012
eligibility for recency testing using STARHS
Data Manager:
HIS data entry and data quality assessment
Secure electronic data transport procedures
Specimen eligibility and, tracking and maintaining results (electronic and hard copies)
Creating reports for HIV care and prevention programs
ISC/Epidemiologist:
• Incidence estimation training
• Program monitoring and evaluation
• eHARS training
Outcome Standards
Outcome standards described in the “Introduction to Policies and Procedures” and “Data Quality”
sections of Technical Guidance for HIV Surveillance Programs, Vol. I: Policies and Procedures can be
applied to HIS. These sections address completeness, case ascertainment, timeliness of reporting,
evaluation of standard data edits, and missing/unknown information. Meeting case surveillance
standards for case ascertainment and timeliness is essential for HIS to be successful given the timesensitive nature of HIS data elements, including TTH data and STARHS result. The quality of the HIV
incidence estimate depends on the quality and completeness of data included in the HIS system. All
outcome standards for HIS relate only to cases that reside within the surveillance area at the time of
diagnosis.
Data Quality
• 97% of case records pass all selected data edits related to HIS data, measured at 12 months
after the close of the report year, with a target of 100%
Completeness of Testing and Treatment History Data
• At least 85% of newly diagnosed HIV disease cases reported for a calendar year should have
HIV testing and treatment data assessed at 12 months after the end of the diagnosis year
Completeness of STARHS Result
• At least 60% of new diagnoses of HIV infection, excluding cases diagnosed with AIDS
within 6 months of HIV diagnosis, reported for a calendar year have a STARHS result from a
specimen obtained at, or within 3 months of HIV diagnosis assessed at 12 months after the
end of the diagnosis year
References
1.
Karon, J.M., Song, R., Brookmeyer, R., Kaplan, E.H., & Hall, H.I. (2008). Estimating HIV
incidence in the United States from HIV/AIDS surveillance data and biomarker HIV test
results. Statistics in Medicine 27(23): 4617-4633
2.
Hall, H.I., Song, R., Rhodes, P., Prejean, J., An, Q., et al. (2008). Estimation of HIV incidence in the
United States. Journal of the American Medical Association 300(5): 520-529.
Cleared by HICSB Management 10/15/2012
3.
4.
5.
Prejean J, Song R, Hernandez A, Ziebell R, Green T, et al. (2011) Estimated HIV Incidence
in the United States, 2006–2009 PLoS ONE August 2011 6 (8).
Janssen RS, Satten GA, Stramer SL, et al. New testing strategy to detect early HIV-1 infection
for use in incidence estimates and for clinical and prevention purposes. Janssen RS, Satten GA,
Stramer SL, et al. New testing strategy to detect early HIV-1 infection for use in incidence
estimates and for clinical and prevention purposes. JAMA 1998; 280:42–48.
Parekh.BS, Hanson DL, Hargrove J, Branson B, Green T, Dobbs T, et al,(2010). Determination
of Mean Recency Period for Estimation of HIV Type 1 Incidence with the IgG capture BED
enzyme immunoassay (BED-CEIA) in Persons Infected with Diverse Subtypes. AIDS Research
and Human Retroviruses 26(00): 1-9
Cleared by HICSB Management 10/15/2012
Technical Guidance for HIV/AIDS Surveillance Programs — HIV Incidence Surveillance
National Center for HIV, STD, and TB Prevention’s
Non-research Determination for HIV Incidence Surveillance
NCHSTP Research/Non-research Determination
(Request to Classify Project as Not Involving Human Subjects or Research)
August 2007National Center for HIV, STD, and TB Prevention’s Non-research Determination for HIV Incidence Surveillance
12-21
Technical Guidance for HIV/AIDS Surveillance Programs — HIV Incidence Surveillance
12-22
National Center for HIV, STD, and TB Prevention’s Non-research Determination for HIV Incidence SurveillanceAugust 2007
Appendix 2 - Training Resources for HIV Incidence Surveillance
The TTH Training Materials (Job aids) were developed to assist HIV Incidence Surveillance
sites in training staff to gather, record and enter testing and treatment history (TTH) data, and to
inform health care providers about their role in HIV incidence surveillance. In addition to the
TTH Job Aids, PowerPoint slides were presented at the 2010 HIV Incidence Surveillance
Workshop: Data Quality and Estimation and are available on on the HICSB SharePoint at
https://partner.cdc.gov/sites/NCHHSTP/HICSB/default.aspxto assist with training efforts.
The TTH Job aids have been developed for four different audiences:
1. TTH Training Materials for Health Care Provider (English and Spanish)
Intended for health care providers who care for HIV patients and record information in the
medical chart or on a provider case report form
Pocket card (4x6”) – reminder of key TTH questions
Quad-fold brochure - basic information on HIV incidence surveillance and the
critical role health care providers play in gathering accurate and complete TTH
information
2. TTH Training Materials for Patient Interview
Intended for personnel who interview patients directly for TTH
Patient Interview bullet point reference - designed to carry on a clipboard as a
quick reference when recording TTH from patient interviews
Patient Interview TTH ACRF - designed to post on a workspace bulletin board
3.
TTH Training Materials for Medical Record Review
Intended for personnel who perform medical records abstraction to obtain TTH information
Medical Record Review bullet-point reference - designed to carry on a clipboard as
a quick reference during medical record abstractions
Medical Record Review TTH ACRF - designed to post on a workspace bulletin
board
4. TTH Training Materials for Data Entry
Intended for personnel who enter TTH information into an electronic database
Data Entry bullet-point reference - quick reference for use at the data entry station
Data Entry TTH ACRF - designed to post on a workspace bulletin board
All files are available on the HICSB SharePoint at
https://partner.cdc.gov/sites/NCHHSTP/HICSB/default.aspx as modifiable files so that sites
can insert local data, change photos and include contact information. For additional information
regarding the use of these materials please contact your CDC epidemiologist.
Cleared by HICSB Management 10/15/2012
Appendix 3 – Test Used by STARHS Laboratory for Recency Classification
BED HIV-1 Capture Enzyme Immunoassay
Currently, within STARHS the EIA is followed by the BED HIV-1 Capture Enzyme
Immunoassay (BED; the assay that serves as the serologic marker of recent infection) which
measures the concentration of anti-HIV IgG to total IgG in a sample. The result is reported as a
standard optical density (SOD), a continuous measure that describes the relative concentration of
anti-HIV IgG. If the SOD for a given sample is over a threshold predetermined to define a “longstanding” infection then the infection is deemed no longer recent. The period of time during
which the SOD is below this threshold is termed the mean recency period. Although the mean
recency period may vary slightly by HIV subtype, the mean recency period for calculating
population-based incidence estimates in the United States is 162 days using the BED6.
The BED for recency testing is performed only on HIV antibody positive serum2,6 and is not
approved as a diagnostic test. Because of the variability in antibody development among
individuals, the predictive value of an individual’s STARHS result is low; the results are reliable
only as part of the population-based HIV incidence estimate. The Food and Drug Administration
(FDA) has allowed the BED to be labeled “For Surveillance use only. Not for diagnostic or
clinical use,” therefore, STARHS results performed for purposes of HIS cannot be returned to
individuals or their health care providers or used for clinical management. For more information
on the BED please access the following link:
http://www.cdc.gov/hiv/topics/surveillance/resources/factsheets/pdf/bed.pdf
Cleared by HICSB Management 10/15/2012
Appendix 4.1
STARHS Specimen Guidance: Specimens
Originating from the Public Health
Laboratory
HIV Incidence and Case Surveillance Branch
Division of HIV/AIDS Prevention
National Center for HIV, STD, Viral Hepatitis and TB Prevention
Centers for Disease Control and Prevention
Cleared by HICSB Management 10/15/2012
Transportation Overview
In the transportation model where specimens originate from the public health laboratory (PHL),
supplemental/confirmatory testing will have been performed at the state/local PHL. The
state/local PHL will store all positive specimens received until STARHS testing eligibility is
determined. All eligible specimens will be pulled, aliquoted into the designated cryogenic vials
provided by the STARHS laboratory, relabeled with a STARHS identification number, and
shipped to the STARHS laboratory by overnight shipping in accordance with the procedures
described in Appendix 4.4 (Guidelines for Preparing Specimens for Shipping and Transporting
to the STARHS Laboratory).
In this model, the PHL will submit laboratory report information to the state/local HIV
surveillance system according to state HIV reporting requirements, but must also include the
laboratory-assigned specimen accession number and the specimen collection date on the report.
Figure 1 graphically depicts the flow of specimens and reports when specimens originate at the
PHL and are then shipped to the STARHS laboratory.
Role of Public Health Laboratories
The PHL must store all remnant HIV positive specimens until eligibility is determined by the
HIV incidence surveillance staff from the state/local health department. The HIV incidence
surveillance staff will provide the PHL with a list of all STARHS-eligible specimens, listed by
specimen accession number. The PHL will then pull all eligible specimens and aliquot them into
the designated cryogenic vial provided by the STARHS laboratory. The PHL will
simultaneously re-label the eligible specimens with a STARHS identification number (ID) using
labels provided to the PHL by the STARHS laboratory. The PHL must also provide the HIV
incidence surveillance staff with a link between the STARHS identification number and the
original specimen accession number. The PHL will ship all eligible specimens, labeled only
with the STARHS identification number, to the STARHS laboratory according to the procedures
described in the Appendix 4.4 (Guidelines for Preparing Specimens for Shipping and
Transporting to the STARHS Laboratory).
Role of State/Local HIV Incidence Surveillance Staff
The state/local HIV incidence surveillance staff from the jurisdiction where patient specimens
originated will determine STARHS eligibility and coordinate with the PHL to ensure that only
eligible specimens are shipped to the STARHS laboratory. The state/local HIV incidence
surveillance staff will also maintain the link, sent by the PHL, between the original specimen
accession number and the STARHS number, and will manage STARHS results.
Role of STARHS Laboratory
The STARHS laboratory will test all specimens received from a PHL (since only eligible
specimens are sent) using the STARHS identification number. Once testing is complete, the
STARHS laboratory will return results to the appropriate state/local health department.
Cleared by HICSB Management 10/15/2012
Figure 1
1. PHL submits HIV case
report with specimen
accession number,
other laboratory
identifiers and
laboratory identifying
information
2. HIV surveillance
program notifies
PHL of STARHS‐
eligible specimens
by specimen
accession number
PUBLIC HEALTH LABORATORY
Assigns STARHS ID
4. PHL links specimen
accession number with
STARHS ID for all
specimens shipped to
STARHS laboratory
3a.PHL aliquots specimen in correct
tubes
3b. PHL assigns STARHS ID and attaches
labels provided by strip of labels from
STARHS laboratory
3c. PHL ships only eligible specimens to
STARHS laboratory
STARHS
LABORATORY
STATE/LOCAL HIV
SURVEILLANCE
PROGRAM
5. STARHS laboratory sends
results by STARHS ID to
state/local HIV incidence
surveillance staff
Legend
Dotted line represents flow of specimen transport
Solid lines represent flow of information (e.g. laboratory information, STARHS laboratory, etc.)
Cleared by HICSB Management 10/15/2012
Appendix 4.2
STARHS Specimen Guidance: Specimens
Originating from a Private/Commercial
Laboratory and Sent to a Public Health
Laboratory
HIV Incidence and Case Surveillance Branch
Division of HIV/AIDS Prevention
National Center for HIV, STD, Viral Hepatitis and TB Prevention
Centers for Disease Control and Prevention
Cleared by HICSB Management 10/15/2012
Transportation Overview
In the transportation model where specimens originate from a private/commercial laboratory and
are sent to the public health laboratory (PHL), confirmatory/supplemental testing will have been
performed at a private/commercial laboratory. The private/commercial laboratory will send all
HIV positive diagnostic specimens to the state/local PHL. The state/local PHL will store all
specimens received from the private/commercial laboratories until STARHS eligibility is
determined. All eligible specimens will be pulled, aliquoted into the designated cryogenic vials
provided by the STARHS laboratory, relabeled with a STARHS identification number, and
shipped to the STARHS laboratory by overnight shipping in accordance with the procedures
described in Appendix 4.4 (Guidelines for Preparing Specimens for Shipping and Transporting
to the STARHS Laboratory).
In this model, the originating laboratory will submit laboratory report information to the HIV
surveillance system in accordance with state HIV reporting regulations, but must also include the
laboratory-assigned specimen accession number, the specimen collection date, and testing
laboratory identification on the report.
Many laboratories send enzyme immunoassay (EIA) positive specimens to a reference laboratory
for confirmatory/supplemental testing which usually results in different laboratory accession
numbers. In this case, care must be taken to ensure that the appropriate specimen accession
numbers are associated with the correct HIV surveillance case report. Figure 1 graphically
depicts the flow of specimens and reports when specimens originate at a private/commercial
laboratory and are sent to the PHL for storage prior to shipment to the STARHS laboratory.
Role of Private/Commercial Laboratories
Private/commercial laboratories are responsible for forwarding two items for recency testing: (1)
a laboratory report to the public health surveillance department per state HIV reporting
regulations with specimen identifiers, specimen collection date, the laboratory-assigned
specimen accession number, and testing laboratory identification information; (2) remnant HIV
positive specimens to the public health laboratory from specimens labeled with the laboratoryassigned specimen accession number and testing laboratory identification information.
Private/commercial laboratories may elect to aliquot 0.5 ml of the remnant specimen to send to
the PHL so that any additional portion of the remnant specimen may be stored at their facility.
Also private/commercial laboratories may send the entirety of the remnant specimen, without
any further manipulation, to the PHL.
Role of Public Health Laboratories
The PHL must store all remnant HIV positive specimens shipped from a private/commercial
laboratory until eligibility has been determined by the HIV incidence surveillance staff. HIV
incidence surveillance staff will provide the PHL with a list of all STARHS-eligible specimens,
listed by specimen accession number. The PHL will then pull all eligible specimens and aliquot
them into the designated cryogenic vial provided by the STARHS laboratory. The PHL will
simultaneously re-label the eligible specimens with a STARHS identification number (STARHS
ID) using labels provided to the PHL by the STARHS laboratory. The PHL must also provide
the HIV incidence surveillance staff with a link between the STARHS ID and the original
Cleared by HICSB Management 10/15/2012
specimen accession number. The PHL will ship all eligible specimens, labeled only with the
STARHS ID, to the STARHS laboratory according to the procedures described in the Appendix
4.4 (Guidelines for Preparing Specimens for Shipping and Transporting to the STARHS
Laboratory).
Role of State/Local HIV Incidence Surveillance Staff
The state/local HIV incidence surveillance staff from the jurisdiction where patient specimens
originate will determine STARHS eligibility and coordinate with the PHL to ensure that only
eligible specimens are shipped to the STARHS laboratory. The state/local HIV incidence
surveillance staff will also maintain the link between the original specimen accession number
and the STARHS ID, and will manage STARHS results.
Role of STARHS Laboratory
The STARHS laboratory will test all specimens received from a PHL (since only eligible
specimens are sent) using the STARHS identification number. Once testing is complete, the
STARHS laboratory will return results to the appropriate jurisdiction’s state/ local HIV
incidence surveillance staff.
Cleared by HICSB Management 10/15/2012
Figure 1. Specimen Originates at a Private/Commercial Laboratory (for example, a University
Hospital Laboratory, Regional or Local Independent Commercial Laboratory) and Sends
Specimen to State Public Health Laboratory (serves as a pass-through facility)
COMMERCIAL OR PRIVATE
LABORATORY
2. Commercial or private laboratory ships
aliquots of all HIV+ specimens to state PHL
with testing laboratory information &
specimen accession number
PUBLIC HEALTH
LABORATORY (PHL)
Repository for all HIV+ specimens
Stores with Specimen Accession
No.
Assigns Specimen Accession No.
3. State/Local
surveillance program
notifies PHL of
STARHS‐eligible
specimens by
specimen accession
number
1. Commercial or
private laboratory
submits HIV case
report with testing
laboratory
information, specimen
accession number and
other laboratory
identifiers
4a. PHL aliquots specimens in correct tubes
4b. PHL assigns STARHS ID and attaches labels
provided by strip of labels from STARHS
laboratory
4c. PHL ships only eligible specimens to
STARHS laboratory
5. PHL links specimen
accession number with
STARHS ID for all
specimens shipped to
STARHS laboratory
STATE/LOCAL HIV
SURVEILLANCE
PROGRAM
STARHS
LABORATORY
6. STARHS laboratory sends results w/ STARHS ID to HIV
incidence surveillance staff
Legend
Dotted lines represent flow of specimen transport
Solid lines represent flow of information (e.g. laboratory information, STARHS laboratory, etc.)
Cleared by HICSB Management 10/15/2012
Appendix 4.3
STARHS Specimen Guidance: Specimens
Originating from a Private/Commercial
Laboratories and Sent Directly to the STARHS
Laboratory
HIV Incidence and Case Surveillance Branch
Division of HIV/AIDS Prevention
National Center for HIV, STD, Viral Hepatitis and TB Prevention
Centers for Disease Control and Prevention
Cleared by HICSB Management 10/15/2012
Transportation Overview
In the transportation model where specimens originate from a private/commercial laboratory and
are sent directly to the STARHS laboratory, the originating private/commercial laboratory
performing the supplemental/confirmatory testing will send all remnant sera/plasma from HIV
positive diagnostic specimens directly to the STARHS laboratory (by-passing the state /local
PHL) by overnight shipping in accordance with the procedures described in Appendix 4.4
(Guidelines for Preparing Specimens for Shipping and Transporting to the STARHS
Laboratory).
The STARHS laboratory will store specimens until STARHS eligibility is determined by the
state/local HIV incidence surveillance staff, at which point, the STARHS laboratory will pull the
specimens, aliquot, re-label them with a STARHS identification number, and perform recency
testing.
In this model, the originating laboratory will submit case report information to the public health
surveillance department according to state HIV reporting requirements, but must also include the
laboratory-assigned specimen accession number, the specimen collection date, and testing
laboratory identification on the report.
Many laboratories send enzyme immunoassay (EIA) positive specimens to another reference
laboratory for confirmatory/supplemental testing, which usually results in different laboratory
accession numbers. In this case, care must be taken to ensure that the appropriate specimen
accession numbers are associated with the correct HIV case surveillance report. Figure 2 depicts
the flow of specimens and reports when specimens originate at a private/commercial laboratory
and are then shipped directly to the STARHS laboratory.
Locating Specimens
Laboratories that perform routine diagnostic supplemental/confirmatory HIV testing or
immunological status tests such as CD4, or viral load counts should report results to the
state/local health department surveillance program per existing state HIV reporting requirements.
In each surveillance area, all laboratories that perform HIV diagnostic tests for residents of the
surveillance area should be identified from a review of local HIV surveillance data and state
laboratory licensing records. These laboratories should be contacted to request that remnants of
HIV-positive diagnostic specimens be made available for recency testing. The reports that these
laboratories send to the state/local health department surveillance programs per the established
reporting requirements should include specimen accession numbers. Some laboratories perform
only the initial screening HIV EIA test, and then send the specimen to a reference laboratory for
confirmatory HIV testing. Originating laboratories are those where a specimen is first sent for
testing. Reference laboratories are those where a specimen is sent for confirmatory testing when
the originating laboratory does not do confirmatory testing. A specimen may be assigned a
different accession number at each laboratory that receives it; therefore it is important to ensure
that the accession number that is assigned by the laboratory that ships the specimen to the
STARHS laboratory is transmitted to the surveillance program as part of the laboratory reporting
Cleared by HICSB Management 10/15/2012
process. Surveillance areas should maintain a directory of laboratory contacts at all reporting
laboratories to facilitate communication in the event that reporting or shipping of specimens is
disrupted or that changes in policy or procedures need to be communicated.
Determining the Disposition of Specimens and Communicating with the
Public Health and STARHS Laboratory
Private/commercial laboratories that ship directly to the STARHS laboratory will send all
HIV positive specimens. The area HIV incidence surveillance staff, using routine HIV
surveillance reporting procedures, determines whether the specimen represents the person’s
first reported positive HIV test result in the HIS area and is eligible for recency testing.
The HIS program must inform the STARHS laboratory of the disposition of the specimen
because in this transportation model a remnant sample of every HIV positive diagnostic
specimen will be shipped to the STARHS laboratory.
Preparing the STARHS Eligibility Lists
The STARHS laboratory receives shipments of HIV-positive specimens from private and
commercial laboratories. Upon receipt, the specimens are checked against the shipping manifest
to verify that the specimen accession numbers on the tubes match those listed on the manifest.
The verified accession numbers are then entered into the STARHS specimen database by 1)
manually entering individual accession numbers, 2) barcode scanning, if tube labels contain
barcodes, or 3) file importing, if the manifest is provided as an Excel file. Once the specimens
have been accessioned into the database, a list is generated and imported into the STARHS
freezer inventory system to link the freezer location of each specimen to the accession number in
the database.
HIV incidence surveillance staff for each site determines the eligibility of the specimens and
provides this information to the STARHS laboratory in the form of an eligibility list. Upon
receipt of an eligibility list, the STARHS laboratory uploads the list into the specimen database
and searches for the specimen accession numbers that match those on the eligibility list. As
matches are identified, the eligible specimens are located and retrieved from the freezer, and
recency testing is conducted.
In order for the matching process to be successful, eligibility lists that are sent by incidence
coordinators to the STARHS laboratory must be produced using a standard format (Table 1). A
standard template for preparing eligibility lists has been developed as an Excel file named ‘2011
CDC STARHS Eligibility list_v0511’. The first tab of this Excel spreadsheet is the template for
producing eligibility lists. The second tab contains the list of standard codes for the state/site
abbreviations and laboratory names. The organization of the spreadsheet and the contents must
adhere to this standard format to allow for uploading of the data into the STARHS specimen
database at the STARHS laboratory.
Cleared by HICSB Management 10/15/2012
Table 1. STARHS Eligibility List Template
2nd Accession #
Accession
Format
#
(If applicable)
Lab Name
(Use standard
lab names
only)
State
(Use
standard
codes only)
Eligibility
Record 'Eligible' or
'Ineligible'
(Sort by eligibility
status)
Eligible
Ineligible
Previously
Requested?
(Indicate only if
answer is 'Yes')
HIV incidence surveillance staff should use only the template from the 2011 CDC STARHS
Eligibility list_v0511.xls provided by the STARHS laboratory for producing eligibility lists. The
template should not be modified. The order of the columns must remain as it appears in the
template or the file will not function properly during the upload process.
For each specimen accession number listed in the Accession Number column, the State and
Eligibility columns must be completed. There should be no empty cells in these columns.
Columns entitled 2nd Accession Number, Lab Name, and Previously Requested should be
completed if applicable, but may have empty cells. See details for each column below.
State: In the State column, HIV incidence surveillance staff should use only the standard code
for their site listed in the ‘standard codes’ tab of the spreadsheet (Table 2). This code should be
completed for each accession number on the list.
Eligibility: In the Eligibility column, HIV incidence surveillance staff should list either
‘Eligible’ or ‘Ineligible’ for each accession number on the list. The terms ‘Test’ and ‘Toss’ have
been used by some in the past, but these should no longer be used.
Lab Name: In the Lab Name column, HIV incidence surveillance staff should use only the
abbreviated laboratory name that is listed in the ‘standard codes’ tab of the spreadsheet (Table 3).
The laboratory name should only be added if it appears on the list. If a new laboratory comes on
board and an abbreviated laboratory name has not been added to the list yet, the Lab Name cell
should be left blank for that accession number. The list will be updated, as needed, and
notification of updated laboratory names will be provided.
2nd Accession Number: If there is a second accession number associated with a particular
specimen, that number should be added to the 2nd Accession Number column. If there is no
additional accession number for a specimen, this cell may be blank.
Previously Requested: In general, once a specimen accession number has appeared as ‘Eligible’
or ‘Ineligible’ on an eligibility list, there should be no need to include it on future lists.
However, if results have not been received after 3 months, it may be necessary to request the
specimen again using the eligibility list. If an accession number that was requested on a previous
eligibility list is included, HIV incidence surveillance staff should indicate ‘Yes’ in the
Cleared by HICSB Management 10/15/2012
Previously Requested column. In some cases the format of the accession number in the
eligibility list may not match exactly with the number in the STARHS laboratory specimen
database. By indicating ‘Yes’ in the Previously Requested column, additional procedures will
be implemented to try to locate the specimen for testing.
Organizing the Eligibility List File for Sending to the STARHS laboratory
After the eligibility list has been completed, the entire list should be sorted using the Eligibility
column so that all of the ‘Eligible’ specimens are grouped together and all of the ‘Ineligible’
specimens are grouped together.
The file should be encrypted as a PGP zip file, preferably as a self-decrypting PGP zip file to
allow the file to be opened on a computer that doesn't have PGP software. The STARHS
laboratory does not use the PGP key system.
Each site should use their site’s standard password to protect their files. This makes it much
easier if the STARHS laboratory needs to re-open the initial list on the CD at a later time.
The encrypted file must be saved to a CD, and labeled with the site name and the date. The
STARHS laboratory receives disks from multiple sites and it is important for the disk to be
labeled so that it can be easily identified.
Untested Specimen Lists
An Untested Specimen List is produced and distributed by the STARHS laboratory to all sites
at least quarterly. This list contains the accession numbers of specimens that have been received
by the STARHS laboratory and have not been tested yet.
HIV incidence surveillance staff may request testing of a specimen listed on the Untested
Specimen List, using the Eligibility List template following all standard procedures for
preparing the eligibility list, as described above.
The specimen accession number is recorded in the Eligibility List template exactly as it appears
on the Untested Specimen List. The list of accession numbers on the Untested Specimen List
is generated directly from the STARHS laboratory specimen database. This will ensure that the
accession number on the eligibility list is in the same format as it appears in the database,
allowing it to be readily located and tested.
Table 2. List of Standard Codes for State/Site Abbreviations
Code
AZ
CHI
CT
FL
State/Site
Arizona DOH
Chicago DOH
Connecticut DOH
Florida DOH
Cleared by HICSB Management 10/15/2012
Code
WA
MI
PHI
SC
State/Site
Washington DOH
Michigan DOH
Philadelphia DOH
South Carolina DOH
HOU
LA
MA
MS
NJ
NYC
NYS
TX
VA
Houston DOH
Louisiana DOH
Massachusetts DOH
Mississippi DOH
New Jersey DOH
NY City DOH
NY State DOH
Texas DOH
Virginia DOH
LAC
SAN
DC
NC
SEA
CO
IN
CA
AL
L A County DOH
San Francisco DOH
Washington DC DOH
North Carolina DOH
Seattle DOH
Colorado DOH
Indiana DOH
California DOH
Alabama DOH
Table 3. List of Standard Codes for Laboratory Names
No.
Standard Name
Laboratory
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
ARUP
MAYO
QUEST
BIO-REFERENCE
PRINCETON
SPECIALTY
LABCORP
WARDE
SUNRISE
NASSAU
UNC
UVAHS
CRL
HERITAGE
MCV
DUKE
ECMC
SBUMC
KALEIDA
STLUKES
WESTCHESTER
MARICOPA
BSI
PAML
ARUP Laboratories
Mayo Medical Laboratories
Quest Diagnostics
Bio Reference Laboratory
Princeton Biomedical Laboratories
Specialty Laboratories, Inc.
Laboratory Corporation of America
Warde Medical Laboratory
Sunrise Medical Laboratory
Nassau County Dept of Health Lab
University of North Carolina @ Chapel Hill
University of Virginia Health System
Clinical Reference Laboratory
Heritage Labs International
Virginia Commonwealth University Health System
Duke University Medical Center ,Durham NC
Erie County Medical Center
Stony Brook University Medical Center
Kaleida Health
St Luke's Roosevelt Laboratory
Westchester County Labs & Research
Maricopa Integrated Health System, Phoenix AZ
Bendiner & Schlesinger Inc.
Pathology Associates Medical Laboratories
Cleared by HICSB Management 10/15/2012
25
26
27
28
29
30
31
UNYTS
ARC
KAISER
SFGH
CDD
ALAMEDA
SANDIEGO
Upstate New York Transplant Services
National Confirmatory American Red Cross
Kaiser
San Francisco General Hospital
Center for Disease Detection, San Antonio TX
ALAMEDA County Public Health Lab, Oakland CA
San Diego County Public Health Laboratory, CA
Role of Private/Commercial Laboratories
The private/commercial laboratories are responsible for forwarding two items for recency
testing: (1) a laboratory report to the HIV surveillance system per local requirements, including
the collection date, the laboratory-assigned specimen accession number, and identification
information about the testing facility; (2) remnant HIV positive serum/plasma from diagnostic
specimens labeled with the laboratory-assigned specimen accession number.
The private/commercial laboratories may elect to aliquot 0.5 ml of the remnant specimen for
shipment to the STARHS laboratory so that any additional portion of the remnant specimens
may be stored at their facility, or they may send the entirety of their remnant specimens, without
any further manipulation, to the STARHS laboratory.
Role of STARHS Laboratory
The STARHS laboratory must store all remnant HIV positive specimens received until eligibility
has been determined by the appropriate jurisdiction’s HIV Incidence Surveillance staff. The
HIV incidence surveillance staff will provide the STARHS laboratory with a list of all specimens
eligible to be tested for recent infection, listed by specimen accession number. If eligibility has
not been determined after 2 years from the time the specimen was received in the STARHS
laboratory, the specimens may be destroyed according to established laboratory methods.
The STARHS laboratory will pull all eligible specimens and aliquot them into the designated
cryogenic vial for testing. The STARHS laboratory will simultaneously re-label the eligible
specimens with a STARHS identification number (STARHS ID). The STARHS laboratory will
provide the appropriate HIV incidence surveillance staff with a link between the STARHS
identification number and the original specimen accession number. All links at the STARHS
laboratory are maintained on a secure internal network that ensures that confidentiality is
protected and maintained to meet the standards for HIV surveillance listed in National Center for
HIV/AIDS, Viral Hepatitis, STD, and TB Prevention’s (NCHHSTP) Data Security and
Confidentiality Guidelines for HIV, Viral Hepatitis, Sexually Transmitted Disease, and
Tuberculosis Programs: Standards to Facilitate Sharing and Use of Surveillance Data for
Public Health Action. The STARHS laboratory will test all eligible specimens and send results
(using only the STARHS ID) to the HIV incidence surveillance staff from the appropriate
jurisdiction. The STARHS results are for surveillance purposes only therefore results will not be
reported back to the originating laboratory, provider, or client.
Cleared by HICSB Management 10/15/2012
The STARHS laboratory will send each HIV incidence surveillance jurisdiction an “Untested
Specimens List” at least quarterly. HIV incidence surveillance staff should examine the list to
identify remnant HIV-positive specimens sent directly to the STARHS laboratory and
representing diagnostic specimens from newly reported cases of HIV infection. These
specimens should be listed on the STARHS Eligibility List.
Role of the State/Local HIV Incidence Surveillance Staff
State/local HIV incidence surveillance staff from the jurisdiction where a specimen originated
will determine STARHS eligibility and coordinate with the STARHS laboratory to ensure that
proper specimens are pulled and tested. State/local HIV incidence surveillance staff will also
maintain the link between the original specimen accession number and the STARHS ID number,
and will manage the STARHS results.
Cleared by HICSB Management 10/15/2012
Figure 1. Specimen Originates at National Commercial Laboratory or Smaller Private
Laboratory and is Sent Directly to the STARHS Laboratory
COMMERCIAL OR PRIVATE
LABORATORY
1 . Commercial or
private laboratory
submits HIV case report
with testing laboratory
information, specimen
accession number and
other laboratory
identifiers
STATE/LOCAL HIV
SURVEILLANCE
PROGRAM
Assigns Specimen Accession No.
2. Commercial or private laboratory
sends aliquots of all HIV+ specimens
with specimen accession number),
collection date, and laboratory
identifiers to the STARHS laboratory
3. State/Local surveillance program notifies
STARHS laboratory of STARHS‐eligible specimens
by specimen accession number
5. STARHS laboratory links specimen accession
number with STARHS ID for all specimens
6. STARHS laboratory sends results w/ STARHS ID to
HIV incidence surveillance staff
STARHS
LABORATORY
4a. STARHS lab pulls and re‐
labels eligible specimens.
4b. STARHS ID assigned to
specimen accession number
and aliquot at STARHS lab.
4c. Specimen tested by STARHS
ID.
Legend
Dotted line represents flow of specimen transport
Solid lines represent flow of information (e.g. laboratory information, STARHS laboratory, etc.)
Cleared by HICSB Management 10/15/2012
Appendix 4.4
Guidelines for Preparing Specimens for
Shipping and Transporting to the STARHS
Laboratory
HIV Incidence and Case Surveillance Branch
Division of HIV/AIDS Prevention
National Center for HIV, STD, Viral Hepatitis and TB Prevention
Centers for Disease Control and Prevention
Cleared by HICSB Management 10/15/2012
Guidelines for Preparing Specimens for Shipping and Transporting to the
STARHS Laboratory
Purpose
This standard operating procedure describes methods for the handling, storage, and shipping
of serum/plasma specimens that will be tested for recent HIV-1 infection using STARHS.
Results from these tests will help estimate HIV incidence.
Introduction
Surplus serum/plasma from positive HIV diagnostic specimens is to be collected, frozen, and
shipped to the STARHS laboratory. Public health laboratories (PHL) that ship only specimens
eligible to be tested for recent infection must use vials and labels specified and supplied by the
STARHS laboratory. Commercial/private laboratories that ship all remnant HIV-positive
specimens to the STARHS laboratory may choose to freeze and ship specimens using the
original primary specimen tube or aliquot tubes provided by the commercial/private laboratory
and labeled with that laboratory’s specimen accessioning label.
Type of Specimens Shipped to STARHS Laboratory
HIV positive serum or plasma from confirmed HIV positive diagnostic samples will ultimately
be shipped to the STARHS laboratory.
Specimen Volume
The optimal quantity of serum/plasma required for recency testing is 0.5ml per aliquot.
However, if less than 0.5ml of the remnant sample is available for recency testing the sample
should still be sent to the STARHS laboratory. The STARHS laboratory is the only
laboratory that should determine if a sample should be rejected due to insufficient
quantity.
Sample Storage
Short-term (less than one week) storage of samples in the refrigerator (2° – 8°C) is acceptable,
but for long-term storage (more than one week), samples must be frozen at -20°C or colder.
Long-term storage includes any period of time that the samples are kept at the originating/testing
laboratory or a “pass through” PHL prior to shipment to the STARHS laboratory, or the interim
period while STARHS eligibility is being determined. Effort should be made to avoid repeated
freezing and thawing of samples, as repeated freeze-thaw cycles may lead to erroneous results.
The freezer must contain adequate space to store specimens. It is recommended that, if not
already in practice, a daily temperature log be kept to ensure the freezer is operating properly.
The freezer should be housed in a location with proper ventilation to avoid overheating and
freezer failure.
Cleared by HICSB Management 10/15/2012
Specimen Numbering
The specimen number on the samples shipped to the STARHS laboratory will either be the
original laboratory-assigned specimen accession number or the STARHS identification number
(STARHS ID) depending on the transport model. Detailed information about specimen
numbering is included in Appendices 4.1-4.3 of this guide.
Specimen Retention
State/local HIV incidence surveillance staff must coordinate with the laboratory storing HIV
positive remnant sera (the STARHS laboratory and/or the state/local PHL) to identify eligible
samples to be tested for recent infection using STARHS. However, not all stored samples will
be eligible for recency testing, and will therefore have to be identified for disposal. State/local
HIV incidence surveillance staff and the storing laboratory should communicate regularly (every
1-3 months) to discuss any specimens that were not on the eligibility list to determine if the
sample can be disposed of or if further investigation is needed. Samples should not be destroyed
or disposed of until eligibility is definitively determined. Specimens at the STARHS laboratory
that have not been requested for testing after a period of two years, may be discarded.
Shipping Program
Since 2008, the STARHS laboratory has directed the shipping and transport of specimens
through the STARHS specimen shipping program. The STARHS specimen shipping program
includes the provision of shipping materials and a courier account for overnight shipping of
specimens to the STARHS laboratory. All laboratories, public health, commercial and private,
that ship specimens to the STARHS laboratory are eligible to utilize this shipping program.
Specimens must be shipped on dry ice by same-day or overnight shipping. Specimens may be
shipped from originating labs to the state PHL which may serve as a “pass through” facility or
may be shipped to the STARHS laboratory as ‘Diagnostic Specimens.’ All laboratories shipping
HIV positive samples must be certified to ship dangerous goods and observe the dry ice
requirement, regardless of where specimens are shipped.
The STARHS laboratory will provide shipping materials (shipping containers, bags/containers
and absorbent materials required per federal regulations, as well as fiberboard specimen boxes
and all required shipping labels) and a courier account number for overnight shipping of
specimens. Dry ice will not be provided by the STARHS laboratory. Each participating
laboratory must identify at least one contact person for shipping samples. The STARHS
laboratory will provide instructions to this individual on how to access the courier account and
utilize the system.
Frequency of Shipments
The frequency of specimen shipments to the STARHS laboratory will be determined by the
shipping laboratory. To determine this frequency the shipping laboratory should consider factors
such as specimen retention policies and freezer/storage space and consult with the STARHS
laboratory delivery service to ensure that specimens do not thaw in transit.
Cleared by HICSB Management 10/15/2012
Additional Information for All Other Private Laboratories
Intra-state shipments from private laboratories to the PHL may be shipped by Federal Express
(or a similar commercial courier) or an established local courier service. Funding permitting,
sites may elect to set-up a billing account with Federal Express (or a similar commercial courier)
to off-set costs incurred by the private laboratory for shipping to the state/local PHL.
Tracking Shipments
The shipping laboratory should notify the receiving laboratory (state PHL or STARHS
laboratory) by fax or email when specimens are shipped, including the tracking number. The
receiving laboratory will be responsible for tracking the shipments and will notify the originating
laboratory if the specimens are not received.
Sample Rejection Criteria
Sample rejection due to thawing, breakage, insufficient quantity, or lost-in-transit will be
determined and recorded by the STARHS laboratory and will be reported to the appropriate HIV
surveillance system.
Confidentiality and HIPAA Regulations
STARHS testing must ensure that confidentiality is protected and maintained to meet standards
for HIV surveillance. The Privacy Rule of the Health Insurance Portability and Accountability
Act (HIPAA) regulations permit protected health information to be shared for the purposes of
public health surveillance activities, allowing the originating laboratories to send specimens
labeled with their laboratory assigned accession number to either the state PHL or the STARHS
laboratory, where eligible samples will be re-assigned a unique STARHS identification number
prior to testing for recent infection. This process will minimize re-labeling errors and also
simplify the shipment procedures for private laboratories. However, the state/local public health
department must have the laboratory accession number to link the test result to the patient
information in the surveillance record. Therefore, the laboratory accession number must be
included on the HIV laboratory report form sent to the state/local public health department by the
originating laboratory.
Preparing and Packing Specimens for Shipping to the STARHS Laboratory
Setting and personnel for specimen processing
1. Centrifugation, aliquoting, and shipping should be performed at or under the auspices of
a laboratory that is CLIA- certified for handling HIV-positive specimens.
2.
All personnel handling specimens should receive blood borne pathogens training. See
OSHA’s Occupational Exposure to Bloodborne Pathogens Standard:
http://www.osha.gov/pls/oshaweb/owadisp.show_document?p_table=STANDARDS&p_
id=10051.
Cleared by HICSB Management 10/15/2012
3.
Personnel handling or processing specimens should have appropriate laboratory training
in the relevant laboratory techniques for handling HIV-positive specimens.
4. The setting in which centrifugation, aliquoting, and shipping occurs should meet
Biosafety level 2 specifications required by the U.S. Department of Health and Human
Services for handling of specimens containing HIV
[http://www.cdc.gov/od/ohs/biosfty/bmbl4/bmbl4toc.htm, Biosafety in Microbiological
and Biomedical Laboratories (BMBL) 5th Edition].
Required Materials
1. Cryogenic vials - Supplied by STARHS laboratory to PHL only.
2. Specimen labels - Supplied by STARHS laboratory to PHL only. Label will identify the
sample (includes barcode and number).
3. Cardboard storage boxes for cryogenic vials with divider insert – Supplied by STARHS
laboratory to all laboratories except when not wanted.
4. A supply of dry ice in pellet form - not supplied by STARHS laboratory.
5. Insulated shipping containers certified to ship frozen diagnostic specimens (HIV+ serum
and dry ice) - Supplied by STARHS laboratory.
6. STARHS/FedEx account access - Supplied by STARHS laboratory.
7.
Materials for shipper packing - Supplied by STARHS laboratory.
8.
STARHS specimen submission form.
Specimen Collection and Processing
1. All processing of specimens should be done by personnel qualified to handle HIVpositive specimens under the auspices of a laboratory equipped for the handling of
HIV-positive specimens [see http://www.cdc.gov/od/ohs/biosfty/bmbl4/bmbl4toc.htm,
th
Biosafety in Microbiological and Biomedical Laboratories (BMBL) 5 Edition ].
1.1
Aliquot the serum (0.5 ml per cryogenic vial). Use labels to identify the specimen
and record this information in the proper setting. (Specimen log for eventual
transfer to access laboratory database for linked samples and temporary database
for unlinked specimens)
1.2
Store aliquots in refrigerator or freezer until eligibility requirements are met and
scheduled shipping date has arrived. Samples for recency testing can be
refrigerated at 2-8oC, but for long term storage and shipping, samples should be
frozen at -20oC.
1.2.1
It is recommended that, if not already in practice, a daily temperature log
be kept to ensure the freezer is operating properly.
Cleared by HICSB Management 10/15/2012
1.2.2
The freezer should be housed in a location with proper ventilation to avoid
overheating and freezer failure.
1.2.3
Staff must ensure adequate freezer space to store specimens.
Shipping
1. Specimens for recency testing should be sent to the STARHS laboratory using the
address provided on the STARHS specimen submission form and under STARHS
laboratory in this document. All specimens will be shipped as diagnostic specimens
using International Air Transport Association (IATA) Packing Instructions 650. Dry ice
will be included with each shipment using IATA Packing Instructions 904.
1.1
Because samples will be shipped using dry ice, shipping personnel must be
trained and certified to ship dangerous goods. See below for a list of
companies that provide training.
1.2
Establish contact with N’ko Lea Ali-Napo ([email protected]) at the
STARHS laboratory.
1.3
An invitation to set up a STARHS / FedEx account for the laboratory will be
provided by the STARHS laboratory.
1.4
A fax or e-mail system should be set up to notify the STARHS laboratory of
incoming shipment.
1.5
Adequate STP320 or equivalent shipping containers must be available. The
STARHS laboratory will return them to the shipping laboratory. These are
expensive shippers and need to be re-used.
1.6
Laboratories should ensure access to the STARHS FedEx account and that they
are able to print shipping courier air bills.
2. Packing and shipping dates should be determined in consultation with N’ko Lea AliNapo at Wadsworth Center Bloodborne Viruses Laboratory. It is recommended that staff at
the laboratory read and walk through all of the following steps before starting the
preparation of the actual shipment to ensure familiarity with the shipping procedures and
requirements. Specimens should not be shipped on Friday or on the day before an official
holiday. Specimens should remain frozen at all times and therefore should not arrive at
the Wadsworth Center when the center is closed.
STARHS Medium Shipper Instructions
The medium shipper will accommodate 2 fiberboard freezer boxes that will hold up to 81
specimens in each box for a maximum total of 162 specimens. All unused freezer boxes, plastic
bags, and Tyvek® envelopes should be left in the box when shipping specimens to the STARHS
laboratory.
Supplies included in each shipper:
1 - Category B label
Cleared by HICSB Management 10/15/2012
1 - UN1845 dry ice label
1 - Class 9 hazard miscellaneous goods label
2 - Fiberboard freezer boxes with divider insert
2 - Large Inner leak proof polybags with absorbent strip (instructions printed on polybag)
2 - Large Tyvek® outer envelopes (instructions printed on envelope)
DIRECTIONS TO PACK MEDIUM SHIPPER
1. Load specimen tubes into the box in the order listed on the manifest.
2. Place the box in the polybag and seal it according to the instructions on the polybag.
3. Insert the polybag and box into the Tyvek® envelope and close the envelope according to
the instructions on the envelope
4. Return both fiberboard specimen boxes regardless of whether they have been used.
5. If both boxes are not used, place the empty box on the bottom of the insulated box and
place the box containing specimens on top.
6. Add the appropriate amount of dry ice.
7. Include a copy of the STARHS Specimen submission form and manifest containing
the specimen list.
8.
Seal the closed box with shipping tape
9.
Place labels (Category B, UN1845, and Class 9) and the FedEx airbill on the box
STARHS Large Shipper Instructions
The large shipper will accommodate 4 fiberboard freezer boxes that will hold up to 81 specimens
in each box for a maximum total of 324 specimens. Please leave all unused freezer boxes, plastic
bags, and Tyvek® envelopes in the box when shipping specimens to the STARHS laboratory.
Supplies included in each shipper:
1 - Category B label
1 - UN1845 dry ice label
1 - Class 9 hazard miscellaneous goods label
4 - Fiberboard freezer boxes with divider insert
4 - Large Inner leak proof polybags with absorbent strip (instructions printed on polybag)
4 - Large Tyvek® outer envelopes (instructions printed on envelope)
Cleared by HICSB Management 10/15/2012
DIRECTIONS TO PACK LARGE SHIPPER
1. Load specimen tubes into the box in the order listed on the manifest.
2. Place the box in the polybag and seal the polybag according to the instructions on the
polybag.
3. Insert the polybag and box into the Tyvek® envelope and close the envelope
according to the instructions on the envelope.
4. Stack the envelopes with the boxes of specimens into the shipping box.
5. Add the appropriate amount of dry ice around the boxes.
6. Place the foam plug over the specimens.
7. Place any unused supplies on top of the foam.
8. Include a copy of the STARHS Specimen submission form and manifest containing
the specimen list.
9. Seal the closed box with shipping tape.
10. Place labels (Category B, UN1845, and Class 9) and the FedEx airbill on the box
NOTE: In order to conserve supplies it is requested that unused supplies be returned with each
shipment.
Questions should be directed to Thomas T. Miller, [email protected] at (518) 474-2163.
Cleared by HICSB Management 10/15/2012
Important STARHS shipping information
Shipping address
For all specimen shipments for recency testing please be sure to use the correct shipping address.
This correct address is printed on the STARHS shipping manifest. Using other addresses could
lead to delays or lost shipments. Specimen shipments should be sent to:
NYSDOH Wadsworth Center
Axelrod Institute
Bloodborne Viruses Laboratory: STARHS
120 New Scotland Avenue
Albany, NY 12208
Attn: N’ko Lea ALI-NAPO
Email notifications
Notifications of specimen shipments and transmission of manifests from commercial laboratories
by email should be sent using the following email addresses only:
[email protected] (Lea Ali-Napo)
[email protected] (William Carmichael)
[email protected] (Brian Granger)
Both of these email addresses should be included when sending an email notification of
shipment. Do not include any other email addresses previously associated with the STARHS
laboratory when sending shipping manifests. It is important that the manifests and other email
notifications regarding shipments be directed only to current staff at the STARHS laboratory.
FAX notifications
FAX notifications of upcoming STARHS shipments should be addressed to Lea Ali-Napo or
William Carmichael. Please continue to use the FAX number (518) 473-0008 for this purpose.
Primary contact for STARHS
N’ko Lea Ali-Napo (Lea), [email protected]
Cleared by HICSB Management 10/15/2012
STARHS Shipment Contact
STARHS / FedEx account access
Thomas T. Miller, [email protected]
Shipment notification and shipment
N’ko Lea Ali-Napo, [email protected]
William Carmichael, [email protected]
Brian Granger, [email protected]
Shipping Address
NYSDOH Wadsworth Center
Axelrod Institute
Bloodborne Viruses Laboratory: STARHS
120 New Scotland Avenue
Albany, NY 12208
Phone (518) 474-2163
FAX (518) 473-0008
Director of Blood-borne Diseases Laboratory
Monica M. Parker, Ph.D., [email protected]
Cleared by HICSB Management 10/15/2012
STARHS Specimen Submission Form
Please complete this form and send it with each shipment. Specimens should be sent to:
NYSDOH Wadsworth Center
Axelrod Institute
Bloodborne Viruses Laboratory: STARHS
120 New Scotland Avenue
Albany, NY 12208
Attn: N’ko Lea ALI-NAPO
SHIPPING FACILITY INFORMATION: RESULTS SENT TO:
Name:________________________ ________ Name________________________________________
Address:_______________________________ Address______________________________________
_______________________________ _____________________________________
Phone Number:_________________________ Phone Number:_______________________________
Fax:__________________________________ Fax_________________________________________
Email:________________________________ Email:_______________________________________
Contact Person:_____________________________
INCIDENCE SURVEILLANCE (HICSB) - List of eligible specimens sent separately
Please place the labels on the tubes with left side even with the bottom of the tube threads.
RANGE OF SPECIMENS SENT:
________________________________________________________________________
________________________________________________________________________
________________________________________________________________________
________________________________________________________________________
FOR NYS DOH USE:
Date Rec’d.
Date Accessioned:
Date Tested:
Date Reported:
Date Specimens Returned:
Date Shipping Container(s) Returned:
Cleared by HICSB Management 10/15/2012
Training and Certification for Shipping Infectious Substances
The following companies provide training for dangerous goods shipping, however, the Centers
for Disease Control and Prevention does not endorse any particular company.
FedEx 800-GO-FEDEX 3 day IATA based training Covers all hazardous materials Cost is $550
Saf-T-Pak 800-814-7484 specifically for infectious and diagnostic substances, and dry ice 3
options---One day seminar, On-site programs, or Interactive CD For interactive CD: for one
sitting, can be done in 3-5 hours. Certificate good for 2 years OR until regulations change Cost is
~$250
Viking Packaging (Oklahoma) 800-788-8525—David Weilert Seminars monthly in Tulsa/ $300
per person Covers all nine classes of hazardous materials Covers shipping under IATA
Certificate good for 2 years Will do group classes in local area---$3,000 plus travel costs
Cleared by HICSB Management 10/15/2012
Guidance for Collection and
Data Entry of HIV Incidence
Surveillance Information
Version 1.2
Contents
Introduction ................................................................................................................... 4
Background .................................................................................................................. 4
Transition to a New HIS Database ................................................................................... 4
Changes in Data Analysis ............................................................................................... 5
Changes in Data Collection Forms ................................................................................... 5
Changes in the HIS Database ......................................................................................... 5
Document-based Data Collection and Data Entry of TTH .................................................... 5
Data Collection ............................................................................................................ 6
Indeterminate HIV Test Results ..................................................................................... 7
Data Entry.................................................................................................................... 7
Data Quality ................................................................................................................. 7
General Guidance .......................................................................................................... 7
Dates .......................................................................................................................... 8
Format of this Document ................................................................................................ 8
1. Critical Data Elements for Testing & Treatment History (TTH) ................................ 10
1.1
Main Source of TTH Information ........................................................................... 10
1.2
Date Patient Reported Information ........................................................................ 12
1.3
Ever had a Previous Positive HIV Test? .................................................................. 14
1.4
Date of First Positive HIV Test .............................................................................. 17
1.5
Ever had a Negative HIV Test? ............................................................................. 19
1.6
Date of Last Negative HIV Test ............................................................................. 21
1.7
Number of Negative HIV Tests within 24 months before First Positive HIV Test ........... 24
1.8
Ever Taken Any Antiretroviral Medications (ARVs)? ................................................. 27
1.9
Name(s) of ARV Medication Taken ........................................................................ 30
1.10 Date ARVs First Began ......................................................................................... 32
1.11 Date of Last ARV Use .......................................................................................... 33
2. Legacy Data and Other Optional TTH Data Elements ............................................... 36
2.1
Are You Now Taking Any ARVs? ............................................................................ 36
2.2
Ever Tested for HIV Before Today? (Legacy Pre-test form) ...................................... 36
2.3
Date of First HIV Test Ever ................................................................................... 37
2.4
Was the First Positive HIV Test Anonymous? .......................................................... 37
2.5
Number of Tests 2 years before Previous Positive HIV Test (Legacy Pre-test form) ...... 38
2.6
Reason for Today’s HIV Test (Legacy Pre-test form) ............................................... 38
2.7
Reason for First Positive HIV Test ......................................................................... 40
June 2012
2
2.8
Name of Facility Where First Tested Positive for HIV ................................................ 41
2.9
State of Facility Where First Tested Positive for HIV ................................................ 41
2.10 Type of Facility Where First Tested Positive for HIV ................................................. 42
2.11 Name of Facility Where Last Tested Negative for HIV............................................... 42
2.12 State of Facility Where Last Tested Negative for HIV ............................................... 42
2.13 Type of Facility Where First Tested Negative for HIV ................................................ 43
3. IVR Database TTH Variables Not Included in eHARS ............................................... 43
3.1
Date of HIV Test (reference date) ......................................................................... 43
4. Required and Optional Data Elements for STARHS Specimen Information
and Results.......................................................................................................... 44
4.1
STARHS Laboratory Name .................................................................................... 44
4.2
Source Lab Specimen ID ...................................................................................... 45
4.3
State Lab Specimen ID (or Other Specimen ID) ...................................................... 46
4.4
Date of Specimen Collection (Required for HIS) ...................................................... 46
4.5
STARHS Result Date ............................................................................................ 47
4.6
Received Date .................................................................................................... 47
4.7
STARHS Assay (Required for eHARS) .................................................................... 48
4.8
Specimen Type (Recommended for HIS) ................................................................ 49
4.9
STARHS ID (Required for eHARS and HIS) ............................................................. 50
4.10 Standard Optical Density ..................................................................................... 50
4.11 Final STARHS Result (Required for HIS) ................................................................. 51
4.12 Reason Specimen Not Sent for STARHS ................................................................. 52
5. IVR Lab Variables Not Included in eHARS ............................................................... 52
5.1
Specimen Approved for STARHS ........................................................................... 52
5.2
State Lab ID....................................................................................................... 53
5.3
HIV Diagnosis Test Type ...................................................................................... 53
5.4
Results Received ................................................................................................. 53
Appendix A: Quick Reference for Data Analysts .......................................................... 56
Appendix B: List of ARV Medications for TTH ............................................................... 62
Appendix C: Data Entry Recommendations for HIV Incidence Surveillance Data
Elements in eHARS……………………………………………………………………………………………..64
June 2012
3
Introduction
Since 2001, the Centers for Disease Control and Prevention (CDC) has funded a number of state and
local health departments to conduct HIV incidence surveillance (HIS) through the collection of
information on HIV testing and treatment history (TTH) and serologic testing algorithm for recent HIV
seroconversion (STARHS) results to generate national, state and local HIV incidence estimates. Since
the implementation of HIV incidence surveillance, practices have evolved as a result of advances in
testing technology, availability of additional information sources and better understanding of the data
necessary to estimate HIV incidence.
This document describes guidance for HIS data collection and data entry within the national
HIV/AIDS reporting system. In addition to describing more specific information about the purpose and
interpretation of HIS data elements, this guidance describes the wording, variable names, and available
values for TTH and STARHS data in eHARS. In this document, the terms person, patient and
individual are used interchangeably when referring to persons newly diagnosed with HIV.
Background
The original HIS database was designed to collect information on demographics, testing and treatment
history, laboratory specimens, STARHS results, and consent. Initially, STARHS used the Vironostika
HIV-1 EIA (Vironostika-LS) assay for which the Food and Drug Administration (FDA) required
individual consent. TTH was collected during an interview with persons seeking HIV testing (pre-test)
or after receiving a new HIV diagnosis (post-test). In March 2005, the FDA allowed the use of a new
assay, the BED HIV-1 Capture enzyme immunoassay (BED), for surveillance purposes only, which
eliminated the need for patient consent. In 2007, CDC reduced the number data elements required for
TTH and expanded the collection of TTH data elements to include provider reports, medical record
review, other databases and the National HIV Monitoring and Evaluation or Program Evaluation and
Monitoring System (NHM&E/PEMS). Most health departments funded for HIS used the post-test data
collection form and collected data at various times, not exclusively at the time of testing.
Transition to a New HIS Database
From the implementation of HIS in 2001 through most of 2011, incidence surveillance data were
stored in the IVR database. In late 2011, HIS data were merged into eHARS, a document-based system
used for storing all HIV surveillance information and producing datasets for local analysis and transfer
to CDC. The IVR database required the entry of pre-test and post-test data, STARHS specimen
information, and import of STARHS laboratory results on separate forms and tables.
In eHARS, there is one TTH document type, not separate pre-test or post-test documents. An eHARS
laboratory document is used both for specimen information prior to receiving the STARHS result and
STARHS results returned from the CDC STARHS laboratory. A case can have multiple TTH and
laboratory documents and all data can be imported into eHARS, using All Document Import (ADI).
Refer to the current eHARS Technical Reference Guide for more information on ADI. At conversion,
pre-test and post-test information in the IVR database were converted to TTH documents in eHARS.
For cases with more than one TTH document in the IVR database (e.g., a pre-test and a post-test
document), separate TTH documents were stored in eHARS. Specimen information and STARHS
results are now entered or imported to Laboratory documents, and HIS areas must ensure that the
STARHS laboratory and any other laboratories that performed STARHS testing in the past were added
by the eHARS Administrator to the local Laboratory Name list for the eHARS variable Lab Name
(CLIA_ID) (See Section IV: Required and Optional Data Elements for STARHS Specimen
Information and Results). At conversion, all pre-test and post-test data were converted to eHARS TTH
June 2012
4
documents, with variable names the same as the IVR pre-test variable names, but some values may be
different. Project areas have to modify previous analysis programs to reflect the new names. This
guidance should be useful in making these modifications.
After conversion, all data are transferred out of eHARS to CDC, but monthly HIS datasets are required
until National Data Processing (NDP) is in place. After that, quarterly datasets will be required. HIS
programs will stop using the IVR database for incidence data, but some areas will continue using it for
Variant, Atypical and Resistant HIV Surveillance (VARHS) until the incorporation of data from
VARHS into eHARS.
Changes in Data Analysis
Currently, TTH variable values do not appear in eHARS Person View (PV), the case-level summary of the
‘best’ information collected for each person that appears on the eHARS screen, but all lab documents with
STARHS tests are listed. All data analyses can be performed using CDC provided programs that extract
datasets from the eHARS Document datasets and Person-based dataset. Eventually, some calculated TTH
variables may be included in PV and the Person-based datasets, based on a hierarchy of the best
information from multiple documents. Surveillance staff will also be able to use the override option to
select a different value for a particular variable from multiple documents. In the interim, CDC will provide
guidance and SAS programs for selecting incidence information from multiple documents for incidence
estimation and other data analysis.
Appendix A comprises a Quick Reference for Data Analysts, a guide for data managers and programmers
who monitor and analyze incidence data. It provides IVR and eHARS variable names, labels, values, and
formats of stored data, as well as the SAS labels that are found in eHARS document-based datasets.
Changes in Data Collection Forms
This document describes changes in data collection that may require project areas to modify local data
collection instruments. Surveillance areas may continue to collect data using current forms until new
ones are developed. In addition, the revised adult case report form (ACRF) has a block of data elements
called HIV Testing and Antiretroviral Use History that reflect the information in this guidance.
Currently, information from this ACRF section must be entered in eHARS on a separate TTH document.
Changes in the HIS Database
The eHARS TTH document looks different from the IVR pre-test and post-test screens. Clicking the
TTH Data tab reveals first the required TTH fields (standard data elements described in Technical
Guidance for HIV/AIDS Surveillance Programs), followed by the optional/legacy fields. The data fields
have simple, short labels. A few fields have changes in the values that can be selected from drop-down
menus. Some values, such as ‘99/99/9999’ for a missing date, are not accepted in eHARS.
Specimen information and STARHS results are captured on an eHARS laboratory document, separate
from the laboratory document for the diagnostic HIV test. The names of specimen variables in eHARS
differ from the names in the IVR database, but the main data elements are captured in eHARS. Some
data elements such as Specimen Approved for STARHS (LAPPRVE) and Were Results Received
(RESULTSRECEIVED) do not appear in eHARS.
Document-based Data Collection and Data Entry of TTH
Document-based data collection and entry differ somewhat from past practice. Using the IVR
database, surveillance areas generally selected one form type (pre-test or post-test) for entering TTH
information. If additional data were collected in another investigation from a second source, some
June 2012
5
surveillance areas recorded those data on the other form, but many just updated the original document.
As a result, HIS data forms contained mixed and sometimes conflicting data.
Surveillance data are often collected by non-incidence surveillance staff such as disease intervention
specialists (DIS) during client interviews, field staff conducting medical chart reviews, or medical
providers filling out case report forms. These collectors usually fill out one form but may collect
information from more than one source. For example, an interviewer at a sexually-transmitted disease
(STD) clinic may use information from a database for Date of Last Negative HIV Test to supplement
patient-reported information.
To be consistent with the principles of document-based data entry, new eHARS documents should be
used to enter data from multiple forms, from multiple sources, or to update information. At the top of
the TTH document screen, a new field, Main Source of TTH Information has replaced the IVR
database variable, Was the Testing History Questionnaire Implemented?, which indicated whether a
patient interview had been conducted. Main Source of TTH Information is important for documentbased HIS because it allows CDC and the project area to track the relative proportions of TTH
documents collected through various reporting methods (e.g., patient interview, medical record review,
passive case reporting) for the purpose of program improvement.
Another change is that eHARS also has additional tabs associated with every document. For guidance
on recommended fields that need to be completed on other eHARS tabs (Form Info, Identification, and
Demographics tabs), consult Appendix C, Data Entry Requirements for HIV Incidence Surveillance
Data Elements in eHARS.
DATA COLLECTION
The general principle in data collection is to collect the most accurate information available to
characterize a person’s HIV testing and treatment history. When investigating a case or completing a
case report form, data collection staff and medical providers may have access to multiple sources of
information, including the patient, medical records, laboratory reports, and other databases. It is
important to examine all the information on hand and ensure that the data recorded on the TTH form
most accurately reflect the patient’s actual HIV testing history and antiretroviral (ARV) use. The
purpose of the TTH is not to study the patient’s knowledge of his or her HIV status, but rather, (a) to
determine testing frequency in order to estimate the probability of being tested during the BED recency
period, and (b) to determine whether ARV use might have affected BED results.
Incidence surveillance staff should record data collected during separate investigations on separate
TTH data collection forms. A case report form sent in by a provider and a data collection form filled
out by field staff during a medical chart review each have a different Main Source of TTH Information
and should be recorded on different forms, even if both collectors used an ACRF.
Sometimes there will be mixed sources during one investigation and the data collector is filling out one
form. A data abstractor conducting a medical chart review should represent the best information
obtained from all sources reviewed, whether from patient self-report found in physician notes or
laboratory results noted in the chart. Similarly, for a provider report, the information will reflect all the
information available, including the provider’s knowledge of a previous positive or negative test that
was not reported by the patient. However, when an interviewer collects most of the TTH during the
patient interview and obtains the rest of the information from a medical chart or database, two forms
should be completed because these represent two different data collection methods. For example,
during an interview, a patient reports never having had a negative HIV test, but the interviewer also
finds in a database the date of a previous negative test for which the patient never returned for the
results. The interviewer should enter the information from the database on a separate TTH form. If
June 2012
6
there are two separate investigations (e.g., an interview and a complete chart review, or two medical
record reviews at different locations or time), two TTH forms should be completed. A TTH form
completed during investigation should summarize all the information obtained during that particular
investigation. If an additional TTH form is received by the health department with data identical to a
previously entered TTH, it is not necessary to create an additional TTH document. However,
surveillance staff must check to see that all elements are identical to the original document.
INDETERMINATE HIV TEST RESULTS
Data collectors should ignore indeterminate test results in recording responses to previous testing
history questions, including previous positive tests, previous negative tests, and number of negative
tests. An indeterminate test is neither positive nor negative for HIV. The guidance for indeterminate
tests differs from the guidance for unknown test results, (e.g., where patient never received test results)
because it may be possible to determine from another source that the unknown test result was either
positive or negative.
Data Entry
Data entry staff should enter data in eHARS as reported on the data collection form. Incidence data
may be entered manually or imported into eHARS. Incidence surveillance staff should always enter
data from different data collection instruments received on separate TTH documents; for example, a
provider case report form and a form filled out by field staff from a medical chart review. Each has a
different Main Source of TTH Information. If medical record reviews were conducted at two different
locations, the data should have been recorded on 2 separate TTH document forms and should
subsequently be entered on separate TTH documents in eHARS. Each eHARS document also may
have a different eHARS Document Source – reflecting the type of setting where the data were collected
– entered on the eHARS Form Info tab. For more information on recommended fields to be completed
on eHARS documents, refer to Appendix C.
When a data conflict is found on a single data collection form, data entry staff should not make
changes to the entries. Data entry staff should follow program procedures to resolve the data conflicts
which may include consultation with the Incidence Surveillance Coordinator or follow-up with the
individual who collected the data.
Data Quality
Sometimes, due to the nature of the data collection process (e.g., different sources) data conflicts may
arise. Each surveillance area should have Standard Operating Procedures (SOP) for resolving data
conflicts on data collection forms. The Incidence Surveillance Coordinator should investigate the
conflict and verify the information collected to determine if any further action is needed. Further
actions may include entering another document with the correct information or overriding a field in
eHARS Person View when incidence data are included in the person view document. Moreover, CDC
recommends that surveillance areas conduct regular data checks and data cleaning activities.
General Guidance
The following are descriptions of the common response options:
“Yes” indicates that there was sufficient evidence that the event occurred. Evidence can be
from patient self-report, health care provider note, or laboratory documentation.
“No” indicates that there was sufficient evidence that the event did not occur. Evidence can be
from patient self-report or health care provider documentation that definitively states that the
June 2012
7
given event never occurred (e.g., no previous positive test, never had a negative test, no ARV
use).
“Don’t know” indicates that the patient reported “don’t know,” the health care provider
documented “unknown,” or there was insufficient evidence for or against (supporting or
denying) the occurrence of the event.
Blank indicates that the usual data sources were not investigated and/or the health care
provider/staff were not asked.
“Refused” means patient refused, health care provider recorded “refused,” or facility refused to
permit the medical record review.
Dates
Dates are very important data elements for HIS. The date of first positive HIV test, date of last
negative HIV test, and dates of ARV use help to classify persons as new or repeat testers and as having
recent or long-term infections. In the hierarchy of incidence estimation algorithms, dates supersede
‘yes’ and ‘no’ answers.
For all TTH dates, eHARS has space for month, day, and year. For most dates, the IVR database and
data collection forms only contained month and year. In many cases, the day or even the month will be
unknown. It is acceptable to leave day or month blank (entering ‘..’ for missing month or day in
eHARS, such as ../../2001), but it is important to try to obtain at least the year, which can be used in
calculations. An approximate date reported by a patient is better than no date.
Format of this Document
This document provides the following for each variable used for HIS:
Name(s) of variable (eHARS TTH, IVR Pre-test and Post-test)
IVR, eHARS, and NHM&E/PEMS labels
Available values
Description
Purpose of the variable
History (if applicable)
Possible sources for the information
Specific information on data collection
Specific instructions for data entry
Examples by source of information
June 2012
8
eHARS and Pre-test Variable Name
Post-test Variable Name
1
2
Variable name:
UCTS (KCTS )
Wording in IVR Database:3
1a. Was the testing history questionnaire implemented?
(interview/ self administration vs. CTS)
Values: ‘Yes’, ‘No’ (legacy), ‘Patient Interview’,
‘Medical
Record
Variable
Description
Review’, ‘From PEMS’, ‘Other’
(label)
Wording on eHARS TTH:
Wording on ACRF:
1. Main source of testing and treatment history information
Values: ‘Patient Interview’, ‘Medical Record Review’, ‘Provider
Report’, ‘NHM&E/PEMS’, ‘Other’
Variable
Display
Main source of testing and treatment history
information
Values
(select one)
Values: ‘Patient Interview’, ‘Medical Record Review’, ‘Provider
Report’, ‘NHM&E/PEMS’, ‘Other’
Wording on NHM&E/PEMS
form
Not on NHM&E/PEMS form—always select ‘PEMS’
1
eHARS variable name, which is the Pre-test variable name
2
Post-test variable name (no longer used in eHARS)
3
From Pre-test version in IVR database
Examples of data collection/entry concerns are presented by source of TTH information because there
are often different instructions with each type of source. This document provides guidance to HIV
Incidence Surveillance Coordinators who will train staff or providers in the collection and data entry of
TTH using the revised adult case report form (ACRF) and other local data collection forms.
Certain formatting conventions are used. Refer to the example below.
Descriptions of variables appear in italics (e.g., Main Source of TTH Information).
Specific values for variables appear in single quote marks (e.g., ‘patient interview’) and refer to
the display values, not necessarily the values stored in databases.
Variable names appear in ALL CAPS (e.g., UCTS).
eHARS variable names for fields on the TTH document appear in ALL CAPS first and
correspond to former Pre-test variables, with retired (mostly former Post-test) variables in
parentheses, such as UCTS (KCTS).
For STARHS specimen information and test results, eHARS variable names appear in ALL
CAPS with IVR Access database variable names in parentheses, such as LOINC_CD
(ASSAY).
June 2012
9
1. Critical Data Elements for Testing & Treatment History (TTH)
1.1 TTH Data Element #1: Main Source of TTH Information
Variable name:
UCTS1 (KCTS2)
Wording in IVR Database:3
1a. Was the testing history questionnaire implemented?
(interview/ self administration vs. CTS)
Values: ‘Yes’, ‘No’ (legacy), ‘Patient Interview’, ‘Medical Record
Review’, ‘From PEMS’, ‘Other’
Wording on eHARS TTH:
1. Main source of testing and treatment history information
Values: ‘1-Provider Report’, ‘2-Patient Interview’, ‘3-Medical
Record Review’, ‘4- NHM&E/PEMS’, ‘5-Other’
Wording on ACRF:
Main source of testing and treatment history information
(select one)
Values: ‘Patient Interview’, ‘Medical Record Review’, ‘Provider
Report’, ‘NHM&E/PEMS’, ‘Other’
Wording on NHM&E/PEMS
form
Not on NHM&E/PEMS form—always select ‘NHM&E/PEMS’
1
eHARS variable name, which is the Pre-test variable name
2
Former Post-test variable name in parenthesis
3
From Pre-test version in IVR database
Description
The Main Source of TTH Information variable reports the method for obtaining most of the
information on the TTH document. This variable is different from the eHARS Document Source
variable which describes the type of provider or facility where the case information was gathered (e.g.,
hospital, laboratory, physician’s office, HIV clinic, or other database).
Purpose
The purpose of this variable is to characterize the source of the majority of data elements for a given
TTH document. CDC and surveillance areas will use this variable to evaluate the amount of
information that comes from various data collection sources which will inform incidence estimation
and program improvement. Also, it will be useful for data quality purposes to know if the source was
patient interview or another source when investigating conflicting information within a document or
across documents.
History
In the original data collection forms, this question was worded “Was the testing history questionnaire
implemented?” (yes/no). This caused confusion, so more categories (‘Patient Interview’, ‘Medical
Record Review’, ‘From PEMS’, ‘Other’) were added in the revised IVR Access database deployed in
July 2006. This variable was not included in the Standard Data Elements after 2007; however, it
became important for incidence estimation, program improvement, and data quality starting in 2010.
Surveillance areas currently should be using this variable. In the past, the IVR database had no specific
category to designate Provider Report, so data entry staff selected ‘other’. After eHARS conversion,
the category ‘provider report’ became available.
June 2012
10
Descriptions of Sources
Patient Interview: Patient was directly asked most of these questions by a disease intervention
specialist (DIS), health department (HD) staff, or a health care provider who asked the series of TTH
questions from a structured TTH form and who has received some training in the proper collection of
TTH information. If a chart review occurs in addition to the interview, two forms should be filled out,
one with ‘interview’ and one with ‘medical record review’ as the Main Source of TTH Information.
Medical Record Review: HD staff obtains the information through review of medical charts or
electronic medical records or databases. A few elements may come from the provider or other source
obtained as part of a medical record abstraction investigation. When HD staff visits a health care
provider’s office and extracts data from medical records, the source is ‘medical record review’, even if
the data collection form is the surveillance area’s provider case report form.
Provider Report: Information was obtained from a case report form submitted by a health care provider
or from a phone call with a provider. The health care provider does not indicate having directly asked
patients for most of this information from a TTH form. Information was found by the provider from
chart notes, laboratory reports, or recollection from discussion with the patient or another health care
provider. If the provider administered a series of questions from an ACRF or other form, then ‘patient
interview’ could be selected, depending on local policies (see note under Data Collection below).
NHM&E/PEMS: The Program Evaluation and Monitoring System (PEMS) is also known as National
HIV Monitoring and Evaluation (NHM&E) and the proposed test form is called the NHM&E HIV
Test Form. Information was obtained from NHM&E or earlier PEMS forms (e.g., Parts 1 and 3),
locally developed CTR forms, or imported from a NHM&E/PEMS database. It is understood that most
of NHM&E/PEMS data come from patient interview but this source selection clarifies the source as
NHM&E/PEMS, which is useful for evaluation. For this guidance, we use the two names, PEMS and
NHM&E, interchangeably.
Other: Information was obtained from another source. If the only source is a database, the ADAP
database for example, select ‘other’.
If the source of data is unknown, then leave Main Source of TTH Information blank.
Data Collection
The data collectors should select the category that best describes the source of the TTH data elements
on this form. If there are data elements collected from two or more of the sources described above
(e.g., an interview and a chart review), then the data collector should use two separate forms, in
accordance with document-based data collection principles.
Note: Each local surveillance area should consider all of the regular sources of information that
contribute to the TTH and assign a value for Main Source of TTH Information for each type of source.
Even if a provider checks ‘patient interview’ on the ACRF, the value should be entered as ‘provider
report’ unless there is evidence that the provider conducted a structured interview using specific TTH
questions. This is because the questions may not be asked or interpreted exactly as this Guidance
document describes. Therefore, Incidence Surveillance Coordinators in consultation with the CDC
epidemiologist may need to make project area-specific decisions on how to classify information from
certain local forms or particular reporting sources. For example, if there are providers at an HIV clinic
that have received training on the collection of TTH, then the Incidence Surveillance Coordinator may
decide that forms received from that clinic should be designated ‘patient interview’.
June 2012
11
Data Entry
If the data source does not appear on the data collection form, the data entry person, in consultation
with the Incidence Surveillance Coordinator, should enter the best answer that characterizes the data
source. For example, if the data come from an ACRF or a local provider case report form sent to the
health department by a provider, enter ‘provider report’. If the information comes from a data
collection form that field staff uses to conduct chart abstraction, even if it is the ACRF, enter ‘medical
record review’. If the form is a known data collection instrument from a STD clinic interview, select
‘patient interview’. If the data are on a NHM&E/PEMS form or from a NHM&E/PEMS database,
enter ‘NHM&E/PEMS’. The data entry person should consult with the Incidence Surveillance
Coordinator if uncertain about the data source.
1.2 TTH Data Element #2: Date Patient Reported Information
1
Variable name and format1:
UQINTD (KQINTD) mm/dd/yyyy
Wording in IVR Database:
1. Today’s Date (Use THQ Interviewing Date)
Wording on eHARS TTH:
2. Date patient reported information
Wording on ACRF:
Date patient reported information
Wording on NHM&E/PEMS form:
Session Date (Part 1) or Date Information Collected (Part 3)
Wording on proposed NHM&E
form:
Date Client Reported Information (Part 3)
Format displayed on eHARS data entry screen
Description
The Date Patient Reported Information variable represents different dates, depending on the
circumstances in which the information was primarily obtained. This variable represents the date of the
patient interview if information was obtained face-to-face, or the date of the provider note for the last
patient encounter at which TTH information was ascertained from a medical record. If there was no
patient contact, then the date of receipt of TTH information from a laboratory report or database should
be used. If none of these dates are available, date of medical record review or date provider completed
the report form should be used.
Purpose
This date may be used in the future for some calculations in analyses and to select the best incidence
values for eHARS Person View (PV) when there are multiple TTH documents. In the eHARS PV
hierarchy, some TTH values may be selected from the earliest TTH document based on this date or, if
the date is blank, the date the document was entered. However, individual variables can be overridden
in PV by the Incidence Surveillance Coordinator to best represent a case’s testing and treatment
history.
Sources
Sources for this variable could include interview date on a NHM&E, CTS or other patient interview
form, last date of a patient visit noted in a medical chart from which the TTH information was
ascertained, date of interview in the NHM&E/PEMS database, date reported on a provider report form,
June 2012
12
and, if Date Patient Reported Information was blank, date health care provider filled out a case report
form or date received at the health department.
Data Collection
In general, the data collector should select the last date of interview or patient visit that contributed to
the testing and treatment information. It represents the point in time when the TTH information was
obtained from a patient. It does not necessarily mean the last patient visit. If data only were obtained
from a laboratory report or from a database, the date should be the date when the data were received by
the provider or health department. If none of these dates are available, use the date when data were
collected (i.e., the date of medical record review or date provider completed the case report form). If
those dates are missing, use the date the information was received by the health department.
Data Entry
If there are two different data collection forms or sources, the data entry person should enter them as
separate TTH documents, each with the appropriate Date Patient Reported Information. If the date is
missing, enter the date the form was received at the health department.
Examples by Source of Information:
Patient Interview: Date of interview should be entered. An interview can be conducted by a health care
provider, Disease Intervention Specialist (DIS), or Health Department (HD) staff using the standard
TTH questions.
More than one interview was conducted—interviewer(s) should complete separate TTH
documents, one for each interview date.
Interview was conducted on two different days, due to an interruption. Interviewer should enter
all of the information on one form and select the last date when interview was completed.
Medical Record Review: Use the last date of the patient encounter or provider’s note that contributed
to the TTH information for this form. Do not use the actual medical record review date unless there is
no other date to use.
Medical record review was conducted at two different locations—abstractor should enter two
separate TTH documents with appropriate dates.
Medical record review occurred at two different points in time—abstractor(s) should enter
separate TTH documents with the appropriate dates.
Chart abstractor started a chart review one day and completed the review on the same form a
week later—abstractor should record the date of the last patient encounter with TTH
information.
Chart abstractor also found a pharmacy record for ARVs dated after the last patient
encounter—abstractor should record the date of the last patient encounter with TTH
information for Date Patient Reported Information because it reflects the time when the patient
answered questions. He should use the pharmacy information for recording dates of ARV use.
Chart abstractor found no TTH information in the provider notes for patient visits, but found a
negative HIV test date in a clinic testing database—she should enter the date the test result was
received. If there is no received date, then enter date of chart abstraction.
Provider Report: Providers should be instructed to record on the provider report form the date when
most of this information was obtained from the patient or another data source, or, lacking that, the date
when the provider completed the report.
June 2012
13
Provider filled out a case report form—provider should record the date of the last patient
encounter that provided TTH information.
Provider filled out a case report form for a person with a positive HIV test result but there was
no patient encounter—provider should enter date that the form was filled out.
Information was only gathered from laboratory reports—provider should enter the date on
which the provider received the laboratory results or, if unavailable, the date he or she filled out
the form.
The last encounter with patient where TTH questions were answered was in April 2010, but the
pharmacy record shows a prescription for ARVs on October 2010—record the date of the
patient encounter, April 2010.
The only information came from pharmacy records—report the date the prescription was filled
or, if missing, the date the provider form was completed.
HIS staff member called the provider because there was no TTH information on the ACRF
submitted to the health department. If the provider does not have any TTH information there
are 2 choices: 1) do not complete a TTH form as no new information was provided, or
2) complete a TTH form by entering ‘don’t know’ for several variables based on the phone call.
In this case, enter the date of the phone call for Date Patient Reported Information.
No reported Date Patient Reported Information on ACRF or provider report form—data entry
person should enter the date the provider completed the form, if available, or when the case
report was received by the HD.
NHM&E/PEMS: HIS data entry person should enter the date of interview obtained from NHM&E,
PEMS or CTS form or database.
Do not leave this date blank. If Date Patient Reported Information is unknown, enter the date when the
document was received at the health department. This date may be used for prioritizing the selection of
incidence data in eHARS PV when there are multiple TTH documents.
1.3 TTH Data Element #3: Ever had a Previous Positive HIV Test?
Variable name and format:
UPASTP1 Yes/No/Refused/Don’t know
Wording in IVR Database (only on
Pre-test form):
4. Have you ever had a positive HIV test result?
Wording on eHARS TTH:
3. Ever had a previous positive HIV test?
Values: ‘N – No’, ‘Y – Yes’, ‘R – Refused’, ‘D – Do not know’
1
Wording on ACRF:
Ever had previous positive HIV test?
Wording on PEMS form:
Part 1, Previous HIV Test? (#26), Self-reported
Result (#27), and Date of Last Test (#28/29)
Wording on proposed NHM&E
form:
Has client ever had a previous positive HIV test? (Part 3)
Values: ‘Yes’, ‘No’, ‘Don’t know’, ‘Declined’
eHARS and Pre-test variable name
June 2012
14
Description
The Ever Had a Previous Positive HIV Test variable provides information about the patient ever
having a positive HIV test before the current test. This variable originally only appeared on the Pre-test
TTH form but was added to the CDC ACRF in 2010 to clarify information on the first positive HIV
test. Surveillance areas should make sure this variable is added to local data collection forms.
Purpose
The purpose of this variable is to ascertain whether a positive HIV test occurred earlier than the
eHARS calculated HIV Disease Diagnosis Date (HIV_AIDS_DX_DT), but was not reported to the
HIV surveillance system; for example, a patient could have been diagnosed in another state/country or
tested anonymously. Having an earlier test, accompanied by an approximate date, might indicate a
long-term HIV infection. This information will also be used for TTH completeness reports starting in
2011 and to evaluate the validity of the variable, Date of First Positive HIV Test.
In the absence of an earlier, self-reported Date of First Positive HIV Test (#4), CDC and local
analytical programs assume that the HIV disease diagnosis date is the collection date of specimen used
for the first positive HIV test. The variable allows the entry of a ‘no’ to confirm that the person
reported never having a previous positive test. We can use this to learn for what proportion of cases it
is known that this is the first positive test.
Sources
This information may come from patient self-report, evidence of an anonymous test, record of test
performed in another country, a doctor’s note, discussion with another provider, or other source that is
probably not a laboratory report. A confirmed positive from an HIV home test kit should be considered
the same as a self-reported anonymous test. It is not reported to the health department.
Data Collection
Data collection may differ by source of information. In a patient interview, a person will likely state
that he or she has or has not had a previous positive test. Persons with a previous positive test often
were tested anonymously and are now converting to confidential testing. Also, the person may have
been tested in another state or was tested previously and is now seeking care. If so, record ‘yes’ and
make an effort to obtain the date of first positive for Date of First Positive HIV Test (variable #4).
Ignore indeterminate and false positive tests. If the only previous HIV test was indeterminate, do not
consider that test as a positive test; in that case, enter ‘no’. For medical record review or provider
report, it is more difficult to ascertain whether there was a previous positive HIV test unless the patient
was specifically asked and the information was recorded in the medical notes. If there is no evidence
either way, enter ‘Don’t know’ on the data collection form. This is useful information for interpreting a
date for the Date of First Positive HIV Test (variable #4 below).
Data Entry
The data entry person should enter the response recorded on the data collection form for this question.
Otherwise, leave the field blank. If the surveillance area does not have this field on current forms, it
should be added to TTH collection forms. If a conflict is identified, the data entry person should follow
standard operating procedures (SOP) for resolving data conflicts and consult with the Incidence
Surveillance Coordinator.
June 2012
15
Examples by Source of Information:
Patient Interview:
Patient self-reported a previous test—interviewer should record ‘yes’.
Patient reported a positive HIV test from a home test kit that was sent into a laboratory—
interviewer should record ‘yes’.
Only previous test was ‘indeterminate’—interviewer should record ‘no’.
Patient reported a positive oral screening test followed by a negative Western Blot—
interviewer should record ‘no’. This test is considered a negative HIV test.
Patient had a false positive HIV test when she was pregnant 3 years ago and was confirmed to
be HIV negative 6 months later—record ‘no’.
Patient did not know—interviewer should record ‘don’t know’.
Patient never got his results—interviewer should record ‘don’t know’.
Patient reported ‘no’ or ‘don’t know’ but the interviewer later found evidence of an earlier
positive test– interviewer should leave the patient’s answer as ‘no’. The date of the earlier
positive test should be entered on a separate TTH document. If it is a documented positive
confirmatory test, it also should be entered on an ACRF or Laboratory document.
Medical Record Review:
Chart abstractor found evidence/report of a previous positive test in the medical record; for
example, a doctor’s note that said ‘patient tested in CA in 2006’—abstractor should record
‘yes’.
Chart abstractor found information about a previous test in a clinic database, regardless of
whether patient knew about it—abstractor should record ‘yes’.
Medical record indicated that the patient never had a previous positive test; for example, a note
saying ‘this is the patient’s first positive HIV test’—abstractor should record ‘no’.
The patient reported that the only previous HIV test patient had was ‘indeterminate’ —
abstractor should record ‘no’.
Patient reported she never got her results—abstractor should record ‘don’t know’.
There was no definitive evidence either way; for example, the provider did not ask the question
or made no note in the chart—abstractor should record ‘don’t know’. This indicates that effort
was made to find the answer but it was unavailable. Do not assume that no evidence is the same
as no previous positive test.
Provider Report: Health care providers should be encouraged to ask patients about previous testing and
record notes in the medical record.
Provider knows from patient report that this was the first positive HIV test that the patient ever
received—provider should record ‘no’.
Provider knows there was an earlier positive HIV test before the one being reported, but has no
date—provider should record ‘yes’.
Provider does not know about any previous positive HIV test—provider should record ‘don’t
know’.
NHM&E/PEMS: The HIV Surveillance data entry person or import program should calculate from
items 26, 27, 28, and 29 from Part 1 to see if there was a previous positive HIV test, or record the
June 2012
16
answer from the TTH question on Part 3 of the new NHM&E form. Enter the information obtained
from a NHM&E, PEMS or CTS form or database. If there is no information, leave blank.
If there was no patient interview, no medical chart review, no provider report, and no other follow-up,
the answer should be left blank.
1.4 TTH Data Element #4: Date of First Positive HIV Test
Variable name and format:
UFPOSD (KFPOSD) mm/dd/yyyy
Wording in IVR Database:
4b. What was the month/year of the very first time you ever
tested positive for HIV? List when you got the test, not when you
got the results.
Wording on eHARS TTH:
4. Date of first positive HIV test
Wording on ACRF:
Date of first positive HIV test
Wording on NHM&E/PEMS
form:
Date of first positive HIV test (Part 3)
Description
The Date of First Positive HIV Test variable is the date of the earliest known positive HIV test for the
patient. It represents the date that the specimen was collected for the very first positive HIV test. This
date could represent an anonymous test that will never be reported to the national HIV surveillance
system. Most of the time, this date is self-reported.
Note: Any documented, positive HIV laboratory test result that is earlier than the current calculated
eHARS HIV Disease Diagnosis Date (HIV_AIDS_DX_DT) should be entered in eHARS on an ACRF
or on a Laboratory document. That will result in the recalculation of HIV_AIDS_DX_DT in eHARS
Person View.
Purpose
This date is important for identifying cases that tested positive for HIV earlier than the eHARS
calculated HIV Disease Diagnosis Date (HIV_AIDS_DX_DT) and therefore may not represent new
diagnoses. For most data analyses, the eHARS HIV Disease Diagnosis Date is used. However, during
incidence estimation, if the TTH self-reported Date of First Positive HIV Test is earlier than the
eHARS date, the TTH date is used for reclassifying BED results and for calculating the inter-test
interval (T). If the self-reported Date of First Positive HIV Test is more than 6 months earlier than the
eHARS diagnosis date, BED results are set to ‘long-term’; if the Date of First Positive HIV Test is 1–6
months before the eHARS date, BED is set to missing and results are imputed. If the Date of First
Positive HIV Test is earlier than eHARS diagnosis date, it will be used in the above calculation, even if
Ever had a Previous Positive HIV Test is ‘no’. If Date of First Positive HIV Test is blank, it is
assumed, for analysis purposes, to be the same as the HIV Disease Diagnosis Date.
Sources
Sources for this variable could be patient self-report, report of an anonymous test, report of a test
performed in another jurisdiction, record of a test in a clinic database, doctor’s note, discussion with
another provider, or other source.
June 2012
17
Data Collection
The self-reported Date of First Positive HIV Test could be the same as the documented date of the
confirmatory HIV test report or eHARS calculated HIV disease diagnosis date. During interviews,
chart abstraction, or completion of the ACRF, some data collectors will enter the date of the current
HIV positive test for the purposes of calculating subsequent answers. When it is known that the current
test date is the date of first positive HIV test, it is important to answer ‘no’ for the Ever had a Previous
Positive HIV Test data element (#3 above). That will clarify why the date is the same as the eHARS
HIV diagnosis date. Do not record any dates for ‘indeterminate’ HIV tests, false positive tests or for
tests with unknown results.
Data Entry
Data entry person should enter the date recorded on the form. If a conflict is identified, follow standard
operating procedures for resolving data conflicts and consult with the Incidence Surveillance
Coordinator.
Examples by Source of Information:
Patient Interview:
Patient reported having a previous positive HIV test (e.g., an anonymous test or a test in
another state)—interviewer should record the date of the first one. The interviewer can use
prompts to assist the patient to remember the date, at least the year of the test.
Patient stated she has never had a previous positive test—record current test date and select
‘no’ for question #3.
Patient said he never had a previous positive, but the interviewer has information that there was
a previous positive test for which the patient never returned to get the results—interviewer
should enter the date of the previous positive test. Interviewer may share the test results in
accordance with local protocols and HIPAA policies.
After the interview, the interviewer found evidence of a previous positive test—enter the date
on a separate TTH document.
Patient did not know if she ever had a positive HIV test. Enter ‘don’t know’ for Ever had a
Previous Positive HIV Test (#3) and leave Date of First Positive HIV Test blank.
Medical Record Review:
Medical record abstractor is not sure what the eHARS date of HIV infection is—abstractor
should enter the date of the earliest known HIV positive test noted in the record. This would be
the date the specimen was collected for the initial HIV test.
Medical record, including self-report, indicated the patient was diagnosed or tested at earlier
time than the current test being investigated—abstractor should record the date noted in the
record and enter ‘yes’ for Ever had a Previous Positive HIV Test (#3).
Medical record indicates the patient was never tested before—abstractor should record ‘no’ for
Ever had a Previous Positive HIV Test (#3) and record the date of the first known positive test.
This should be the date of the current test.
There is no information in the chart about a previous test—abstractor should record ‘don’t
know’ for Ever had a Previous Positive HIV Test (#3) and leave Date of First Positive HIV
Test blank.
June 2012
18
Provider Report: Often health care providers enter the current test date when completing a provider
case report form.
Patient reported a previous test date to the provider or the provider has access to information
about a previous positive—provider should record the date and enter ‘yes’ for Ever had a
Previous Positive HIV Test (#3).
Patient reported to the provider that this was his first HIV test ever—provider enters ‘No’ for
Ever had a Previous Positive HIV Test (#3) and records the date of current test.
Provider doesn’t know if there was a previous positive—provider should record ‘don’t know’
for Ever had a Previous Positive HIV Test (#3) and leave Date of First Positive HIV Test blank.
NHM&E/PEMS: The HIV Surveillance data entry person should enter the information obtained from
NHM&E/PEMS (Part 3) or CTS form or database. If there is no information, leave blank.
If the value entered on any form or into any database is ‘999999’ or there was no patient interview, no
medical chart review, no provider report, and no other follow-up, leave the Date of First Positive HIV
Test blank.
1.5 TTH Data Element #5: Ever had a Negative HIV Test?
Variable name and format:
UNGTST (KNGTST) Yes/No/Refused/Don’t know
Wording in IVR Database:
4f. Have you ever had an HIV test that was negative?
Wording on eHARS TTH
5. Ever had a negative HIV test?
Values: ‘N – No’, ‘Y – Yes’, ‘R – Refused’, ‘D – Do not know’
Wording on ACRF
Ever tested HIV negative?
Wording on NHM&E/PEMS
form:
Has client ever had a negative HIV test? (Part 3)
Values: ‘Yes’, ‘No’, ‘Don’t know’, ‘Declined’
Description
This variable, Ever Had a Negative HIV Test, captures whether or not the person ever had a negative
HIV test result at any time in the past. In the initial IVR database, this variable appeared only on the
Post-test form but was added to the Pre-test form with the revised IVR Access database, version 3.1
released in July 2006.
Purpose
This variable is one of three data elements (variables #5, 6, and 7) used to classify cases as new testers
or repeat testers. This distinction is important for incidence estimation because the probability of being
classified as recent by the BED assay is calculated separately for new testers and repeat testers.
Persons with a ‘yes’ answer—indicating a previous negative HIV test—are classified as repeat testers,
and those with a ‘no’ answer are classified as new testers. For the purposes of incidence estimation,
when Ever Had a Negative HIV Test is unknown or missing, Date of Last Negative Test is blank, and
Number of Negative HIV Tests Within 24 Months before First Positive HIV Test is unknown, missing,
or ‘0’ (zero), persons will be classified as a new or repeat tester using multiple imputation. Multiple
imputation is a statistical process that will assign cases to either the ‘new’ or ‘repeat’ tester groups.
June 2012
19
Sources
Sources for this variable could be patient self-report, laboratory report of a negative HIV test, record of
a test in a clinic database, doctor’s note, discussion with another provider, or other source.
Data Collection
Since this data element is used to classify cases as new testers or repeat testers for incidence
estimation, it is important to have accurate information because misclassification can strongly impact
the accuracy of the incidence estimate. The data collector should not make the assumption that there
was never a negative test when he or she could not find any information about a negative test. Because
people are tested for HIV in many different venues or with different providers, the absence of
information about previous testing does not mean that previous tests did not occur.
If an interviewer has knowledge of a previous negative HIV test, he or she should prompt the patient to
recall it. If the interviewer finds evidence of a previous negative test after the interview, even though
the patient answered ‘no’, the interviewer should record the Date of the Last Negative HIV Test (#6)
but he or she does not need to change the patient’s response. The algorithm for classifying new testers
and repeat testers gives priority to the Date of Last Negative HIV Test and will assign the person to the
repeat tester category. It is best practice for the data collector to enter the date of the negative test on a
separate TTH form when those data were found separately from other information.
If the patient reports that he or she doesn’t know, if the health care provider does not know whether the
patient had any negative HIV tests, or there was insufficient documented evidence in the medical
record supporting or denying the occurrence of a negative test, the data collector should select ‘don’t
know’. ‘Don’t know’ and ‘can’t find evidence’ are treated similarly. The field should be left blank if
the usual data sources have not been investigated and/or the health care provider did not ask the
patient.
If the patient’s only previous HIV test was positive, select ‘no’. Ignore indeterminate tests. If the only
previous test was indeterminate, select ‘no’, because there is no evidence of a previous negative.
Similarly, an undetectable viral load is not evidence of a negative HIV test.
Data Entry
The data entry person should enter the data as recorded on the form. If the data entry person identifies
a conflict, he or she should consult with the Incidence Surveillance Coordinator and follow the process
outlined in the SOP. If there is no information on the form, leave blank.
Examples by Source of Information:
Patient Interview:
Patient was asked this question—interviewer should record the answer as given.
Patient reported having a negative test result from a home HIV test kit that was sent to a
laboratory for testing—interviewer should record ‘yes’.
Patient was tested but did not know the test results—interviewer should record ‘don’t know’
Patient reported ‘no’ or ‘don’t know’ to Ever Had a Negative HIV Test but interviewer found
evidence of a previous negative HIV test before or during interview—interviewer may prompt
the patient by sharing the previous negative test results in accordance with local protocols and
HIPAA policies. Regardless of patient response, record the Date of the Last Negative Test (#6,
below).
June 2012
20
Patient answered “no,” but interviewer later finds evidence of a previous negative HIV test,
interviewer should not change the response but record the Date of the Last Negative Test (#6,
below). The date should be recorded on a separate form, if possible.
Patient was unsure about whether he had a negative test—interviewer should record ‘don’t
know’.
Medical Record Review: Chart abstractors should examine the medical record for a provider note or
laboratory evidence of a negative test.
There was a report of a negative HIV test, with or without a date—abstractor should enter
‘yes’.
There was a note indicating that the patient had never tested before—abstractor should record
‘no’.
There was no note or other indication that the patient ever had a negative test or not—abstractor
should record ‘don’t know’. This indicates that effort was made to find the answer but it was
unavailable.
Provider Report: Providers should be encouraged to ask patients about previous HIV tests.
Patient reported having a previous negative HIV test (with or without a date)—provider should
report ‘yes’.
Provider has access to information about at least one previous negative HIV test—provider
should report ‘yes’.
Patient indicates she never tested before or she never had a negative HIV test—provider should
report ‘no’.
Provider finds no evidence of a previous negative test in her notes or patient reports ‘don’t
know’—provider should report ‘don’t know’.
NHM&E/PEMS: Data entry person should enter the information obtained from NHM&E/PEMS (part
3) or CTS form or database. If there is no information, leave blank.
If there was no patient interview, no medical chart review, no provider report, or no other follow-up,
the answer should be left blank.
1.6 TTH Data Element #6: Date of Last Negative HIV Test
Variable name and format:
ULSTND (KLSTND) mm/dd/yyyy
Wording in IVR Database:
4f. What was the month and year that you got your last negative
HIV test? List when you got the test, not when you got the
results.
Wording on eHARS TTH
6. Date of last negative HIV test
Wording on ACRF-TTH
Date of last negative HIV test
Wording on ACRF-LAB
Date of last documented negative HIV test
Wording on NHM&E/PEMS
Form
Date of last negative HIV test (Part 3)
Description
This variable, Date of Last Negative HIV Test, is the date of the last known negative HIV test and
represents a point in time when the person was known not to be infected with HIV.
June 2012
21
Purpose
This is one of the most important data elements for incidence estimation. It is used to classify cases as
repeat testers, like data element Ever Had a Negative HIV Test (#5), and is also used to calculate the
inter-test interval between last negative and first positive tests for repeat testers.
Sources
Sources for this variable may include patient interview, medical record review, physician note on a
chart, ACRF, provider case report, field staff data collection form, databases at testing sites, PEMS or
CTS forms, or other source. Field staff should receive training to note the date of last negative HIV test
when following up on new cases for any reason.
Data Collection
Because this is one of the most important data elements for incidence surveillance, extra effort should
be made in collecting this date. Presence of a date of last negative HIV test classifies a person as a
repeat tester, even if the patient’s answer to question #5 is ‘no’. During incidence estimation, the
probability of being classified as recent by the BED assay is calculated separately for new testers and
repeat testers. Any available information about testing history is useful in this calculation, including an
approximate date or only the year tested, because it will classify a person as a repeat tester.
It is important to train interviewers, providers, and chart abstractors to record the last known date of a
negative HIV test, even if the data collector does not know if negative results for the patient have
occurred for later HIV tests performed at other facilities. If there are two dates for negative HIV tests,
the most recent one should be entered. If the provider or chart abstractor is using the standard ACRF
and has evidence of a documented negative HIV test with a test type, the date should be entered in the
Laboratory Data section of the ACRF in the Date of Last Documented Negative HIV Test field.
Otherwise, he or she should enter the date in the HIV Testing and ARV Use History section of the
ACRF. Do not include dates of tests with unknown or indeterminate results.
Data Entry
If the negative HIV test date is obtained from a documented laboratory report that contains test type
(e.g., EIA), it should be entered in Date of Last Documented Negative HIV Test on the Lab Data tab of
the ACRF form in eHARS. Any other dates of a negative HIV test which may come from patient selfreport, provider report, chart notes, NHM&E/PEMS, or other databases should be entered on the TTH
document, even if it is not known if it was absolutely the most recent test that the patient had. If
another test date is found later, it should be entered on a separate eHARS document. If data entry
person is unsure which form should be used to enter data, he or she should consult with the Incidence
Surveillance Coordinator.
Note: Until the implementation of eHARS version 4.0, it is recommended to also enter any
documented negative laboratory test date found on the ACRF into the field for Date of Last
Negative HIV Test on a TTH document, as well as on the Lab Data tab of the ACRF in the field
for Date of Last Documented Negative HIV Test. Because this data field is not properly contributing
to the calculated variable for date of last negative test in eHARS, the date should also be entered on the
eHARS TTH document until the defect is fixed in eHARS version 4.0. After version 4.0, the Date of
Last Documented Negative HIV Test will be entered as any other negative HIV test. There will no
longer be a separate field on the eHARS screen for this data element. There will be a calculated
variable for the date of last negative HIV test before first positive test.
June 2012
22
Examples by Source of Information:
Patient Interview: Interviewer should record the date of the last known negative HIV test reported by
the patient, even if the date only contains the year tested.
Patient reported having a negative HIV test but cannot remember the date—interviewer should
prompt patient to estimate the month and year tested. If unsuccessful, leave the date blank. Also
record ‘yes’ for Ever Had a Negative HIV Test (#5).
Patient recalled the date of one negative HIV test but was unsure if she had another one after
that date—interviewer should record reported date.
Interviewer found evidence of a previous negative test from another source, even if the patient
forgot or was unaware of the test (e.g., didn’t return for results)—interviewer should record the
date. It is best practice to record the date on a separate TTH form. Retain patient’s answer for
Ever Had a Negative Test (#5).
Patient reported she donated blood 3 months ago and says she must have been HIV negative or
the blood bank would have called her—do not record this date. Do not assume that is evidence
of a negative HIV test.
Medical Record Review: Chart abstractors should examine the chart for a provider note, patient selfreport, or laboratory evidence of the last known negative HIV test.
Provider note says ‘patient reports negative HIV test in July 2007’—abstractor should record
month and year of the reported test, not the date of the provider note.
There is a report of a documented negative HIV test date—abstractor should record date in the
Laboratory section of the ACRF along with the test type (e.g., HIV-1 EIA).
A documented, negative HIV test is found in the medical record and recorded in the Laboratory
section of the ACRF. A later patient-reported date is also described in the physician note.
Record the later patient-reported date in the Testing and ARV Use History section of the ACRF
or other data collection form.
Provider Report: Health care providers should be encouraged to ask patients about previous negative
HIV tests and to record the most recent reported date, even if it only contains the year.
Provider documented a laboratory result with test type—provider should report in the
Laboratory section of the ACRF under Date of Last Documented Negative HIV Test.
Provider from a clinic that runs a community HIV testing program consults a database of past
tests and finds a date of last negative test for the patient but with no test type—provider should
report the date in the Testing and ARV Use History section of the ACRF or other provider
report form.
Patient reported there is no previous HIV test—provider should leave date blank and report
‘no’ for Ever Had a Negative HIV Test (#5) on the provider report form.
NHM&E/PEMS: Enter the information obtained from NHM&E/PEMS (Part 3) or CTS form or
database. If there is no information, leave blank.
If the value entered on any form or into any database is ‘999999’ or there was no patient interview, no
medical chart review, no provider report, and no other follow-up, the answer should be left blank.
June 2012
23
1.7 TTH Data Element #7: Number of Negative HIV Tests within 24 months before First
Positive HIV Test
Variable name and format
UNUMTSTS (KNUMTSTS, UPNUMTSTS)
IVR Values: 1-99, Refused (R), Don’t Know (D);
eHARS Values: 0-99, R, D
Wording on Pre-test form,
IVR Database (UNUMTSTS):
5b. For people who have NEVER had a positive test: In the past
two years, how many times did you get tested for HIV? Today’s
test should be included for you in the count.
Wording on Pre-test form,
IVR Database (UPNUMTSTS)
5a. For people who have had a positive test: in the two years
before your first positive HIV test (that is, the two years before
the date in question 4b) how many times did you get tested for
HIV? Your first positive test should be included for you in the
count.
Wording on Post-test form,
IVR Database (KNUMTSTS)
9. In the two years before your first positive HIV test (in
Question #3) how many times did you get tested for HIV? Your
first positive HIV test (in Question #3) should be included for
you in the count.
Wording on eHARS TTH
7. Number of negative HIV tests within 24 months before first
positive test (Do not include first positive HIV test)
Wording on ACRF
Number of negative HIV tests within 24 months before first
positive test
Wording on PEMS Form
Number of tests in the two years before the current (or first
positive) test. Include the current (or first positive) test. (Part 3)
Values = 1-99
Wording on proposed PEMS
form
Number of negative HIV tests within 24 months before the
current (or first positive) HIV test (Part 3)
Values = 0-99, Don’t know, Declined
Description
This variable, Number of Negative HIV Tests within 24 Months before First Positive HIV Test, is used
to quantify the number of negative HIV tests within the 24 months preceding the first positive HIV
test.
Purpose
The purpose of this variable is to indicate testing frequency in the 24 months before the first positive
HIV test, which is further used to calculate the inter-test interval for repeat testers in the absence of an
actual date of last negative test. If a person reported having a previous test (Ever had a Negative HIV
Test = ’yes’) but Number of Negative HIV Tests within 24 Months before First Positive HIV Test is ‘0’,
then the inter-test interval is assumed to be >24 months. If there is no Date of Last Negative HIV Test
and the value of the Number of Negative HIV Tests within 24 Months before First Positive HIV Test is
‘R’, ‘D’, or blank, multiple imputation is used to calculate the inter-test interval for repeat testers.
This variable is also used to classify cases as repeat testers when the Number of Negative HIV Tests
indicates one or more previous negative tests and there is no information for Ever had a Negative HIV
Test (#5) and Date of Most Recent Negative Test (#6).
June 2012
24
History
Historically the TTH forms and IVR database included the current positive HIV test in this count,
which caused confusion. The value ‘1’ was inconsistently reported and could not be used to classify
individuals as a new tester or repeat tester. Starting on the date of conversion of HIS data to eHARS at
a given surveillance area, only previous negative tests were counted in the newly-entered eHARS TTH
documents; the current test, or first positive test, is no longer included. Field staff, providers,
interviewers, and data entry staff need to be trained on the specific solution for the surveillance area,
depending on the wording of data collection instruments.
For data collected but not entered prior to eHARS conversion, Incidence Surveillance Coordinators
must correct the values for this data element before data are entered in eHARS or instruct data entry
staff on how to enter the data correctly.
Sources
Sources for this variable can be patient self-report, medical record review, doctor notes, and laboratory
reports.
Data Collection
Field staff who perform data collection and providers who complete case report forms need to be
informed that as of the date the surveillance area converted incidence data to eHARS, this variable will
no longer include the first positive HIV test.
The best source of this information is patient interview because a provider or medical record is
unlikely to have the total of all tests in the 24-month period preceding the first positive HIV test. As a
result this field is often blank or ‘don’t know’.
If the interviewer, abstractor or provider knows of at least one previous negative HIV test in the 24
months before first positive test, he or she should enter ‘1’ even if there could be more unknown tests
in this period. In the absence of a ‘yes’ for Ever Had a Negative HIV Test (#5) and a date for Date of
Last Negative HIV Test (#6), entering a ‘1’ for this variable classifies the person as a repeat tester. In
addition, it indicates that a person received at least one negative HIV test in the two years prior to the
HIV diagnosis date, which is used to estimate the inter-test interval for repeat testers.
If it is known that the patient did not have any HIV tests in the previous 24 months, then ‘0’ should be
recorded.
If the patient doesn’t remember whether they had a negative test, or the provider or abstractor has no
evidence about whether or not there was a previous test, enter ‘don’t know’. Do not record zero (‘0’).
Do not count an indeterminate test as a negative test. Ignore indeterminate tests.
Note: In analysis, when there is no known Date of First Positive HIV Test that is earlier than the HIV
Disease Diagnosis Date, it is assumed that the Date of First Positive HIV Test is the same as the
calculated eHARS HIV Disease Diagnosis Date (HIV_AIDS_DX_DT). If the patient/provider/chart
reviewer does not know whether the patient ever had a previous positive HIV test, but he has
information on the number of tests in the past two years, then this information should be captured. For
example, the patient remembers one negative HIV test in the past 24 months but doesn’t know if he
had a previous positive. Entry of ‘1’ allows us to categorize the patient as a repeat tester and estimate
frequency of testing.
June 2012
25
Data Entry
Because data for this data element have been entered inconsistently, the guidance has changed, as of
each surveillance site’s date of conversion to eHARS, to stop including the first positive HIV test in
the count. This change may require some data adjustment for this data element at the point of data
entry for documents entered in eHARS on or after conversion to eHARS. The data entry person should
consult with the Incidence Surveillance Coordinator for the appropriate data entry from the area’s data
collection forms.
For areas that used the IVR Pre-test form: At conversion, data for both of the Pre-test variables,
UNUMTSTS and UPNUMTSTS, were converted to eHARS. Data for UNUMTSTS (number of tests
for persons with no previous positive HIV test) appears at the top of the form and data for
UPNUMTSTS (number of tests for persons with a previous positive test) appear under the Legacy
Data and Other Optional Data Elements section at the bottom of screen. After conversion, do not enter
any more data in the UPNUMTSTS field. Data can still be used for data analysis.
Examples by Source of Information:
Note: These examples are based on the revised adult case report form and the eHARS TTH form,
which do not include the first positive HIV test in this count.
Patient Interview:
Patient never had an HIV test before—interviewer should record ‘0’.
Patient indicated a previous positive but no previous negative HIV test—interviewer should
record ‘0’.
Patient self-reported previous testing—interviewer should record the number of negative tests
that occurred in the 24 months before first positive HIV test.
Patient reported having a negative test in the past but not in the last two years—question #5
should be ‘yes’ and number of tests should be ‘0’.
Patient had one test in the 24 months before first positive but never got the results—do not
count the test—interviewer should record ‘don’t know’.
Patient had two negative tests and one HIV test for which she never got the results in the 24
months before first positive—interviewer should count the two negative tests before the current
test but not the one with an unknown result.
Patient knows he did not have a negative HIV test in the in the 24 months before first positive
but doesn’t know whether he was ever tested before then—interviewer should record ‘0’ for
number of negative tests and ‘don’t know’ for Ever Had a Negative HIV test.
Patient doesn’t remember—interviewer should select ‘don’t know’.
Interviewer finds evidence of one or more previous negative tests in the 24 month period—
interviewer should record the number.
Question was never asked—interviewer should leave blank.
Medical Record Review: Field staff needs to receive training that this variable changed as of the date
of incidence conversion to eHARS and no longer includes the first positive HIV test in the count. It is
generally difficult to ascertain the number of negative tests in the past two years from chart abstraction.
In the absence of evidence to the contrary, use the current positive HIV test to calculate the number of
negative tests in the 24 month interval before the positive test.
June 2012
26
There was evidence in the medical record of at least one previous negative test in the 24
months prior to the date of first positive HIV test (but the total number is unknown)—record
‘1’. Abstractor should try to find a Date of Most Recent Negative Test and enter for #6.
There was evidence in the record of more than one negative test in the 24 month period—
abstractor should record the number of tests.
Clinic has a database with all previous tests conducted for the individual at that facility—
abstractor should record the number of negative HIV tests in the 24 month period, even if there
might have been other tests performed at other facilities.
Chart abstractor does not know if the person had a previous positive HIV test, but finds
evidence of two negative tests in the 24 months before the current positive HIV test—record
‘2’.
Chart abstractor finds that there are no patient visits in the 24 months before first positive test—
enter ‘Don’t Know’. Do not enter ‘0’.
Patient reported to provider that he had no previous HIV tests other than an indeterminate HIV
test a few weeks before the first confirmed positive HIV test—abstractor should record ‘0’.
Ignore the indeterminate test.
Patient had a false positive HIV test when she was pregnant 24 months ago and was confirmed
to be HIV negative 6 months later. She had no other HIV tests until the current one. Record ‘1’
for the negative test that occurred 18 months before the first positive test. Do NOT enter any
information for the false positive test as it was not confirmed. Abstractors should ignore any
false positive and indeterminate tests.
There is no mention of the number of negative HIV tests in the record—abstractor should
record ‘don’t know’.
Provider Report: Providers should be instructed that this variable changed as of the date of incidence
conversion to eHARS to only include negative tests in the 24 months prior to the first positive HIV
test.
Provider knows the patient never had a previous test—report ‘0’.
Patient self-reported at least one negative test in the 24 months preceding first positive test —
report ‘1’ (or more tests, depending on the number reported).
Provider has no knowledge about previous negative HIV tests—record ‘don’t know’.
Provider never asked the question of the patient—leave blank.
NHM&E/PEMS: Enter the information obtained from NHM&E/PEMS (Part 3) form or database. If
there is no information, leave blank.
If there was no patient interview, no medical chart review, no provider report, and no other follow-up,
the answer should be left blank.
1.8 TTH Data Element #8: Ever Taken Any Antiretroviral Medications (ARVs)?
Variable name and format:
UHRT (KHRT) Yes/No/Refused/Don’t Know
Wording in IVR Database:
7. In the past 6 months, have you taken any medicines shown in
the picture on the last page to treat or try to prevent HIV or
Hepatitis?
Wording on eHARS TTH
8. Ever taken any antiretroviral medications (ARVs)?
June 2012
27
Values: ‘N - No', ‘Y - Yes', ‘R – Refused’, ‘D – Do not know’
Wording on ACRF
Ever taken any antiretrovirals (ARVs)?
Wording on NHM&E/PEMS
Form
Has client used or is client currently using antiretroviral
medication (ARV)? (Part 3)
Values: Yes, No, Don’t know, Declined
Description
The variable, Ever Taken Any Antiretroviral Medications, is used to determine whether the patient took
any antiretroviral medication to prevent or treat HIV or hepatitis at any time before the collection of
the specimen used for the BED test. This variable is used in conjunction with the Date ARVs First
Began and Date of Last ARV Use TTH variables and the STARHS specimen collection date.
A list of current medications used to treat HIV is available at:
http://www.crine.org/templates/cri/pdfs/cri_pillchart_jan09_ver3.pdf
A database to search medications by name is available at:
http://aidsinfo.nih.gov/DrugsNew/Default.aspx
Purpose
This data element is important because ARV use may cause the BED results to appear ‘recent’ when
the infection is not recent. It is critical to provide dates to know whether ARVs might have had any
effect on the BED results. In the estimation model, for cases that have been exposed to ARVs in the 6
months prior to HIV disease diagnosis date, the BED results are set to missing and imputed. If ARVs
were started after the collection of the specimen used for the BED assay or ended more than 6 months
before that date, the BED results will be used as reported.
For Variant, Atypical, and Resistant HIV Surveillance (VARHS), this variable is used to determine
eligibility. Any use of ARVs on or prior to the date of HIV diagnosis will make the case ineligible for
VARHS.
Note that this data element is not being used to monitor treatment.
Sources
Sources of the information may come from patient self-report, physician’s notes, medical chart,
pharmacy records, or the AIDS Drug Assistance Program (ADAP).
History
This data element changed in 2007 when HIS transitioned from primarily conducting patient
interviews to collecting data elements from all sources. At that time, it was decided to change the data
element to ‘Ever Taken Any ARVs’ instead of recording ARV use only in the six months prior to initial
HIV diagnosis so that the variable could also be used for VARHS eligibility. This introduced some
uncertainty with the timeframe of this variable. Now this data element covers any use of ARVs before
or after diagnosis. However, only ARV use prior to the specimen collection date for the BED assay,
which could include a specimen collected within 3 months after diagnosis, is used in calculations for
incidence estimation.
June 2012
28
Data Collection
For any reported use of ARVs, even if ARV use began after the date of HIV diagnosis, record ‘yes’ but
be sure to include Date ARVs First Began (#10) and Date of Last ARV Use (#11). Even if start/end
dates are unknown, record ‘yes’. When it is unknown whether the patient ever used ARVs, record
‘don’t know’. Do not assume that the absence of ARV use information indicates that the patient never
used ARVs. Note: When there is no patient interview, it is difficult to know about previous ARV use.
If the patient has sporadic ARV use before HIV diagnosis, enter ‘yes’ and make an effort to obtain
dates of first and last use. See #10 and #11 below. If the patient did not take ARVs until after HIV
diagnosis, make an effort to obtain the month and year when medications began (even an estimated
date) that is after the eHARS HIV Disease Diagnosis Date.
Data Entry
ARV use is still relatively rare so this element is often blank or ‘don’t know’. Enter the data as
recorded on the form. If there is a question, the data entry person should consult with the Incidence
Surveillance Coordinator. If there is no information on the form, leave the field blank.
Examples by Source of Information:
Patient Interview: If the patient answers ‘yes’, it is important to ask for approximate dates when ARVs
began and ended, if appropriate (see #10 and #11).
Patient reports taking prophylactic doses (prescribed or not) of ARVs—interviewer should
enter ‘yes’.
ARV use occurred after the HIV diagnosis date but patient does not recall dates—interviewer
should enter ‘yes’. It is especially important for the interviewer to elicit approximate dates to
show that ARV use occurred after HIV diagnosis date in this case. If no date is provided for
Date ARVs First Began, then the specimen may not be eligible for VARHS.
Medical Record Review: Medical record abstractors should pay attention to dates of ARV use since
many patients will be put on ARVs shortly after initial diagnosis.
Abstractor finds ARV use before the first positive HIV test date—abstractor should enter ‘yes’.
Also record the dates Date ARVs First Began (#10) and Date of Last ARV Use (#11).
Abstractor finds ARV use began shortly after diagnosis—abstractor should record ‘yes’ and
include dates for #10 and #11.
Chart note indicates patient was part of a Pre-exposure Prophylaxis (PrEP) clinical trial—
abstractor should record ‘yes’ and include dates for #10 and #11.
Chart note indicates the patient was being treated for hepatitis—check medications list to verify
that the medication is an ARV and record ‘yes’ and include dates for #10 and #11. If the
medication is not an ARV, record ‘no’.
Chart note indicates the patient never was on ARVs—enter ‘no’.
There is no mention of whether or not ARVs were used– enter ‘don’t know’.
Provider Report: Providers should be encouraged to ask patients if they ever used ARVs, including
PrEP or non-prescribed ones, and report dates of ARV use.
Provider knows patient has taken ARVs in the past but is uncertain of the timing—provider
should record ‘yes’.
June 2012
29
Provider started patient on ARVs after date of first positive HIV test—provider should record
‘yes’ and Date ARVs First Began (#10) and Date of Last ARV Use (#11).
Patient was given ARVs after an occupational exposure—provider should report ‘yes’.
Patient reports taking trimethoprim-sulfamethoxazole (TMP-SMX, Bactrim, Septra, Cotrimoxazole), an antibiotic used to prevent Pneumocystis carinii Pneumonia (PCP) in HIV
patients that is not an ARV—report ‘no’.
If the provider has no information about ARV use, enter ‘don’t know’.
NHM&E/PEMS: Enter the information obtained from NHM&E/PEMS (Part 3) or CTS form or
database. If there is no information, leave blank.
If there was no patient interview, no medical chart review, no provider report, and no other follow-up,
the answer should be left blank.
1.9 TTH Data Element #9: Name(s) of ARV Medication Taken
1
Variable name and format:
UHRTA1 (MEDS) 2-digit code for various meds1
Wording in IVR Database:
7a. Which ones did you take? (If you are not sure of when you
took the medicines, please include the ones you MIGHT have
taken in the six months before your first positive test)
Wording on eHARS TTH
9. If yes, name(s) of ARV medication taken
Wording on ACRF
If yes, ARV medications:
Wording on PEMS Form
If yes, specify antiretroviral medication (Part 3). Note: there is
space for 4 medication codes which are listed on back of form
See Appendix B for list of medications in eHARS
Description
This variable, Name(s) of ARV Medication Taken, lists at least one of the ARV or hepatitis medications
that the patient has taken but may not include all medications used.
Purpose
The purpose of this data element is to verify that at least one medication taken was actually an
antiretroviral used to prevent or treat HIV or hepatitis. This is mostly used for verification during a
patient interview. Chart abstractors and health care providers should be able to recognize the difference
between ARVs and other medications, so this variable is less critical for collection. CDC has not used
this variable for any data analysis, but individual surveillance areas may want to know all ARVs taken.
Sources
Sources of this information may be patient self-report, physician’s notes, medical chart, or pharmacy
records. A list of current medications used to treat HIV is available at:
http://www.crine.org/templates/cri/pdfs/cri_pillchart_jan09_ver3.pdf
and a database to search names of medications is available at
http://aidsinfo.nih.gov/DrugsNew/Default.aspx.
June 2012
30
Data Collection
It is not necessary to list every drug combination that may have been used; record at least one ARV
medication. It is important to record dates of first and last use (#10 and #11).
Data Entry
Enter the medications as recorded on the form. In eHARS, hold down the Control key and select all of
the recorded medications from the pick list. If the ARV does not appear in the pick list, check the Web
sites listed above to verify the medication is an ARV. If the ARV is a new ARV drug that does not
appear on the eHARS list, enter ‘other’. If there is a question whether the medication is an ARV, the
data entry person should consult with the Incidence Surveillance Coordinator. If there is no
information on the form, leave the field blank.
Examples by Source of Information:
Patient Interview: For patients that report ‘yes’ to Ever Taken Any Antiretroviral Medications (#8), ask
which ARVs they have taken. If patient is uncertain of the names, show a pill chart. Only one ARV
medication name is required. If patient reports taking only trimethoprim-sulfamethoxazole (TMPSMX), also known as Bactrim, Septra, or Co-trimoxazole, an antibiotic used to prevent Pneumocystis
carinii Pneumonia (PCP) in persons living with HIV and which is not an ARV—interviewer should
report ‘no’ to Ever Taken Any Antiretroviral Medications (#9) and leave this field blank.
Patient names medications that are not used to treat or prevent HIV or hepatitis, probe further.
If it is apparent that the person did not take ARVs, interviewer should report ‘no’ for #8 and
leave this variable blank.
Patient does not know the names of medications taken—select ‘unspecified’ or leave blank if it
is suspected that the patient was not taking ARVs. Incidence Surveillance Coordinators need to
consider any notes made by the interviewer and exercise judgment in determining whether the
answer to Ever Taken Any Antiretroviral Medications (# 8) should be ‘yes’ or ‘don’t know’.
Medical Record Review: Examine physician notes (including patient self-report), noted prescriptions,
and pharmacy records to see if patient was taking an antiretroviral medication. List at least one med
taken. Record dates of use.
Provider note indicates patient was on ARVs before seeking treatment—abstractor should list
at least one medication.
Provider note indicates patient has taken non-prescribed drugs (e.g., medications obtained
outside of a clinical setting or “on the street”), for prevention of HIV—abstractor should list
medication or, if unknown, record ‘unspecified’.
Provider Report: Instruct providers that they do not have to list all drugs, that one is sufficient.
Patient has moved to this state and has been on various ARVs for over 5 years—provider
should enter at least one of the medications.
Patient has taken ARVs in the past but provider is not sure which ones—provider should record
‘unspecified’.
NHM&E/PEMS: Enter the information obtained from NHM&E/PEMS (Part 3) or CTS form or
database. If there is no information, leave blank.
If there was no patient interview, no medical chart review, no provider report, and no other follow-up,
the answer should be left blank.
June 2012
31
1.10
TTH Data Element #10: Date ARVs First Began
Variable name and format:
UHRTBD (KHRTBD) mm/dd/yyyy
Wording in IVR Database:
7b. What was the first day you took any of the medications
shown in the pictures? Please make your best guess if you are
not sure.
Wording on eHARS TTH
10. Dates ARVs taken: Date first began
Wording on ACRF
Dates ARVs taken: Date first began
Wording on NHM&E/PEMS
Form
Date ARV began
Description
This date, Date ARVs First Began, represents the earliest date of any ARV use.
Purpose
The purpose of this data element is to determine whether the patient took any antiretroviral medication
to prevent or treat HIV or hepatitis before or after the date that the specimen used for the BED test was
collected. This variable is important because ARV use may cause the BED results to appear ‘recent’
when the infection is not recent. Therefore, in incidence estimation, for cases that have been exposed
to ARVs in the 6 months prior to BED specimen collection date, the BED results are set to missing and
imputed. For ARV start dates after the specimen collection date, the reported BED results are
maintained.
This date is also used to determine if the patient was on ARVs on or prior to the date of diagnosis in
order to assess eligibility for VARHS.
Sources
Source for this item includes: patient self-report, chart abstraction, doctor notes, ADAP database, and
pharmacy records.
Data Collection
This data element is important for determining the period when any ARV use started. Medical record
abstractors should pay attention to start dates of ARV use, even those after initial HIV diagnosis.
Providers should be informed that the start date is the critical piece of information. Record the earliest
month and year of ARV use. Note that this date is not necessarily related to the time that the specific
medication named in #9 was taken.
Do not be concerned if ARV use has been sporadic. It is true that some BED results that were not
affected by ARVs may not be used because the Date ARVs First Began was well before HIV diagnosis
and the Date of Last ARV Use from a subsequent use is after diagnosis, but this relatively rare scenario
is an acceptable limitation.
Data Entry
Enter the data as recorded on the form. If there is a question, the data entry person should consult with
the Incidence Surveillance Coordinator. If there is no information on the form, leave the field blank.
June 2012
32
Examples by Source of Information:
Patient Interview: Enter the first date the patient used ARVs. Ask about prophylactic use (prescribed or
not).
Patient used non-prescribed ARVs intermittently in the past year—interviewer should record
the first month/year of ARV use.
Patient is uncertain about the start date—interviewer should use prompts to obtain the
estimated month and year.
Patient was diagnosed in another state 2 years ago and took ARVs until he moved to your
jurisdiction 9 months ago. Now he has been retested in order to enter care and the provider has
put the patient back on ARVs. Provider should record the earliest known date of ARV use.
Patient cannot remember date—leave blank.
Medical Record Review: Examine physician notes (including patient self-report), noted prescriptions,
and pharmacy records to find dates when patient was taking any antiretroviral medication. Record the
first date ARV use began before HIV diagnosis.
ARV use began before the eHARS HIV diagnosis date and continued after diagnosis—
abstractor should record the earliest date.
ARV use started later than HIV diagnosis—abstractor should record the earliest date.
ARV use was intermittent for several years—abstractor should record the earliest date.
ARVs used but dates are unavailable—leave date blank.
Provider Report: Providers should be encouraged to ask for dates of ARV use for patients that report
ever using ARVs. Providers should be reminded that this date is not related to the time that the specific
medication named in #9 was taken.
Provider should record first known date of ARV use, even if use has been sporadic.
If the provider has no information about ARV use dates—leave date blank.
NHM&E/PEMS: Enter the information obtained from NHM&E/PEMS (Part 3) or CTS form or
database. If there is no information, leave blank.
If the value is ‘999999’ or there was no patient interview, no medical chart review, no provider report,
and no other follow-up, the answer should be left blank.
1.11
TTH Data Element #11: Date of Last ARV Use
Variable name and format:
UHRTED (KHRTED) mm/dd/yyyy
Wording in IVR Database:
7b. What was the last day you took any of the medications
shown in the pictures? Please make your best guess if you are
not sure.
Wording on eHARS TTH
11. Date of last ARV use
Wording on ACRF
Date of last use
Wording on NHM&E/PEMS
Form
Date of last ARV use (Part 3)
June 2012
33
Description
This variable, Date of Last ARV Use, represents the date when ARVs were last taken by the patient.
The Date Patient Reported Information should be recorded as the Date of Last ARV Use if the patient
was still on ARVs as of that date.
Purpose
The purpose of this data element is to determine whether the patient took any antiretroviral medication
to prevent or treat HIV or hepatitis in the 6 months before the date that the specimen used for the BED
test was collected. This variable is important because ARV use may cause the BED results to appear
‘recent’ when the infection is not recent. Therefore, for incidence estimation, for cases that have been
exposed to ARVs in the 6 months prior to BED specimen collection date, the BED results are set to
missing and imputed in the incidence estimation program.
Sources
Sources for this item include: patient self-report, chart abstraction, doctor notes, ADAP database, and
pharmacy records.
Data Collection
Record the last date when the patient was known to be taking ARV medications, prescribed or not. If
ARVs are currently being taken, record the date when the patient was last known to be taking ARVs.
That is likely to be the same as Date Patient Reported Information (#2), the date of the interview, chart
note or provider encounter.
Data Entry
Enter the data as recorded on the form. If there is a question, the data entry person should consult with
the Incidence Surveillance Coordinator. If there is no information on the form, leave the field blank.
Examples by Source of Information:
Patient Interview: Enter the last date the patient was known to use ARVs. If the patient is currently on
ARVs, that date would be the interview date.
Patient had intermittent use—interviewer should select the last date used.
Patient took some ARVs to prevent infection but stopped two years ago—interviewer should
record the approximate date of last use, even if it is only the year of use.
Patient is uncertain about the last ARV use date—interviewer should use prompts to obtain the
estimated month and year.
Patient still taking ARVs—interviewer should record date of interview.
Medical Record Review: Examine physician notes (including patient self-report), noted prescriptions,
and pharmacy records to find date when patient was last known to be taking ARVs or stopped taking
antiretroviral medication.
Patient still taking ARVs—record the latest date of physician note or last prescription.
Patient was prescribed ARVs but discontinued use—abstractor should record the date of last
known use.
Provider Report: For patients that ever used ARVs, providers should be encouraged to ask for the last
date of ARV use.
June 2012
34
Provider has no information about ARV use dates—leave blank.
Patient is still taking ARVs—report date of last known use or date of encounter.
NHM&E/PEMS: Enter the information obtained from NHM&E/PEMS (Part 3) or CTS form or
database. If there is no information, leave blank.
If the value is ‘999999’ or there was no patient interview, no medical chart review, no provider report,
and no other follow-up, the answer should be left blank.
For required STARHS specimen and laboratory results data elements, skip to
page 44.
June 2012
35
2. Legacy Data and Other Optional TTH Data Elements
2.1
TTH Data Element #12: Are You Now Taking Any ARVs?
Variable name and format:
QHRTNW (KHRTNW) Yes/No/Refused/Don’t Know
Wording in IVR Database1:
7c. Are you now taking any of the medications shown in the
pictures?
Wording on eHARS TTH:
12. Are you now taking any ARVs?
Values: ‘N – No’, ‘Y – Yes’, ‘R – Refused’, ‘D – Do not know’
Wording on ACRF
1
Not in ACRF TTH section
From Pre-test version in IVR database
Collection and Data Entry/History
The purpose of the Are You Now Taking Any ARVs variable was to ascertain if the patient was on
ARVs at the time the specimen was collected for the BED test because ARV use could cause false
‘recent’ results of the BED test. This variable became optional as of 2007. Since TTH information
could be obtained through chart review, provider report, or interview before or after the HIV diagnosis,
this variable lacked a temporal component. For example, if the information was obtained through
medical record abstraction 4 months after diagnosis, the patient could be ‘currently’ on ARVs, though
they were not taking them at the time the specimen was collected for diagnosis and BED testing. The
dates specified in the Date ARVs First Began and Date of Last ARV Use variables are more precise
indicators.
2.2
TTH Data Element #13: Ever Tested for HIV Before Today? (Legacy Pre-test form)
Variable name and format:
(UPTESTS) Yes/No/Refused/Don’t Know
Wording in IVR Database:
4. Have you ever been tested for HIV before today?
Wording on eHARS TTH:
13. Ever been tested for HIV before today?
Values: ‘N – No’, ‘Y – Yes’, ‘R – Refused’, ‘D – Do not know’
Wording on ACRF
Not in ACRF TTH section
Collection and Data Entry
The Ever Tested for HIV before Today variable appeared on the pre-test form and was not used by the
majority of incidence surveillance areas. It was part of a series of questions for a patient interview.
When it was used as part of an interview, the variable was part of a skip pattern for a series of
questions. For analysis, if a patient reported ‘No’, then he or she should be classified as a new tester
without a previous positive test.
History
Ever Tested for HIV before Today became optional in 2007 because better information could be
obtained from other variables regarding first positive and first negative tests.
June 2012
36
2.3
TTH Data Element #14: Date of First HIV Test Ever
Variable name and format:
UFTSTD (KFTSTD) mm/dd/yyyy
Wording in IVR Database:
6. When was the very first time you ever got tested for HIV
(when you got the test, not when you got the results)? Please
make your best guess if you are not sure.
Wording in eHARS:
14. When was the first time you ever got tested for HIV?
Wording on ACRF
Not in ACRF TTH section
Collection and Data Entry/History
The purpose of the Date of First HIV Test Ever variable was to ascertain previous testing. However, it
lacked the specificity of whether the test was positive or negative. This data element became optional
in 2007 because better information could be obtained from other variables regarding date of first
positive and first negative HIV tests.
2.4
TTH Data Element #15: Was the First Positive HIV Test Anonymous?
Variable name and format:
UFPOSA (KFPOSA) Yes/No/Refused/Don’t Know
Wording in IVR Database:
4c. When you first tested positive for HIV were you given a
number or code to use to get your results instead of your name?
Wording on eHARS TTH:
15. When you first tested positive for HIV, was the HIV test an
anonymous test?
Wording on ACRF
Not in ACRF TTH section
Collection and Data Entry/History
The purpose of the Was the First Positive HIV Test Anonymous variable was to clarify the reason there
was a positive HIV test before the HIV diagnosis date. There are a number of reasons that a previous
positive test may not have been reported before; for example, the person was tested in another state,
was tested in a state that did not have HIV reporting, or had an anonymous test. This data element
became optional in 2007.
June 2012
37
2.5
TTH Data Element #16: Number of Tests 2 years before Previous Positive HIV
Test (Legacy Pre-test form)
Variable name and format:
1
UPNUMTSTS
Wording in IVR Database :
5a. For people who have had a positive test: In the two years
before your first positive HIV test (that is, the two years before
the date in question 4b), how many times did you get tested for
HIV? Your first positive test should be included for you in the
count.
Wording on eHARS TTH:
16. For persons who had a previous positive test (Pre-test Legacy only; enter new data in #7): In the two years before first
positive test, how many times did you get tested for HIV?
Values: 1-99; R, D
Wording on ACRF
1
Not in ACRF TTH section
From Pre-test version in IVR database
Collection and Data Entry/History
The Number of Tests 2 Years before Previous Positive HIV Test variable appeared on the Pre-test form
and was not used by the majority of incidence surveillance areas. The Pre-test form had two separate
variables. This variable (UPNUMTSTS) was for persons with a previous positive test where the first
positive test would be included in the count. The other variable (UNUMTSTS) was for persons with no
previous positive tests who were interviewed before the HIV test and did not know they were positive.
In the latter case, the first positive test had to be added to the count later. The purpose of this variable
is the same as Number of Negative HIV Tests within 24 Months before First Positive HIV Test (#7).
Note: For areas that used the Pre-test form, during eHARS conversion data for this variable were
converted to the variable UPNUMTSTS (eHARS column name) which can be used in analyses for
data entered prior to conversion. After conversion to eHARS, do not enter data to this field. All counts
of previous negative tests are now entered in the required variable, Number of Negative HIV Tests
within 24 Months before First Positive HIV Test (#7).
2.6
TTH Data Element #17: Reason for Today’s HIV Test (Legacy Pre-test form)
Variable names and format:
UREAS3_ 1 through UREAS3_5, UR3_5SP
Yes/No/Refused/Don’t Know
Wording in IVR Database:
3. Why are you getting the HIV test today? Are you getting the
test…
Wording in eHARS TTH
17. Reason for getting today’s HIV test (Legacy Pre-test form):
Wording on ACRF
Not in ACRF TTH section
Collection and Data Entry
The purpose of these variables, collectively named Reason for Today’s HIV Test, was to ascertain
whether the patient had reasons for getting an HIV test that would affect the probability of being tested
in the BED window period, such as being concerned about a possible recent exposure. These variables
were collected for reason for testing at the time of the interview, which occurred before the test results
were known.
June 2012
38
There are 5 standard yes/no questions and 19 other reasons for testing. The following are the standard
reasons, using current eHARS wording, with the corresponding variable in parentheses:
a. Think you might have been exposed to HIV in the past 6 months? (UREAS3_1)
b. Get tested on a regular basis and it is time to get tested again? (UREAS3_2)
c. Just checking to make sure you are HIV negative? (UREAS3_3)
d. Required to get the test by insurance, military, court, or other agency? (UREAS3_4)
e. Other reason you wanted to get tested? (UREAS3_5)
f. If other reason, describe: ____________ (UR3_5SP)
The variable for #17f., Other Reason for Today’s Test (describe), is an open text field in eHARS.
However, in the IVR database, the following were offered as commonly used other reasons. It is
recommended to use these (entered exactly as written or beginning the same way) because they can be
analyzed more easily.
The other reasons are:
1.
MD recommendation / rule out HIV diagnosis
11. Requested by partner, family, other
2.
Patient has symptoms, recent illness, wt loss, OI
12. Confirm previous HIV + test
3.
Current STD and/or STD screening
13. Needed to initiate care
4.
Named as contact or partner/ ex is HIV+/symptomatic
14. Starting/ending a relationship
5.
Possible exposure >6 months ago
15. HIV vaccine trial
6.
Incarceration
16. Immigration screen
7.
Blood/plasma donation or referred by blood bank
17. Community screening /free test /test offered
8.
Prenatal screening or pregnancy
18. Occupational exposure
9.
Hospitalization, pre-op test, other procedure
19. NAAT testing
10.
Entry to drug/alcohol treatment
History
In developing the first incidence estimates, CDC found that reasons for testing, such as possible
exposure, were not associated with a higher likelihood of having a BED recent result. As a result, these
data elements were no longer required as of 2007, but some surveillance areas continue to collect this
information for their own purposes.
These particular variables, Reason for Today’s HIV Test, only appeared on the Pre-test form. For areas
that used the Pre-test form, these reasons for testing were for newly diagnosed persons that did not
self-report a previous positive HIV test. For persons that reported having a previous positive test, a
second question about reasons was included on the Pre-test form, which appears under the following
data element, Reason for First Positive HIV Test (#18).
June 2012
39
2.7
TTH Data Element #18: Reason for First Positive HIV Test
Variable name and format:
URS4E_1 through URS4E_5 (KREAS6_1 thru _5)
Yes/No/Refused/Don’t Know
Wording in IVR Database:
4e. Why you got your first positive test? Did you get the test…
Proposed wording in eHARS
18. Reason for getting the first positive HIV test
Wording on ACRF
Not in ACRF TTH section
Collection and Data Entry
The purpose of these variables, collectively named Reason for First Positive HIV Test, was to ascertain
whether the patient was motivated to get tested by an exposure or had other reasons for testing that
would affect the probability of being tested in the BED window period. The above named variables
were collected for reason for testing at the time of the self-reported first positive test. There are 5
standard reasons (using current eHARS wording), with each a separate variable named in parentheses:
a. Thought that you might have been exposed to HIV in the 6 months before the test? (URS4E_1)
b. Got tested on a regular basis and it was time to get tested again? (URS4E_2)
c. Just checking to make sure you were HIV negative? (URS4E_3)
d. Required to get the test by insurance, military, court, or other agency? (URS4E_4)
e. Other reason you wanted to get tested? (URS4E_5)
f. If other reason, describe: ____________ (URS4E_5SP)
The variable for 18f., Other Reason for First Positive HIV Test (describe), is an open text field in
eHARS. However, in the IVR database, the following were offered as commonly used other reasons. It
is recommended to use these (entered exactly as written or beginning the same way) because they can
be analyzed more easily. The other reasons are:
1.
MD recommendation / rule out HIV diagnosis
11. Requested by partner, family, other
2.
Patient has symptoms, recent illness, wt loss, OI
12. Confirm previous HIV + test
3.
Current STD and/or STD screening
13. Needed to initiate care
4.
Named as contact or partner/ ex is HIV+/symptomatic
14. Starting/ending a relationship
5.
Possible exposure >6 months ago
15. HIV vaccine trial
6.
Incarceration
16. Immigration screen
7.
Blood/plasma donation or referred by blood bank
17. Community screening /free test /test offered
8.
Prenatal screening or pregnancy
18. Occupational exposure
9.
Hospitalization, pre-op test, other procedure
19. NAAT testing
10.
Entry to drug/alcohol treatment
History
In developing the first incidence estimates, CDC found that reasons for testing, such as exposure, were
not associated with a higher likelihood of having a BED recent result. As a result, these data elements
were no longer required as of 2007, but some surveillance areas continue to collect this information.
The questions were asked once on the Post-test form but were included on the Pre-test form in addition
to the reasons for “today’s test” for those persons who reported having a previous positive HIV test. At
conversion, the Post-test variables for Reason for First Positive HIV Test (KREAS6_1 through _5)
June 2012
40
were mapped to the Pre-test variables for Reason for First Positive HIV Test (URS4E_1 through
URS4E_5).
2.8
TTH Data Element #19: Name of Facility Where First Tested Positive for HIV
Variable name and format:
UFPS_SITE (KFPS_SITE) Text field
Wording in IVR Database:
4d. What was the name of the place where you got your first
positive HIV test (on the date in question 4b)? For example,
this could be the name of a health clinic, blood bank, doctor’s
office, or STD clinic. Site Name:
Wording in eHARS TTH:
19. Name of facility where first tested positive for HIV
Wording on ACRF
Not in ACRF TTH section
Collection and Data Entry/History
The Name of Facility Where First Tested Positive for HIV variable is for local use to record the facility
where the person first tested positive for HIV. It can be used to monitor the sources of new diagnoses
as well as provide information for follow-up activities (e.g., chart review). This data element is not
transferred to CDC through eHARS.
This data element was no longer required as of 2007, but some surveillance areas continue to collect
this information for local use.
2.9
TTH Data Element #20: State of Facility Where First Tested Positive for HIV
Variable name and format:
UFPS_STATE (KFPS_STATE) 2-digit State code
Wording in IVR Database:
4d. Site State:
Wording in eHARS TTH:
20. State of facility where first tested positive for HIV
Wording on ACRF
Not in ACRF TTH section
Collection and Data Entry/History
The State of Facility Where First Tested Positive for HIV variable is used to record the state where the
person first tested positive for HIV. It can be used to explain why the self-reported diagnosis date is
different from the eHARS HIV Disease Diagnosis Date. This data element is not transferred to CDC.
This data element is no longer required as of 2007, but some surveillance areas continue to collect this
information for local use. For persons previously tested in other countries, “Foreign Country” is not an
available choice on the eHARS drop down menu, but information can be stored in a local field.
June 2012
41
2.10
TTH Data Element #21: Type of Facility Where First Tested Positive for HIV
Variable name and format:
UFPSTYP (KFPS) eHARS FACILITY_TYPE code
Wording in IVR Database:
4d. Site Type:
Wording in eHARS TTH:
21. Type of facility where first tested positive for HIV
Wording on ACRF
Not in ACRF TTH section
Collection and Data Entry/History
The Type of Facility Where First Tested Positive variable represents the type of testing facility where
the person was first tested for HIV. This data element was to be used, in conjunction with the eHARS
Facility Type at HIV Diagnosis, as a variable used in multiple imputation of missing information for
testing group and BED results during calculation of incidence estimates, but because the variable is
also included on the ACRF in eHARS, it is no longer required for HIS as of 2007. Some surveillance
areas continue to collect this information for local use.
2.11
TTH Data Element #22: Name of Facility Where Last Tested Negative for HIV
Variable name and format:
ULSTNGS_SITE (KLSTNGS_SITE) text
Wording in IVR Database:
4g. What was the name of the place where you had your last
negative HIV test? For example, this could be the name of a
health clinic, blood bank, doctor’s office, or STD clinic. Site
Name:
Wording in eHARS TTH:
22. Name of facility where last tested negative for HIV
Wording on ACRF
Not in ACRF TTH section
Collection and Data Entry
The Name of Facility Where Last Tested Negative for HIV variable represents the place where the
person last had a negative HIV test. This information is not transferred to CDC through eHARS. This
data element is no longer required as of 2007, but some surveillance areas continue to collect this
information for local use.
2.12
TTH Data Element #23: State of Facility Where Last Tested Negative for HIV
Variable name and format:
ULSTNGS_STATE (KLSTNGS_STATE) 2-digit State code
Wording in IVR Database:
4g. Site State:
Wording in eHARS TTH:
23. State of facility where last tested negative for HIV
Wording on ACRF
Not in ACRF TTH section
The State of Facility Where Last Tested Negative for HIV variable represents the state where the
person last had a negative HIV test. This data element is no longer required as of 2007, but some
surveillance areas continue to collect this information for local use. For persons previously tested in
other countries, “Foreign Country” is not an available choice on the eHARS drop down menu, but
information can be stored in a local field.
June 2012
42
2.13
TTH Data Element #24: Type of Facility Where First Tested Negative for HIV
Variable name and format:
ULSTNGS (KLSTNGS) eHARS FACILITY_TYPE code
Wording in IVR Database:
4g. Site Type:
Wording in eHARS TTH:
24. Type of facility where last tested negative for HIV
Wording on ACRF
Not in ACRF TTH section
The Type of Facility Where Last Tested Negative for HIV variable represents the type of facility where
the person last had a negative HIV test. This data element is no longer required as of 2007, but some
surveillance areas continue to collect this information for local use.
3. IVR Database TTH Variables Not Included in eHARS TTH Document
3.1 Date of HIV Test (reference date)
Variable name and format:
UPTESTD (KPTSTD) mm/dd/yyyy
Wording in IVR Database:
2. Date of this HIV test
Wording in eHARS TTH:
Not in eHARS
Wording on ACRF
Not in ACRF TTH section
Collection and Data Entry/History
The purpose of the Date of HIV Test variable was to assist interviewers and field staff in the collection
of data related to TTH around time of HIV diagnosis. It was used to note the date of the current
positive HIV test in order to ask questions related to events before this date during an interview or
chart abstraction. This variable became optional in 2007 and does not appear in eHARS. Data will not
be converted into eHARS.
June 2012
43
4. Required and Optional Data Elements for STARHS Specimen
Information and Results
These data elements are presented in the order that they appear on the Lab Data tab of the eHARS
Laboratory Report document. In eHARS, STARHS specimen and results information are entered like
any other laboratory test. Therefore, some of these variables have multiple uses, depending on the type
of laboratory test that is selected, and will have generic labels on the eHARS Laboratory document.
When a STARHS laboratory test type is selected in eHARS, it opens a screen with all of the STARHS
laboratory variables. Certain variables are required to enter a laboratory document and are noted here
as ‘Required for eHARS.’ Variables that are required for HIS are so designated.
There can be multiple laboratory documents for the same specimen; for example, one for specimen
information entered on one date and one for the results imported at a later date, each with the same
STARHS ID. For analysis, the data from all Lab documents for a unique STARHS ID will be
combined.
4.1 STARHS Laboratory Name
Variable name and format:
CLIA_UID1 (LABID)2 CLIA code for STARHS lab
Wording in IVR Database:
LabID
IVR Values:
21D0649758 – Maryland Department of Health Laboratory
50D0661430 – University of Washington Laboratory
33D0654341 – Wadsworth Center–Biggs Laboratory
(Legacy)
eHARS Value:33D2005937 – Wadsworth Center–David
Axelrod Institute (Use in eHARS)
Wording in eHARS Lab Document:
1
eHARS variable
2
IVR database variable
Lab Name
Description
For HIV incidence surveillance, the STARHS Laboratory Name variable represents the name of the
laboratory that conducted STARHS testing. Currently, all tests are conducted at the single CDCfunded STARHS laboratory using BED HIV-1 Capture EIA (BED), but in the past some other
laboratories performed STARHS testing using Vironostika HIV-1 EIA (Vironostika-LS).
The list of laboratory names in eHARS is created by the local eHARS Administrator. For most
STARHS specimens, the value for STARHS Laboratory Name will be the name of the CDC-funded
STARHS laboratory, currently the Wadsworth Center – David Axelrod Institute.
Purpose
The purpose of this variable is to identify the name of the laboratory which performed STARHS
testing. In the IVR and eHARS databases, the laboratory name is stored as a unique Clinical
Laboratory Improvement Amendments (CLIA) code assigned to that laboratory.
June 2012
44
Source
Sources for STARHS Laboratory Name are tracking databases and the STARHS results file received
from the STARHS laboratory.
Data Entry
The STARHS Laboratory Name is the name of the laboratory performing the HIV-related test, in this
case, the STARHS testing algorithm. This data element may be manually entered or imported. The
STARHS Laboratory Name can be selected from the list of laboratory names that appear on the eHARS
screen (user interface) in the local eHARS installation. The value can also be imported from the
STARHS Results file received from the STARHS laboratory using the All Document Import (ADI)
Laboratory Report Default template. The SAS program used to import STARHS data will translate the
values to the appropriate CLIA code.
Examples
Wadsworth Center – David Axelrod Institute – 33D2005937
4.2 Source Lab Specimen ID
Variable name and format:
SAMPLE_ ID (LSRCEID)
Wording in IVR Database:
Source Lab Specimen ID
Wording in eHARS Lab Document:
Sample ID (Specimen)
Description
The Source Lab Specimen ID variable was originally intended to capture the specimen ID from the
originating laboratory for a STARHS specimen, usually a commercial laboratory. However, this data
element is for local use and can be used in any way that facilitates identifying a specific specimen.
Specimens may have several specimen numbers.
Purpose
This variable is one of two used locally to track specimen identifiers: Source Lab Specimen ID
(LSRCEID) for the originating or commercial laboratory and State Lab Specimen ID (SSTATEID) for
public health laboratory identifiers. In eHARS, there are two generic specimen IDs that can be used,
Sample ID (SAMPLE_ID) and Accession Number (ACCESSION_NO). At conversion to eHARS, the
IVR variable, Source Lab Specimen ID, will be mapped to SAMPLE_ID and State Lab Specimen ID
will be mapped to ACCESSION_NO. CDC does not use these data elements.
Data Entry
Local surveillance areas should decide which specimen ID will be entered for this variable. The main
goal is to be consistent; therefore, these decisions should be included in the Standard Operating
Procedures (SOPs). These data may be manually entered or imported using the ADI Laboratory Report
Default template.
June 2012
45
4.3 State Lab Specimen ID (or Other Specimen ID)
Variable name and format:
ACCESSION_NUMBER (SSTATEID)
Wording in IVR Database:
State Lab Specimen ID
Wording in eHARS Lab Document:
Accession Number
Description
The State Lab Specimen ID variable was intended to capture the specimen ID from the State Public
Health laboratory. However, this data element is for local use and can be used in any way that
facilitates identifying a specific specimen.
Purpose
This variable is also used to track specimens, particularly from State Public Health laboratories. At
conversion the IVR variable, SSTATEID, will be mapped to the eHARS field labeled Accession
Number (ACCESSION_NO).
Data Entry
Local surveillance areas should decide what specimen IDs will be entered for this variable. To be
consistent, these decisions should be included in the SOPs. These data may be manually entered or
imported using the ADI Laboratory Report Default template.
4.4 Date of Specimen Collection (Required for HIS)
Variable name and format:
SAMPLE_DT1 (LDTEOBT)2 mm/dd/yyyy
Wording in IVR Database:
Date of Specimen Collection
Wording in eHARS Lab Document:
Collection date
Description
The Date of Specimen Collection variable is the date the specimen was drawn for the HIV-related test
that was used for STARHS testing. This is the draw date of the remnant specimen from the initial HIV
diagnostic specimen or another HIV-related test within 3 months of HIV diagnosis.
Purpose
The purpose of this data element is to determine eligibility of the specimen for STARHS testing. It is
an HIS-required variable for STARHS specimens. This data element is compared to the eHARS
calculated HIV Disease Diagnosis Date (Calc OBS 285) which includes earliest AIDS date in the
calculation. If the date is later than 3 months from HIV diagnosis, this specimen should not be sent for
STARHS testing. If it is tested, the STARHS result from this specimen is not used in the HIV
incidence estimation.
Sources
Sources of this date are paper or electronic laboratory reports and laboratory tracking databases.
June 2012
46
Data Entry
The Date of Specimen Collection variable can be entered manually or imported using the ADI
STARHS Lab template. This variable is required for HIS. When entering any lab document in eHARS,
if there is no specimen collection date entered there will be a warning given after submitting the
document. However, the specimen data will be stored in eHARS. For HIS, there must be a collection
date on at least one lab document for a given STARHSID for the results to be used in data analysis or
incidence estimation.
4.5 STARHS Result Date
Variable name and format:
RESULT_RPT_DT (TESTDATE) mm/dd/yyyy
Wording in IVR Database:
Test Date
Wording in eHARS Lab Document:
Result Date
Description
The STARHS Result Date variable is the date that the STARHS lab performed the STARHS test on the
remnant sample.
Purpose
This variable is not used for data analysis but is useful for data timeliness evaluation, monitoring the
time from reporting date or shipping date to STARHS testing.
Sources
The source of this data element is the STARHS Results file sent by the STARHS laboratory.
Data Entry
These data may be manually entered in eHARS or imported using ADI Laboratory Report Default
template.
4.6 Received Date
Variable name and format:
RECEIVE_DT mm/dd/yyyy
Wording in IVR Database:
Not in IVR Database
Wording in eHARS Lab Document:
Received Date
Description
The Received Date variable is an eHARS data element for the date the laboratory result was received at
the health department. This variable did not appear in the IVR database.
Purpose
This data element is not used by incidence surveillance. Local surveillance areas may choose to use it
to monitor the time when STARHS results were received from the STARHS lab. Otherwise, the
variable can be left blank.
June 2012
47
Data Entry
These data may be manually entered or imported using the ADI Laboratory Report Default template.
4.7 STARHS Assay (Required for eHARS)
Variable name and format:
LOINC_CD (ASSAY)
Wording in IVR Database:
Assay
Wording in eHARS Lab Document:
Test
Values available in eHARS
ST-001
ST-002
ST-003
ST-888
ST-999
=
=
=
=
=
BED
Vironostika
Avidity
Other
Unknown (Blank in IVR database)
Description
For eHARS, this variable is used to designate what lab test is being reported. For incidence
surveillance, this variable, STARHS Assay, designates the assay used for the STARHS algorithm. The
assay type is designated by a LOINC code. This variable is required to enter any STARHS specimen
information in eHARS.
Purpose
In eHARS, STARHS specimen and results information are entered like any other laboratory test;
therefore, eHARS requires the selection of one of the codes for STARHS tests when entering HIS data.
For HIS, the purpose of this variable is to distinguish various STARHS testing algorithms that have
been used. Until March 2005, HIS used the Less Sensitive (LS) Enzyme Immunoassay (EIA) test for
STARHS. Currently, HIS uses the BED HIV-1 Capture Enzyme Immunoassay (BED) for STARHS. In
the future, other tests, such as an avidity test, might be used.
Sources
This variable was imported with the STARHS laboratory results into the IVR database. Because the
only current STARHS assay is the BED, the value of ST-001 can imported or manually entered from a
drop-down menu as STARHS (BED).
History
The value of ST-999 (STARHS Unknown) was used during the conversion of data into eHARS for any
STARHS result that had a blank value. Note: Do not use STARHS (Unknown), when entering
information for a specimen that has not yet been sent for STARHS testing or does not yet have a
STARHS result. Use STARHS (BED). Most SAS programs exclude non-BED results, so the entered
information entered as STARHS (Unknown) will be lost in data analysis.
Data Entry
These data may be manually entered or imported using the ADI Laboratory Report Default template.
June 2012
48
4.8 Specimen Type (Recommended for HIS)
Variable name and format:
SPECIMEN (LSPECTY)
Wording in IVR Database:
Specimen Type
Values:
‘1-blood finger stick’
‘2-blood venipuncture’
‘3-blood spot’
‘4-oral mucosal transudate’
‘5-urine’
‘8-other’
‘9-unknown’
Wording in eHARS Lab Document:
Sample Type
Values: ‘BLD’, ‘SAL’, ‘URN’, ‘OTH’, ‘UNK’
Description
The Specimen Type variable represents the type of specimen that was obtained from the patient, such
as blood or oral fluid. In the IVR database, there were seven values that could be selected but eHARS
only offers five.
Purpose
This data element is useful to ascertain if the specimen was blood (serum/plasma) and therefore
eligible for BED testing. This data element is available for all HIV tests in eHARS. CDC recommends
that incidence surveillance areas start using this variable in eHARS for all diagnostic or other
specimens for HIV tests within 3 months of diagnosis to indicate whether or not the specimen could be
used for STARHS. This information will be very useful in tracking eligible specimens and, for
evaluation, to quantify the denominator for the proportion of cases with a STARHS result among those
with blood or serum specimen. This is the incidence evaluation standard as stated in Technical
Guidance for HIV/AIDS Surveillance Programs:
At least 85% of new diagnoses of HIV infection reported for a calendar year diagnosed using a
serum/plasma specimen or having a follow-up HIV-related test conducted on a serum/plasma
specimen within 3 months of the diagnosis should have a specimen transported to the CDC
STARHS laboratory, assessed at 12 months after the end of the diagnosis year.
Sources
Sources for this data element are paper or electronic laboratory reports or databases.
Data Entry
These data may be manually entered or imported using the ADI Laboratory Report Default template.
June 2012
49
4.9 STARHS ID (Required for eHARS and HIS)
Variable name and format:
STARHS_SAMPLE_ID (SSTARHSID)
Wording in IVR Database:
StarhsID
Wording in eHARS Lab Document::
STARHS ID (Specimen)
Description
The STARHS ID variable is the identifier assigned to a specimen in order to be STARHS tested. There
is only one STARHS ID assigned to a specimen that is sent for testing. If the specimen cannot be
tested (e.g., due to insufficient quantity), the same STARHS ID cannot be re-used. An individual may
have more than one specimen, each with a unique STARHS ID, even if the aliquots were from the
same HIV diagnostic specimen.
Purpose
This variable is used to link specimen information with STARHS results received from the STARHS
laboratory.
Sources
Sources for this variable are health departments, public health laboratories, tracking databases, and the
Results file received from the STARHS laboratory.
Data Entry
These data may be manually entered or imported into eHARS using the ADI Laboratory Report
Default template. Data entry staff will have to manually enter the STARHS ID on an eHARS laboratory
document in order to link the STARHS results file, which has no STATENO, to cases in eHARS
during a separate import process. Otherwise, the value can be imported using a SAS program that will
combine specimen information from a tracking database, STATENO, and the STARHS Results file
received from the STARHS laboratory.
4.10 Standard Optical Density
Variable name and format:
RESULT (SOD)
Wording in IVR Database:
SOD
Wording in eHARS Lab Document:
Standard Optical Density
Description
The Standard Optical Density variable represents the specific value of SOD found by STARHS
testing.
Purpose
This variable is used to substantiate the final result interpretation described in the data element below,
Final STARHS Result.
June 2012
50
Sources
The only source for this data element is the STARHS result file received from the STARHS
laboratory.
Data Entry
CDC strongly recommends that this variable be imported into eHARS using the ADI Laboratory
Report Default template to avoid errors.
4.11 Final STARHS Result (Required for HIS)
Variable name and format:
RESULT_INTERPRETATION (RESULT)
Wording in IVR Database:
Lab Result
Wording in eHARS Lab Document:
Final Result
Values:
01 - Long term
02 – Recent
91 – Quantity not sufficient
92 – Specimen never received
93 – Broken in transit
94 – Other, indeterminate
95 – Not sufficient antibodies
99 – Undefined result
Description
The Final STARHS Result variable represents the main BED testing result. It also contains values to
explain why a STARHS specimen does have a STARHS result.
Purpose
This data element is used to classify newly-diagnosed HIV infections as recent or long-standing (long
term). This is one of the most important variables for HIS and is required for specimens that receive
STARHS testing.
If the result of STARHS testing is not recent or long term, a reason for not testing is reported by the
STARHS laboratory and represented by the codes 91–99 described above. These codes are used
exclusively by the STARHS laboratory. Reasons for not shipping specimens to the STARHS
laboratory are captured in Reason Specimen Not Sent for STARHS, which is described later in this
document.
Sources
The only source for this data element is the STARHS result file received from the STARHS
laboratory.
Data Entry
CDC strongly recommends that this variable be imported into eHARS using the ADI Laboratory
Report Default template to avoid errors. Do NOT manually enter a reason for not testing in the eHARS
Final STARHS Result field.
June 2012
51
4.12 Reason Specimen Not Sent for STARHS
Variable name and format:
SREASON
Wording in IVR Database:
Reason Specimen Not Tested
Wording in eHARS Lab Document:
Reason Specimen Not Sent for STARHS
Description
The Reason Specimen Not Sent for STARHS variable represents the reason why a specimen was not
sent to the STARHS laboratory for STARHS testing. In the IVR database, this variable also
represented the reason why the STARHS laboratory did not have a result, and was imported from the
STARHS result file received from the STARHS laboratory. In eHARS, those two types of reasons will
be represented by separate variables, Reason Specimen Not Sent for STARHS and Final STARHS
Result.
Purpose
This data element is to explain and track reasons why the specimen was not sent for STARHS testing.
This is useful for resolving STARHS specimens without results, monitoring problems with laboratories
that send specimens, and evaluation of the incidence surveillance area’s effort in obtaining specimens
for STARHS testing.
Data Entry
Information on why a specific specimen was not sent to the STARHS laboratory for STARHS testing
should be manually entered in eHARS. Do not assume, without information, that a specimen had
insufficient quantity and that is why a laboratory did not send the specimen to the public health
laboratory or STARHS laboratory. Only enter a reason when there is sufficient information to support
it.
5. IVR Lab Variables Not Included in eHARS
5.1 Specimen Approved for STARHS
2
Variable name and format:
(LAPPRVE)2 Yes/No/Don’t Know
Wording in IVR Database:
Approved for STARHS
Wording in eHARS Lab Document:
Not in eHARS
IVR database variable
Description
The Approved for STARHS variable designated that a specific specimen was eligible for STARHS
testing, either because it was collected within 3 months of date of HIV diagnosis or consent was
obtained. In the macro that assembled the monthly data transfer, CDC used the variable to select which
STARHS specimen result would be transferred to CDC until July 2010.
Purpose
This variable is no longer used by CDC and it is not in eHARS.
June 2012
52
5.2 State Lab ID
2
Variable name and format:
(SSTATELID)2 CLIA code for State Laboratory
Wording in IVR Database:
State Lab ID
Wording in eHARS Lab Document:
Not in eHARS
IVR database variable
Description
The State Lab ID variable was used locally to link specimens to a public health laboratory.
Purpose
This variable was never used by CDC and it is not in eHARS.
5.3 HIV Diagnosis Test Type
Variable name and format:
(LTSTYP)2
Wording in IVR Database:
State Lab ID
Values:
5220-9
21009-6
5472-6
25835-0
5017-9
25836-8
Wording in eHARS Lab Document:
2
= EIA / Elisa
= Western Blot
= CD4
= Viral Load (NASBA)
= Viral Load (bDNA)
= Viral Load (RT-PCR)
Not in eHARS
IVR database variable
Description
The HIV Diagnosis Test Type variable was used in the IVR database to indicate the type of
confirmatory HIV test from which the STARHS specimen might be obtained.
Purpose
This variable, used locally for tracking potential STARHS specimens, was never used by CDC and it is
not in eHARS. All HIV tests are entered separately into eHARS ACRF or Lab documents and can be
found in the Document datasets.
5.4 Results Received
2
Variable name and format:
(RESULTRECEIVED)2 Yes/No
Wording in IVR Database:
Result Received
Wording in eHARS Lab Document:
Not in eHARS
IVR database variable
June 2012
53
Description
The Results Received variable was used in the IVR database to indicate that a specimen had a
STARHS result imported into the database.
Purpose
This variable, used for monitoring STARHS result, was never used by CDC and it is not in eHARS.
June 2012
54
Appendices
Appendix A: Quick Reference for Data Analysts
Appendix B: List of ARV Medications for TTH
Appendix C: Data Entry Recommendations for HIV Incidence Surveillance Data Elements in
eHARS
June 2012
55
Appendix A
Quick Reference for Data Analysts
Table A.1: Testing and Treatment History Document in HIS and eHARS Datasets
HIS Dataset Before Conversion
IVR Screen
Question
Number1
eHARS Document Dataset After Conversion
Length
SAS Format2 SAS Label3
eHARS Screen Label on
TTH Document
Variable Name
(Pre/Post)
Valid Values
SAS
Format
Variable Name
Valid Values
UCTS
1-Provider Report
$1
2-Patient Interview
3-Medical Record Review
4-NHM&E/PEMS
5-Other
$2
Main source of
TTH
Pre
Post
1
1
1. Main source of testing and
treatment history information
UCTS/ KCTS
0- No
1-Yes
2-Patient Interview
3-Medical Record
Review
4-From PEMS
5-Other
$50
1
1
2. Date patient reported
information
UQINTD/ KQINTD
Date
yyyymmdd UQINTD
$8
mm/dd/yyyy
$10
yyyymmdd
Date patient
reported info
4a
N/A
3. Ever had a previous positive UPASTP
HIV test?
1=Yes
0=No
7=Refused
9=Don't know
$1
Y=Yes
N=No
R=Refused
D=Don't know
$1
$YNRD
Ever previous
positive test
4b
3
4. Date of first positive HIV test UFPOSD/ KFPOSD
Date
yyyymmdd UFPOSD
$8
mm/dd/yyyy
$10
yyyymmdd
Date first
positive test
4f
8
5. Ever had a negative HIV test? UNGTST/ KNGTST
1=Yes
0=No
7=Refused
9=Don't know
$1
Y=Yes
N=No
R=Refused
D=Don't know
$1
$YNRD
Ever tested
negative
4f
8a
6. Date of last negative HIV test ULSTND/ KLSTND
Date
mm/dd/yyyy
$10
yyyymmdd
5b
9
7. Number of negative HIV tests UNUMTSTS/
within 24 months before first
KNUMTSTS
positive test (Do not include first
positive HIV test)
1 to 99
R=Refused
D=Don't know
yyyymmdd ULSTND
$8
$2
UNUMTSTS
0 to 99
R=Refused
D=Don't know
$2
7
10
8. Ever taken any antiretroviral UHRT/ KHRT
medications (ARVs)?
1=Yes
0=No
7=Refused
9=Don't know
$1
UHRT
Y=Yes
N=No
R=Refused
D=Don't know
$1
7a
10a
9. If yes, name(s) of ARV
medication taken
Same as Appendix B, $2
not including 32-36
UHRTA1
See Appendix B
$256
$DRUG
Names of
(piped values) ARVs taken
7b
10b
10. Dates ARVs taken: Date first UHRTBD/ KHRTBD
began
mm/dd/yyyy
$10
yyyymmdd
January 2011
PREMED1-N/
POSTMED1-N4
Date
UPASTP
UNGTST
yyyymmdd UHRTBD
$8
Date last
negative test
$2,
Number
restricted to
negative tests
integer 1-99 or 24 mos.
R or D
before first
positive
$YNRD
Ever taken
any ARVs
Date first
began ARVs
56
HIS Dataset Before Conversion
IVR Screen
Question
Number1
Pre
Post
7d
10d
eHARS Document Dataset After Conversion
Valid Values
Length
SAS Format2 SAS Label3
yyyymmdd UHRTED
$8
mm/dd/yyyy
$10
yyyymmdd
Date last ARV
use
eHARS Screen Label on
TTH Document
Variable Name
(Pre/Post)
Valid Values
SAS
Format
11. Date of last ARV use
UHRTED/ KHRTED
Date
Variable Name
Optional/Legacy Variables
7c
10c
12. Are you now taking any
ARVs?
QHRTNW/ KHRTNW
1=Yes
0=No
7=Refused
9=Don't know
$1
QHRTNW
Y=Yes
N=No
R=Refused
D=Don't know
$1
$YNRD
Now taking
ARVs
4
N/A
13. Ever been tested for HIV
before today? (Legacy Pre-test
only)
UPTESTS
1=Yes
0=No
7=Refused
9=Don't know
$1
UPTESTS
Y=Yes
N=No
R=Refused
D=Don't know
$1
$YNRD
Ever tested for
HIV before
6
7
14. When was the first time you UFTSTD/ KFTSTD
ever got tested for HIV?
Date
yyyymmdd UFTSTD
$8
mm/dd/yyyy
$10
yyyymmdd
Date first HIV
test
4c
4
15. When you first tested
positive for HIV, was the HIV
test an anonymous test?
1=Yes
0=No
7=Refused
9=Don't know
$1
UFPOSA
Y=Yes
N=No
R=Refused
D=Don't know
$1
$YNRD
Was first
positive test
anonymous
5a
N/A
16. In the two years before first UPNUMTSTS
positive test, how many times
did you get tested for HIV?
1 to 99
R=Refused
D=Don't know
$2
UPNUMTSTS
0 to 99
R=Refused
D= Don't know
$2
$2,
restricted to
integer 1-99 or
R or D
3a
N/A
1=Yes
0=No
7=Refused
9=Don't know
$1
UREAS3_1
Y=Yes
N=No
R=Refused
D=Don't know
$1
$YNRD
3b
N/A
17. Reasons for getting today's UREAS3_1
HIV test (Legacy Pre-test form
only): a. Think you might have
been exposed to HIV in the 6
months before the test
b. Get tested on a regular basis UREAS3_2
and it is time to get tested again
Number of
tests 2 yrs
before
previous
positive
(legacy pretest)
Today’s test
reason:
exposed past
6 mos.
1=Yes
0=No
7=Refused
9=Don't know
$1
UREAS3_2
Y=Yes
N=No
R=Refused
D=Don't know
$1
$YNRD
Today’s test
reason: tests
regularly
3c
N/A
c. Just checking to make sure
you are HIV negative
UREAS3_3
1=Yes
0=No
7=Refused
9=Don't know
$1.
UREAS3_3
Y=Yes
N=No
R=Refused
D=Don't know
$1
$YNRD
Today’s test
reason:
checking if
negative
3d
N/A
d. Required to get the test by
insurance, military, court, or
other agency
UREAS3_4
1=Yes
0=No
7=Refused
9=Don't know
$1.
UREAS3_4
Y=Yes
N=No
R=Refused
D=Don't know
$1
$YNRD
Today’s test
reason:
required by
agency
June 2012
UFPOSA/ KFPOSA
57
HIS Dataset Before Conversion
IVR Screen
Question
Number1
eHARS Document Dataset After Conversion
eHARS Screen Label on
TTH Document
Variable Name
(Pre/Post)
Valid Values
SAS
Format
Variable Name
Valid Values
Length
SAS Format2 SAS Label3
Pre
Post
3e
N/A
e. Other reason you wanted to
get tested
UREAS3_5
1=Yes
0=No
7=Refused
9=Don't know
$1
UREAS3_5
Y=Yes
N=No
R=Refused
D=Don't know
$1
$YNRD
Today’s test
reason: other
3e
N/A
f. If other reason, describe:
UR3_5SP
Text field
$50
UR3_5SP
Text
$50
text field
Today’s test
reason:
describe
4e(a)
6a
18. Reason for getting the first
positive HIV test: a. Thought
you might have been exposed
to HIV in the 6 months before
the test
URS4E_1/ KREAS6_1 1=Yes
0=No
7=Refused
9=Don't know
$1
URS4E_1
Y=Yes
N=No
R=Refused
D=Don't know
$1
$YNRD
First positive
reason:
exposed past
6 mos.
4e(b)
6b
b. Got tested on a regular basis URS4E_2/ KREAS6_2 1=Yes
and it was time to get tested
0=No
again
7=Refused
9=Don't know
$1
URS4E_2
Y=Yes
N=No
R=Refused
D=Don't know
$1
$YNRD
First positive
reason: tests
regularly
4e(c)
6c
c. Just checking to make sure
you were HIV negative
URS4E_3/ KREAS6_3 1=Yes
0=No
7=Refused
9=Don't know
$1
URS4E_3
Y=Yes
N=No
R=Refused
D=Don't know
$1
$YNRD
First positive
reason:
checking if
negative
4e(d)
6d
d. Required to get the test by
insurance, military, court, or
other agency
URS4E_4/ KREAS6_4 1=Yes
0=No
7=Refused
9=Don't know
$1
URS4E_4
Y=Yes
N=No
R=Refused
D=Don't know
$1
$YNRD
First positive
reason:
required by
agency
4e(e)
6e
e. Other reason you wanted to
get tested
URS4E_5/ KREAS6_5 1=Yes
0=No
7=Refused
9=Don't know
$1
URS4E_5
Y=Yes
N=No
R=Refused
D=Don't know
$1
$YNRD
First positive
reason: other
4e(e)
6e
f. If other reason, describe:
URS4E_5SP/
KReas6_5SP
Text field
$50
URS4E_5SP
Text
$50
$50
First positive
reason:
describe
4d
5
19. Name of facility where first
tested positive for HIV
UFPSTYP_SITE/
KFPS_SITE
Text field
$50
UFPS_SITE
Text
$50
$50.
Facility name
where first
tested positive
4d
5
20. State of facility where first
tested positive for HIV
UFPSTYP_STATE/
KFPS_STATE
2-digit State codes
$2
UFPS_STATE
2-digit State codes
$2
$STATE
4d
5
21. Type of facility where first
tested positive for HIV
UFPSTYP/ KFPS
Same as eHARS
Facility codes2
$7
UFPSTYP
eHARS Facility codes2
$7
$FAC_TYP
State where
first tested
positive
Facility type
where first
tested positive
June 2012
58
HIS Dataset Before Conversion
IVR Screen
Question
Number1
eHARS Document Dataset After Conversion
eHARS Screen Label on
TTH Document
Variable Name
(Pre/Post)
Valid Values
SAS
Format
Variable Name
Valid Values
Length
SAS Format2 SAS Label3
Text
$50
$50.
Facility name
where last
tested
negative
$2
$STATE
State where
last tested
negative
$7
$FAC_TYP
Facility type
where last
tested
negative
Pre
Post
4g
8b
22. Name of facility where last
tested negative for HIV
ULSTNGS_SITE/
KLSTNGS_SITE
Text field
$50
ULSTNGS_SITE
4g
8b
23. State of facility where last
tested negative for HIV
ULSTNGS_STATE/
KLSTNGS_STATE
2-digit State codes
$2
ULSTNGS_STATE 2-digit State codes
4g
8b
24. Type of facility where last
tested negative for HIV
ULSTNGS/ KLSTNGS Same as eHARS
Facility codes2
$7
ULSTNGS
eHARS Facility codes2
1
Where a question has no number in the IVR database, the approximate position is given
For SAS formats, see eHARS Technical Reference Guide, Chapter 17
3
SAS labels not included in eHARS version 3.2
4
The dataset may contain up to N medications for a given TTH; each will be stored in a separate variable
2
Table A.2: Consent Document in HIS and eHARS Datasets
HIS Dataset Before Conversion
eHARS Document Dataset After Conversion
eHARS Screen Label on
Consent Document
Variable Name
Valid Values
SAS Format
Variable Name
Valid Values
Length
SAS Format1 SAS Label
Date of first approach for
Consent
C1DATE
Date
yyyymmdd
$10
CDATE1
mm/dd/yyyy
$8
$8
cdate1
Did the participant consent
on this date?
CCONSENT1
1–Yes
2–No
$1
CCONSENT1
Y-Yes
N-No
U-Unknown
$1
$YNU
cconsent1
What visit was this?
CCONSENTVISIT
1–Pre-test
2–Post-test
3–Other follow-up
11.
CCONSENTVISIT1
01–Pre-test
02–Post-test
03–Other follow-up
$2
$CONSENT
cconsentvisit1
Date of second approach, if
applicable
C2DATE
Date
yyyymmdd
$10
CDATE2
mm/dd/yyyy
$8
$8
cdate2
Did the participant consent
on this date?
CCONSENT2
1–Yes
2–No
$1
CCONSENT2
Y-Yes
N-No
U-Unknown
$1
$YNU
cconsent2
What visit was this?
CCONSENTVISIT2
2–Post-test
3–Other follow-up
11.
CCONSENTVISIT2
01–Pre-test
02–Post-test
03–Other follow-up
$2
$CONSENT
cconsentvisit2
1
For SAS formats, see eHARS Technical Reference Guide, Chapter 17
June 2012
59
Table A.3: STARHS Specimen and Result Information in HIS and eHARS Datasets
HIS Dataset Before Conversion
eHARS Document Dataset After Conversion
IVR Screen
Label (in order
of appearance
in eHARS)
eHARS
Variable Name1 Valid Values
Screen Label
on Lab
Document
SAS Format Variable Name
Valid Values
Test Lab ID
Lab Name
LABID1–3
21D0649758
33D0654341
50D0661430
CLIA Code3
$11
CLIA_UID
CLIA laboratory codes3 $11
on CLIA_CODE table
Current: 33D2005937
$11
CLIA_UID
Source Lab
Specimen ID
Sample ID
LSRCEID1–3
Text
$20
SAMPLE_ID
Text
$50
$50
SAMPLE_ID
State Lab
Specimen ID
Accession
Number
SSTATEID1–3
Text
$20
ACCESSION_NUMBER
Text
$50
$50
ACCESSION_NO
Date of
Specimen
Collection
Collection Date LDTEOBT1–3
Date
yyyymmdd
$8
SAMPLE_DT
mm/dd/yyyy
$8
yyyymmdd
$8
SAMPLE_DT
Test Date
Result Date
Date
yyyymmdd
$8
RESULT_RPT_DT
mm/dd/yyyy
$8
yyyymmdd
$8
RESULT_RPT_DT
Not in IVR
Received Date
RECEIVE_DT
mm/dd/yyyy
$8
yyyymmdd
$8
RECEIVE_DT
Assay
Test
LOINC_CD
LOINC codes for lab
tests on LOINC_CODE
table
$7
$LNC_CD
LOINC_CD
SPECIMEN
BLD–Blood
OTH–Other
SAL– Saliva
URN–Urine
UNK–Unknown
$3
$SPC_TYP
SPECIMEN
$15
STARHS_SAMPLE_ID
Text
$15
$15
STARHS_SAMPLE_ID
$10
RESULT
#.###
$10
$10
RESULT
TESTDATE1–3
ASSAY1–3
BED=BED
BVLS=BVLS
(Vironostika LS)
OTLS=OTLS
(Vironostika LS)
OTV=OTV
(Vironostika LS)
AVID=AVID
Specimen Type Sample Type
LSPECTY1–3
1-Blood Finger Stick $15
2-Blood
Venipuncture
3-Blood Spot
4-Oral Mucosal
Transudate
5-Urine
8-Other
9-Unknown
SStarhsID
STARHS ID
(Specimen)
SSTARHSID1–3
Test SOD
Standard
SOD1–3
Optical Density
June 2012
$7
Length
SAS Format2 SAS Label
60
HIS Dataset Before Conversion
1
eHARS Document Dataset After Conversion
eHARS
Variable Name Valid Values
Screen Label
on Lab
Document
SAS Format Variable Name
Valid Values
Test Result
Final Result
01–Long term
02–Recent
91–QNS
92–Not rec'd by
STARHS Lab
93–Broken
94–Other
$2
RESULT_INTERPRETATION
01–Long term
$100
02–Recent
91–Quantity not
sufficient
92–Specimen never
received
93–Broken in transit
94–Other, indeterminate
95–Not sufficient
antibodies
99–Undefined result
$ST_RSLT
RESULT_INTERPRETATION
Reason
Specimen Not
Tested
Reason
SREASON1–3
Specimen Not
Sent for
STARHS
1–QNS
2–Specimen never
rec'd at public lab
3–Broken in transit
4–Other
$1
SREASON
1–Quantity not sufficient $1
2–Specimen never
received at public lab
3–Specimen broken in
transit
4–Other
5–Not sufficient
antibodies
$ST_RSN
SREASON
RESULT1–3
Source Lab CLIA Not in eHARS
code
LSRCEL1–3
Laboratory CLIA
code3
$20
Not in eHARS
State Lab ID
Not in eHARS
SSTATELID1–3
State Laboratory
CLIA code3
$20
Not in eHARS
HIV Test
Not in eHARS
LTSTTYP
5220-9=EIA/Elisa
$15
21009-6=Western
Blot
5472-6=CD4
25835-0=Viral Load
(NASBA)
5017-9=Viral Load
(bDNA)
25836-8=Viral Load
(RT-PCR)
Not in eHARS
Approved for
STARHS
Not in eHARS
LAPPRVE1–3
0=no
1=yes
2=pending
Not in eHARS
$1
Length
SAS Format2 SAS Label
IVR Screen
Label (in order
of appearance
in eHARS)
1
HIS datasets include specimen information and STARHS results for up to 3 specimens; variables are numbered for each corresponding sample.
2
For SAS formats, see eHARS Technical Reference Guide, Chapter 17
3
To look up specific CLIA codes, refer to: http://www.hipaaspace.com/Medical_Billing/Coding/Clinical_Laboratory_Improvement_Amendments/CLIA_Codes_Lookup.aspx
June 2012
61
Appendix B
List of ARV Medications for TTH
Medications appear alphabetically in eHARS and are stored in the database using the numeric codes
listed. More than one medication can be selected by holding down the Control key. The variable
UHRTA1 in the document-based datasets consolidates all the medications selected in a document in a
piped fashion. For example, if the person reported taking Ziagen, lamivudine, and zidovudine, the
values would appear as ‘20|03|26’.
Name of Medication
Code
Agenerase (amprenavir)
22
Aptivus (tipranavir, TPV)
30
Atripla (efavirenz / emtricitabine / tenofovir DF)
32
Combivir (lamivudine/ zidovudine)
24
Crixivan (indinavir sulfate)
06
Emtriva (emtricitabine, FTC)
11
Epivir (lamivudine, 3TC)
03
Epzicom (3TC / ABC)
28
Fortovase (saquinavir)
25
Fuzeon (enfuvirtide, T-20)
10
Hepsera (adefovir)
19
Hivid (zalcitabine, ddC)
02
Hydroxyurea
23
Invirase (saquinavir mesylate)
18
Intelence (etravirine)
34
Isentress (raltegravir)
36
Kaletra (lopinavir / ritonavir)
16
Lexiva (fosamprenavir, 908)
31
Norvir (ritonavir)
07
Other
88
Prezista (darunavir, DRV)
33
Rescriptor (delavirdine mesylate)
09
Retrovir (zidovudine, ZDV, AZT)
26
January 2011
62
Name of Medication
Code
Reyataz (atazanavir sulfate)
15
Saquinavir (Fortavase, Invirase)
08
Selzentry (maraviroc)
35
Sustiva (efavirenz)
21
Trizivir (abacavir sulfate / lamivudine / zidovudine)
13
Truvada (FTC / TDF)
27
Unspecified
99
Videx (didanosine, ddl)
01
Videx EC (didanosine, ddl)
14
Viracept (nelfinavir mesylate)
17
Viramune (nevirapine)
05
Viread (tenofovir)
12
Zerit (stavudine, d4T)
04
Ziagen (abacavir sulfate)
20
June 2012
63
Appendix C
Data Entry Recommendations for HIV Incidence Surveillance Data Elements in eHARS
Overall Guidance
Data entry in eHARS is document-based. Document-based data entry means that data entered or
imported on an individual document in eHARS were collected at the same time and from the same
source. Documents in eHARS that relate to HIV Incidence Surveillance (HIS) include the Testing and
Treatment History (TTH) Document and the Laboratory Document (where specimen information and
STARHS results are entered).
On the TTH and Laboratory documents in eHARS, there are a number of information tabs which
contain many data fields. These are the tabs (in italics) and their descriptions for TTH and Laboratory
documents:
Testing and Treatment History document:
1. Form Info – includes Stateno, reporting city, source of document, surveillance method, etc.
2. Identification – includes name, address, identification numbers
3. Demographics – includes DOB, sex, marital status, education, ethnicity, race, country of birth,
etc.
4. Testing and Treatment Information – includes all required and optional TTH data elements
5. Local Fields – fields that can be added by eHARS Administrator, for local use, not sent to CDC
6. Comments – text field for extra information for local use, not sent to CDC
Laboratory document:
1. Form Info – includes Stateno, reporting city, source of document, surveillance method, etc.
2. Identification – includes name, address, identification numbers
3. Demographics – includes DOB, sex, marital status, ethnicity, race
4. History – includes risk factors for HIV
5. Lab Data – STARHS lab name, sample ID, collection date, test type, STARHS ID, STARHS
results, etc.
6. Local Fields – fields that can be added by eHARS Administrator, for local use, not sent to CDC
7. Comments – text field for extra information for local use, not sent to CDC
Data Entry in eHARS
When entering or importing data into a document in eHARS, only data contained on a given data
collection form should be entered in that document. For example, if demographic information, such as
date of birth, race or ethnicity, is included on a TTH data collection form, these demographic variables
should be entered on the Demographics tab of the TTH document. Similarly, if name appears on the
data collection form, it should be entered on the Identification tab. However, if the TTH data
collection form does not contain any identification or demographics fields or if the fields are blank,
then do not enter anything on the Identification or Demographics tabs and do not seek this
information from another document in eHARS. Furthermore, if these data are available from another
source, they should be entered on a separate eHARS document. Note: demographic and
identification information entered on eHARS TTH and Lab documents can affect Person View
calculations, for example race, ethnicity, and address, depending on document hierarchies.
Many surveillance areas use an ACRF form that contains a section for HIV Testing and Antiretroviral
Use History. In this case, data for the ACRF should be entered on an ACRF document and the
supplemental TTH information on the ACRF should be entered on a separate TTH document in
June 2012
64
eHARS. It is not necessary to re-enter the identification or demographic information on the
Identification and Demographics tabs on the eHARS TTH document as these data will have been
entered on the eHARS ACRF document.
There is one tab that is very important for all eHARS documents, the Form Info tab. All documents in
eHARS include a minimal number of required/recommended data elements about the document
which need to be completed on the Form Info tab, as recommended below for incidence-related
documents. These are used for evaluation and data quality purposes.
Required, Recommended and Optional Data Elements
The following clarifies the required, recommended, and optional data elements that appear on
incidence-related eHARS documents. In this document, ’Required’ data elements are those that are
important for incidence estimation or that are necessary to enter an incidence document in eHARS.
‘Recommended’ data elements are very important for evaluation or matching purposes, but may not
always be available. ‘Optional’ variables are used for local surveillance and evaluation needs or for
enhancing case information. Tabs or fields found in eHARS that are not listed below are optional and
can be left blank.
TESTING AND TREATMENT HISTORY DOCUMENT
Form Info tab
Stateno
Reporting City
Date Received at Health Dept.
Source
Surveillance Method
Report Medium
Date Form Completed
Person Completing Form
Facility completing form
(Required)
(Optional)
(Optional)
(Recommended) – location where data were obtained
(Recommended) – used for evaluation
(Optional)
(Optional) – recommended if this appears on the form
(Optional)
(Optional)
Identification tab [Enter only if this information is on the data collection form]
Name type
First Name
Middle Name
(Recommended for matching purposes)
(Recommended for matching purposes)Last Name
(Recommended for matching purposes)
Address information
ID
Demographics tab [Enter only if this information is on the data collection form]
Sex at Birth
Date of Birth
Ethnicity
Race
Country of Birth
June 2012
(Recommended for matching purposes)
65
Testing and Treatment Information tab
Required Data Elements (leave blank if no attempt was made to obtain information):
1. Main source of testing and treatment history information
2. Date patient reported information
3. Ever had a previous positive HIV test?
4. Date of first positive HIV test
5. Ever had a negative HIV test?
6. Date of last negative HIV test
7. Number of negative HIV tests within 24 months before first positive test (Do not include first
positive HIV test)
8. Ever taken any antiretroviral medications (ARVs)?
9. If yes, name(s) of ARV medication taken
10. Dates ARVs taken: Date first began
11. Date of last ARV use
LAB DOCUMENT – STARHS RESULTS
Form Info tab
Stateno
Date Received at Health Department
Source
Did this report initiate a new investigation?
Reporting Medium
Surveillance Method
(Required)
(Optional)
(Recommended) – should be ‘A05’
(Optional) – should be ‘no’
(Optional)
(Recommended) – for evaluation, should be ‘Follow-up’
Identification tab [Enter only if this information is on the data collection form]
Name type
First Name
Middle Name
Last Name
(Recommended for matching purposes)
(Recommended for matching purposes)
(Recommended for matching purposes)
Lab Data tab
Lab Name
Sample ID (Specimen)
Accession Number
Collection Date
Result Date
Received Date
Test (test type)
Sample Type
STARHS ID (Specimen)
Final Result
Standard Optical Density
Reason Specimen Not Sent for STARHS
June 2012
(Recommended)
(Recommended)
(Recommended)
(Required for STARHS specimen document)
(Recommended for valid BED results)
(N/A)
(Required) – necessary to display fields for BED test
information
(Recommended)
(Required)
(Required if available)
(Required with valid STARHS Final Result)
(Optional)
66
Local HIV Incidence Estimation Guide
Version:
Date:
Prepared by:
2.2
January 24, 2011
Rebecca Ziebell
Qian An
Xing Dong
Version Tracking
Version
2.0
Date Finalized
October 21, 2010
2.1
2.2
January 7, 2011
January 24, 2011
Comments
Initial version of user guide for 2nd round of local HIV
incidence estimation
Updates to subsection 2.2 and sections 3 and 4
Updates to data dictionary table in subsection 3.1
2
Table of Contents
1. Background ............................................................................................................................................... 4
2. Required Materials.................................................................................................................................... 4
2.1. Locally Available Materials................................................................................................................. 4
2.2 CDC-Provided Materials ...................................................................................................................... 4
3. HIV Incidence Estimation .......................................................................................................................... 5
3.1. PGM1_PREPARE_DATA ...................................................................................................................... 6
Table 1. Assess Missingness of STARHS Results, Testing Group, and T (Among Observed Repeat
Testers)................................................................................................................................................ 12
Table 2. Assess Distribution of Observed STARHS Result Data........................................................... 12
Table 3. Assess Distribution of Observed Testing Group Data ........................................................... 12
3.2. PGM2_IMPUTE_BED_GROUP_T ...................................................................................................... 13
Table 4. Incidence Data Summary Before and After Imputation........................................................ 14
Table 5. Determine Which Potential Strata Meet 20%/40/10 Requirement ..................................... 15
3.3. PGM3_DETERMINE_STRATA............................................................................................................ 17
Table 6. Determine Stratification Variable Based on BED-Recent Heterogeneity.............................. 19
3.4. PGM4_ESTIMATE_INCIDENCE ......................................................................................................... 20
4. Interpreting HIV Incidence Estimates ..................................................................................................... 24
4.1. PGM5_COMPARE_ESTIMATES ......................................................................................................... 24
4.2. PGM6_EVALUATE DIFFERENCES ...................................................................................................... 27
5. Contact Information ................................................................................................................................ 28
3
1. Background
The purpose of this document is to serve as a reference for HIV incidence surveillance staff during
preparation of local HIV incidence estimates. For more general information on the purpose of HIV
incidence estimation and the history of HIV incidence surveillance in the United States, please refer to
the resources available on the CDC website: http://www.cdc.gov/hiv/topics/surveillance/incidence.htm
The following acronyms and abbreviations will be used throughout this document:
Abbreviation
ARV
BED
eHARS
HICSB
HIS
STARHS
TTH
XX
Definition
Antiretroviral drug
BED HIV-1 EIA Capture Assay
Enhanced HIV/AIDS Reporting System
CDC’s HIV Incidence and Case Surveillance Branch
HIV Incidence Surveillance
Serologic Testing Algorithm for Recent HIV Seroconversion
Testing and Treatment History
Placeholder for abbreviation of name of local surveillance area
2. Required Materials
This section details the materials needed to successfully estimate HIV incidence at the state or local
level. Unless otherwise noted, the reader can assume implied file extensions of *.sas for SAS programs,
*.sas7bdat for data sets, and *.pdf for output tables. Throughout this document, the names of SAS
programs will appear in ALL CAPS, data set names in SMALL CAPS, variable names in italics, and PDF
filenames in bold.
2.1. Locally Available Materials
The local HIV incidence estimation process requires that HIS staff have access to several SAS data sets
exported from the local eHARS system. The latest available frozen, year-end PERSON, DOCUMENT and LAB
data sets are specifically required as inputs to the local HIV incidence estimation process. The process
also uses the PERSON_MI imputed risk factor data set1, which should be based on the same frozen, yearend snapshot of the eHARS system. The final required local data set is the monthly HIS data set (e.g.,
2
XX_HIS_YYYYMMDD ), which need not be based on the same frozen, year-end eHARS data sets; instead,
the most recent monthly HIS data set can be used for local HIV incidence estimation.
2.2 CDC-Provided Materials
CDC has developed a set of six SAS programs to facilitate the estimation of HIV incidence at the local
level. These programs should be opened and executed in SAS 9.1.3 or SAS 9.2., not in SAS Enterprise
Guide. The programs were designed to be run with a minimum of user interaction. Sections that do
require user input appear immediately beneath program headers and are clearly marked as follows:
1
PERSON_MI refers to the output of the MI SAS PROGRAMS 2009 V1 program, which was provided to core
surveillance staff by CDC’s Quantitative Sciences and Data Management Branch on February 4, 2010.
2
XX refers to the two- or three-character abbreviation of the project area name; see section 3.1 for more details.
YYYYMMDD represents the date of data set creation. The monthly HIS data set must be in the revised PERSON-based
format implemented in late summer 2010.
4
*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*;
*
ATTENTION: USER INTERACTION REQUIRED!
*;
*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*;
Instructions for interacting with these sections will be provided both in this document and within the
programs themselves.
The SAS programs require as input the locally available data sets mentioned in section 2.1 as well as two
additional SAS data sets provided by CDC, REFP_NEW and REFP_REPEAT. These data sets contain the
reference probabilities of being detected as BED-recent for new and repeat testers, respectively.
All CDC-provided materials—i.e., six SAS programs and two SAS data sets—should be stored in a single
secure, user-defined folder on the local network. All output data sets and PDF files generated by the SAS
programs will be automatically saved to this location.
3. HIV Incidence Estimation
As previously mentioned, CDC has developed a set of SAS programs for use in estimating HIV incidence
at the local level. This section covers the first four programs—PGM1_PREPARE_DATA,
PGM2_IMPUTE_BED_GROUP_T, PGM3_DETERMINE_STRATA, and PGM4_ESTIMATE_INCIDENCE—which
should be run in succession to generate single-year HIV incidence estimates for each desired year. The
fifth and sixth programs, PGM5_COMPARE_ESTIMATES and PGM6_EVALUATE_DIFFERENCES, and their
associated output will be discussed in section 4 of this document.
The order and timing of program execution are important, as programs 2–5 refer to the initial setup
completed in PGM1_PREPARE_DATA and described in section 3.1 below. Therefore, the user should run
the first five programs within the same instance of SAS—i.e., without closing and restarting the
application. Program 6 does not rely on the initial setup and can be run separately at any time.
The four programs used to estimate HIV incidence are:
PGM1_PREPARE_DATA
o Merges and cleans HIV incidence-related data elements
o Creates initial HIV incidence estimation data set
o Generates pre-imputation data review tables
PGM2_IMPUTE_BED_GROUP_T
o Imputes missing values of STARHS results, testing group assignments (i.e., new vs.
repeat tester status), and time from last negative to first positive HIV test (T)
o Creates imputed HIV incidence estimation data set
o Generates post-imputation data review tables
PGM3_DETERMINE_STRATA
o Accepts user input re: regrouping of potential estimation stratification variables
o Creates final HIV incidence estimation data set
o Generates BED-recent heterogeneity data review table
PGM4_ESTIMATE_INCIDENCE
o Estimates HIV incidence, both overall and stratified (if applicable)
5
o
Outputs HIV incidence estimation results tables
3.1. PGM1_PREPARE_DATA
PGM1_PREPARE_DATA begins with an interactive section in which the user must enter the desired year
or range of years for which to estimate HIV incidence, the local area abbreviation, and the names and
locations of the input data sets described in sections 2.1 and 2.2 above. This introductory section is
clearly noted in the program and uses the “%let” style of setting macro variables (which should be
familiar to individuals who have used other SAS programs provided by HICSB’s Incidence and Viral
Resistance Team):
*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*;
*
ATTENTION: USER INTERACTION REQUIRED!
*;
*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*;
* Set values for macro variables below. End each line with a semicolon.
*;
*
Years to be analyzed/estimated (separate range with hyphen, no spaces)
*;
*
ex. %let estyears = 2006-2008
*;
*
Project area abbreviation
*;
*
ex. %let area = XX
*;
*
Location and name of latest monthly HIS data set
*;
*
ex. %let hislib = H:\Secure\HIS
*;
*
%let hisdata = xx_his_20100719
*;
*
Location and name of frozen year-end eHARS PERSON data set
*;
*
ex. %let perlib = H:\Secure\eHARS\Frozen\2009
*;
*
%let perdata = person
*;
*
Location and name of frozen year-end eHARS DOCUMENT data set
*;
*
ex. %let doclib = H:\Secure\eHARS\Frozen\2009
*;
*
%let docdata = document
*;
*
Location and name of frozen year-end eHARS LAB data set
*;
*
ex. %let lablib = H:\Secure\eHARS\Frozen\2009
*;
*
%let labdata = lab
*;
*
Location and name of frozen year-end imputed risk factor data set
*;
*
ex. %let irflib = H:\Secure\eHARS\Frozen\2009
*;
*
%let irfdata = person_mi
*;
*
Location of estimation-related files (incl. REFP_NEW and REFP_REPEAT)
*;
*
ex. %let estlib = H:\Secure\HIS\2006-2008 Estimation
*;
*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*;
%let
%let
%let
%let
%let
%let
%let
%let
%let
%let
%let
%let
%let
estyears = ;
area = ;
hislib = ;
hisdata = ;
perlib = ;
perdata = ;
doclib = ;
docdata = ;
lablib = ;
labdata = ;
irflib = ;
irfdata = ;
estlib = ;
*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*;
Local area abbreviations (XX) should be entered as follows:
HIS funded at state level: Two-letter state abbreviation (e.g., Alabama = AL, New York = NY)
HIS funded at city or county level:
6
o
o
Multi-word name: First letter of each word (Los Angeles County = LAC, New York City =
NYC, San Francisco = SF)
Single-word name: First three letters of name (Chicago = CHI, Houston = HOU,
Philadelphia = PHI)
After entering the macro variable values, the user can click the Submit button to execute
PGM1_PREPARE_DATA. The program includes a check to ensure that the baseline data set contains
sufficient data to estimate HIV incidence for the desired year(s). If local data do not meet the minimum
requirements—i.e., for each year, 300 diagnoses of HIV disease and 15% STARHS result completeness—
then the program will stop executing and display an alert window. The window alerts the user to the
specific year or years that lack sufficient data. At this point the user can return to the interactive portion
at the beginning of the program and enter a different year or years for which to estimate HIV incidence,
then resubmit PGM1_PREPARE_DATA.
Successful execution of PGM1_PREPARE_DATA accomplishes the following:
Identification of the baseline sample population—i.e., adult and adolescent diagnoses of HIV
disease in the desired year(s), residing in the selected HIS area3 at time of diagnosis—from the
frozen, year-end eHARS PERSON data set
Collection of available documented dates of last negative HIV tests from frozen, year-end eHARS
DOCUMENT and LAB data sets
Addition of HIS-specific data elements from latest monthly HIS data set, XX_HIS_YYYYMMDD
Selection of “best” TTH data elements for each case:
o Earliest date of first positive HIV test
o Latest date of last negative HIV test
o Earliest ARV start date
o Latest ARV end date
o Most complete set of Ever Tested Negative, Number of Negative HIV Tests in 24 Months
Before First Positive, and Facility of First Positive Test data elements
Elimination of non-useable STARHS results (i.e., from non-BED assays, specimens collected
outside the three-month post-diagnosis timeframe, or persons who used ARV during six months
prior to specimen collection)
Resolution of apparent conflicts between TTH data elements
Creation of various calculated variables for use in imputation and estimation programs
Creation of the initial HIV incidence estimation data set, XX_HIS_EST_DATA_YEAR(S)4
Generation of pre-imputation data review tables
3
PGM1_PREPARE_DATA is designed to prepare an initial HIV incidence estimation data set at the level of the
funded project area. In some cases—e.g., when the funded area is “State, excluding Major City”—the desired area
for estimation may be something other than the area that was funded. In these cases, two or more HIS areas (e.g.,
State and Major City) will need to collaborate to create a statewide, deduplicated version of the monthly HIS data
set to use as input for PGM1_PREPARE_DATA. Local HIS staff should contact their assigned HICSB epidemiologist
for assistance.
4
Single-year filename: XX_HIS_EST_DATA_YEAR; multi-year filename: XX_HIS_EST_DATA_STARTYEAR_ENDYEAR
7
The table below contains a list of pertinent calculated XX_HIS_EST_DATA_YEAR(S) variables in the order of
their creation in PGM1_PREPARE_DATA. All other variables in XX_HIS_EST_DATA_YEAR(S) are sourced
directly from the eHARS PERSON data set or the monthly HIS data set.
Variable Name
hars_fpos_yrmo
Variable Label
Date of HIV disease diagnosis
hars_fpos_yr
Year of HIV disease diagnosis
hiv2aid
Months from HIV diagnosis to
AIDS diagnosis
AIDS diagnosis at HIV
diagnosis
aids
aids6
AIDS diagnosis 1–6 months
after HIV diagnosis
res_st
State of residence at HIV
disease diagnosis
res_county
County of residence at HIV
disease diagnosis
hars_geoco
County FIPS of residence at
HIV disease diagnosis
res_city
City of residence at HIV
disease diagnosis
hars_geocty
City FIPS of residence at HIV
disease diagnosis
8
Calculation Method & Valid Values
Numeric YYYYMM version of variable
hiv_aids_dx_dt from eHARS PERSON data set
Numeric YYYY version of variable
hiv_aids_dx_dt from eHARS PERSON data set
Calculated from eHARS PERSON data set
variables hiv_dx_dt and aids_dx_dt
If calculated variable hiv2aid is less than or
equal to 0, then aids = “1”; otherwise, aids =
“0”.
If calculated variable hiv2aid is between 1 and
6 (inclusive), then aids6 = “1”; otherwise,
aids6 = “0”.
If state of residence at HIV Dx (rsh_state_cd) is
not missing, then state of residence at HIV
disease Dx equals rsh_state_cd; otherwise,
state of residence at HIV disease Dx equals
state of residence at AIDS Dx (rsa_state_cd).
If state of residence at HIV Dx (rsh_state_cd) is
not missing, then county of residence at HIV
disease Dx equals county of residence at HIV
Dx (rsh_county_name); otherwise, county of
residence at HIV disease Dx equals county of
residence at AIDS Dx (rsa_county_name).
If state of residence at HIV Dx (rsh_state_cd) is
not missing, then county FIPS of residence at
HIV disease Dx equals county FIPS of residence
at HIV Dx (rsh_county_fips); otherwise, county
FIPS of residence at HIV disease Dx equals
county FIPS of residence at AIDS Dx
(rsa_county_fips).
If state of residence at HIV Dx (rsh_state_cd) is
not missing, then city of residence at HIV
disease Dx equals city of residence at HIV Dx
(rsh_city_name); otherwise, city of residence
at HIV disease Dx equals city of residence at
AIDS Dx (rsa_city_name).
If state of residence at HIV Dx (rsh_state_cd) is
not missing, then city FIPS of residence at HIV
disease Dx equals city of residence at HIV Dx
(rsh_city_fips); otherwise, city FIPS of
residence at HIV disease Dx equals city FIPS of
residence at AIDS Dx (rsa_city_fips).
Variable Name
hars_fps
Variable Label
Type of facility of HIV disease
diagnosis
hars_lstn_yrmo
Date of last negative test
lstnd
TTH date of last negative test
(char, yyyymmdd)
tth_lstn_yrmo
TTH date of last negative test
(num, yyyymm)
lstnd_tth_form
TTH form (pre/post) used for
date of last negative test
fposd
TTH date of first positive test
(char, yyyymmdd)
tth_fpos_yrmo
TTH date of first positive test
(num, yyyymm)
fposd_tth_form
TTH form (pre/post) used for
date of first positive test
hrtbd
TTH date of start of ARV use
(char, yyyymmdd)
tth_arv_start_yrmo TTH date of start of ARV use
(num, yyyymm)
hrtbd_tth_form
TTH form (pre/post) used for
date of start of ARV use
hrted
TTH date of end of ARV use
(char, yyyymmdd)
tth_arv_end_yrmo TTH date of end of ARV use
(num, yyyymm)
hrted_tth_form
TTH form (pre/post) used for
date of end of ARV use
other_tth_form
TTH form (pre/post) used for
key non-date variables
ngtst
TTH ever tested negative
9
Calculation Method & Valid Values
If type of facility of HIV Dx
(hf_facility_type_cd) is not missing, then type
of facility of HIV disease Dx equals
hf_facility_type_cd; otherwise, type of facility
of HIV disease Dx equals type of facility of
AIDS Dx (af_facility_type_cd). Valid values:
eHARS facility type codes.
Numeric YYYYMM version of latest available
negative screening test result from eHARS LAB
table
Later of pre- and post-test TTH variables
ulstnd and klstnd
Calculated from lstnd
Valid values: “PRE”, “POST”
Earlier of pre- and post-test TTH variables
ufposd and kfposd
Calculated from fposd
Valid values: “PRE”, “POST”
Earlier of pre- and post-test TTH variables
uhrtbd and khrtbd
Calculated from hrtbd
Valid values: “PRE”, “POST”
Later of pre- and post-test TTH variables
uhrted and khrted
Calculated from hrted
Valid values: “PRE”, “POST”
Pre- or post-test form chosen as source of key
non-date variables based on number of
useable values on each form. Valid values:
“PRE”, “POST”.
If calculated variable other_tth_form equals
“PRE”, then ngtst equals pre-test TTH variable
ungtst; otherwise, ngtst equals post-test TTH
variable kngtst. Valid values: “0” (No), “1”
(Yes), “7” (Refused), “9” (Unknown).
Variable Name
numtsts
Variable Label
TTH number of tests in two
years before first positive test,
no previous positive (char)
pnumtsts
TTH number of tests in two
years before first/previous
positive test (char)
num_tests
TTH number of tests in two
years before first positive test
(num)
fps
TTH type of facility of HIV
diagnosis
spec_date
raw_result
result
Date of STARHS specimen
collection (SAS date)
Raw STARHS result
STARHS result
newdx
Presumed new diagnosis
cal_fpos_yrmo
Calculated date of first
positive test (num, yyyymm)
Calculated date of last
negative test (num, yyyymm)
STARHS result group
cal_lstn_yrmo
bed
10
Calculation Method & Valid Values
If calculated variable other_tth_form equals
“PRE”, then numtsts equals pre-test TTH
variable unumtsts; otherwise, numtsts equals
post-test TTH variable knumtsts.
If calculated variable other_tth_form equals
“PRE”, then pnumtsts equals pre-test TTH
variable upnumtsts; otherwise, pnumtsts is
blank.
If calculated variable other_tth_form equals
“PRE”, then num_tests equals the greater of
pre-test TTH variables unumtsts and
upnumtsts; otherwise, num_tests equals
knumtsts. Note: 77 = Refused, 99 = Unknown.
If calculated variable other_tth_form equals
“PRE”, then fps equals pre-test TTH variable
ufpstyp; otherwise, fps equals post-test TTH
variable kfps. Valid values: eHARS facility type
codes.
Calculated from specimen variable ldteobt
Unmodified/original STARHS result
Modified STARHS result, set to missing when:
value of calculated variable raw_result is
invalid (i.e., not long-term or recent);
specimen was collected >3 months after HIV
disease Dx date; BED assay was not used; or
there is evidence of ARV use during 6 months
prior to specimen collection.
If calculated variable tth_fpos_yrmo is more
than 6 months earlier than calculated variable
hars_fpos_yrmo, then newdx equals “N”;
otherwise, newdx equals “Y”.
Earlier of calculated variables hars_fpos_yrmo
and tth_fpos_yrmo
Later of calculated variables hars_lstn_yrmo
and tth_lstn_yrmo
If concurrent HIV and AIDS diagnoses (aids =
“1”), then bed = “a”; else if result is missing,
then bed = “x”; else if STARHS result is longterm (result = “01”), then bed = “b”; else if
STARHS result is recent (result = “02”), then
bed = “0”.
Variable Name
group
Variable Label
Testing group
t
Time (in months) from
calculated last negative test to
calculated first positive test
Calculated type of facility of
HIV diagnosis (raw)
cal_fps
cat_fps
Calculated type of facility of
HIV diagnosis (categorical)
Calculation Method & Valid Values
If calculated date of last negative test
(cal_lstn_yrmo) is not missing and not equal to
HIV disease Dx date (hars_fpos_yrmo) OR ever
tested negative equals yes (ngtst = “1”) OR
number of tests in two years before first
positive (num_tests) is greater than 1 and less
than 77, then group = “1” (repeat testers); else
if ever tested negative equals no (ngtst = “0”),
then group = “0” (new testers); otherwise,
group = “x” (missing).
Difference between calculated variables
cal_lstn_yrmo and cal_fpos_yrmo
If calculated date of first positive
(cal_fpos_yrmo) equals HIV disease Dx date
(hars_fpos_yrmo) and type of facility of HIV
disease Dx (hars_fps) is not missing, then
cal_fps = hars_fps; otherwise, cal_fps = TTH
type of facility of HIV diagnosis (fps).
Values: “1-CTS/STD”, “2-Inpatient”, “3Outpatient”, “4-Other”, “5-Unknown”
PGM1_PREPARE_DATA ends by generating three tables (described below), which are automatically
output to the XX Year(s) Pre-Imputation Data Review PDF file in the “estlib” directory. These tables
provide an overview of local data quality related to the imputation of missing data that occurs in the
next SAS program, PGM2_IMPUTE_BED_GROUP_T. The tables are purely informational and require no
direct action on the part of the user; they are simply intended to help the user understand the
imputation process. Since imputation and estimation are performed separately for each year, Tables 1–3
are similarly broken out by year.
Tables 1–3 are designed to address different aspects of one of the basic assumptions of the imputation
process—namely that missing data are missing-at-random (MAR). MAR means that missingness only
depends on observed data points, not on the value of the unobserved data. The assumption of MAR
data is impossible to prove, but most imputation models address the issue by incorporating as many
potential predictors of missingness as possible so that any remaining, unexplained missing data are likely
to be MAR. In accordance with this approach, the imputation model used for local HIV incidence
estimation incorporates a variety of covariates that may be associated with the missingness and/or the
distribution of non-missing data in the three variables being imputed—i.e., STARHS results, testing
group assignments, and T. Table 1 addresses potential covariates associated with missingness of the
three variables, while Tables 2 and 3 address potential covariates associated with non-missing values
(which are used by the imputation model to replace missing values).
11
Table 1. Assess Missingness of STARHS Results, Testing Group, and T (Among Observed Repeat
Testers)
For each year, Table 1 displays the missingness of data to be imputed by a list of covariates: sex,
race/ethnicity, age at diagnosis, raw (i.e., not imputed) transmission category, and type of facility of
diagnosis. When reviewing this table, the user should pay attention to whether the percentage of cases
with missing STARHS results or testing group assignments is similar across each level of a covariate (e.g.,
for both sexes or for all age groups). If the missing percentage differs across levels, then missingness is
said to be associated with that covariate. The covariate should then be considered during imputation
either by including it in the imputation model or performing a stratified imputation by the covariate.
The imputation model used in local HIV incidence estimation incorporates all of the covariates displayed
in Table 1, even those that may not be associated with missingness, since including them has no
negative impact on the model. Therefore, Table 1 requires no direct action on the part of the user; it is
simply meant to provide insight into local data.
Table 1 should indicate low missingness of T among observed repeat testers; the imputation of T is
mainly reserved for imputed repeat testers.
Some HIS areas may be aware of other covariates (e.g., county of residence) that are locally associated
with missingness and should be included in the imputation model. Areas should contact their assigned
HICSB epidemiologists to discuss modifying the SAS programs to include these covariates.
Table 2. Assess Distribution of Observed STARHS Result Data
Table 2 is designed to highlight whether non-missing STARHS results are associated with any particular
covariates. Table 2 output is broken out by year and can be reviewed separately by year as well. The
covariates included in Table 2 are the same as those in Table 1: sex, race/ethnicity, age at diagnosis, raw
transmission category, and type of facility of diagnosis. If the percentage of BED-recent or long-term
results differs across levels of a covariate, this indicates that different types of cases have different
chances of being detected as BED-recent or long-term and that the covariate in question should be
included in the imputation of STARHS results. Again, the imputation model used in local HIV incidence
estimation already incorporates all of the potential covariates listed in Table 2, since their inclusion has
no negative impact on the model. Therefore, Table 2 requires no direct action on the part of the user; it
is purely informational.
Table 3. Assess Distribution of Observed Testing Group Data
Table 3 allows the user to assess whether non-missing testing group assignments and T values are
associated with any covariates, which include the aforementioned demographic categories as well as
observed STARHS result. Similar to Tables 1 and 2, Table 3 output is separated by year.
When reviewing Table 3, the first area of focus should be the percentage of new testers and whether it
differs across levels of the potential covariates. As a reminder, new testers are those who indicated that
they never had a negative test result prior to their first positive. If the percentage of new testers differs
across levels of one or more covariates, this indicates that the likelihood of being a new tester differs
with the covariate(s) in question, which should then be included in the imputation of missing testing
12
group values. Again, the local HIV incidence estimation process automatically includes all of the
displayed covariates in the imputation of testing group, so no action is required on the part of the user.
The second area of focus in Table 3 is whether the median value of T among observed repeat testers is
associated with any covariates. In preparing the national HIV incidence estimates, CDC analysts noticed
that STARHS result and age at diagnosis are highly associated with the distribution of T. Therefore, these
two covariates are included in the imputation of T in the default version of the local HIV incidence
estimation programs. At the local level, Table 3 may show that other covariates are highly associated
with the distribution of T; however, these associations may be the result of small sample sizes or known
biases in data collection. HIS areas may choose to modify the imputation of T based on these results, but
CDC does not recommend this course of action. If HIS areas elect to modify the imputation of T, they
should be careful never to stratify by more than two covariates. Limiting to no more than two
stratification variables ensures that each stratum contains a sufficient number of cases with observed T
to prevent an overly sparse distribution from which to impute T values.
3.2. PGM2_IMPUTE_BED_GROUP_T
After reviewing the pre-imputation data review tables, the user should run
PGM2_IMPUTE_BED_GROUP_T, which can be submitted and executed without user input. Successful
execution of PGM2_IMPUTE_BED_GROUP_T accomplishes the following:
For each year:
o Replacement of missing transmission category values in the initial HIV incidence
estimation data set, XX_HIS_EST_DATA_YEAR(S), with 20 possible values based on the
imputed risk factor data set, PERSON_MI
o Definition of dummy variables for use as covariates in multiple imputation
o Multiple imputation of missing STARHS results and testing group assignments
o Imputation of missing T among repeat testers based on observed distributions of T by
age at diagnosis and STARHS result strata
Creation of XX_IMPUTED_YEAR(S) data set
Generation of post-imputation data review tables
Missing STARHS result and testing group values are imputed via multiple imputation using SAS’s MI
procedure. The number of imputations was set at 20 based on the percentage of missing data and the
desired precision. The following basic syntax is used:
proc mi ;
by ;
class ;
monotone (discrim );
var ;
run;
As previously discussed, the imputation model used in local HIV incidence estimation incorporates a
variety of covariates. The control variables used in the imputation of STARHS results are race/ethnicity,
age at diagnosis, imputed transmission category, type of facility of diagnosis, and observed testing
13
group. The covariates used in imputation of missing testing group are race/ethnicity, age at diagnosis,
imputed transmission category, type of facility of diagnosis, whether AIDS was diagnosed concurrently
with HIV, whether AIDS was diagnosed 1–6 months after HIV, whether STARHS result was observed or
imputed, and imputed STARHS result. Furthermore, since the number of possible transmission
categories differ by sex, both STARHS results and testing group assignments are imputed separately for
males and females.
Missing STARHS results are imputed prior to missing testing group assignment. Imputation of missing
STARHS results excludes both concurrent AIDS cases and those that were diagnosed with AIDS between
one and six months after HIV, as these cases are known to represent long-term HIV infections. STARHS
results from concurrent AIDS cases are not used in estimating HIV incidence, and cases that were
diagnosed with AIDS between one and six months after HIV are automatically assigned BED-long-term
results.
As mentioned in the discussion of Table 1 (see page 12), the user may wish to incorporate additional
covariates that are known to be associated with local missingness of STARHS results and/or testing
group assignments. These modifications should be made in the multiple imputation section of
PGM2_IMPUTE_BED_GROUP_T. However, the variables in PROC MI’s MONOTONE and VAR statements
must be listed a specific order. Therefore, the local user should consult with his or her assigned HICSB
epidemiologist before modifying the PROC MI syntax.
Missing values of T among observed and imputed repeat testers are imputed through multinomial
distributions within each stratum determined by age at diagnosis and STARHS result. Within each
stratum, the observed distribution of T among observed repeat testers is derived and used to impute
missing values of T among observed and imputed repeat testers.
Program 2 ends by generating Tables 4 and 5, which are automatically output to the XX Year(s) PostImputation Data Review PDF file in the “estlib” directory. These tables should be reviewed carefully, as
each may indicate the need for additional user interaction with later SAS programs.
Table 4. Incidence Data Summary Before and After Imputation
The purpose of Table 4 is to confirm the successful imputation of missing STARHS result, testing group,
and T values. If any errors or data anomalies occur during imputation, they will be highlighted in Table 4,
which displays the following hierarchical summary statistics for each year:
-
Total number of new diagnoses of HIV disease
- Among all diagnoses, numbers and percentages of cases with and without testing group
assignments
- Among cases with testing group assignments, numbers and percentages of new
and repeat testers
- Among repeat testers, numbers and percentages of cases with and
without T
- Among all diagnoses, numbers and percentages of cases with concurrent HIV and AIDS
diagnoses (“AIDS cases”), cases in which AIDS was diagnosed between one and six
14
months after HIV (“AIDS6 cases”), and cases that were not diagnosed with AIDS within
six months of HIV
- Among non-AIDS and non-AIDS6 cases, numbers and percentages of cases with
and without STARHS results
- Among cases with STARHS results, numbers and percentages of BEDrecent and long-term cases
PGM2_IMPUTE_BED_GROUP_T should impute values for all missing values of STARHS result, testing
group, and T. As a result of the imputation process, certain frequencies and percentages should look
different in the post-imputation data set; however, other data should remain unchanged. With that in
mind, the user should look for the following characteristics in each yearly iteration of Table 4:
1. The total number of new diagnoses should remain the same before and after imputation. The
imputation process should fill out missing data, not increase the number of cases in the data set.
2. After imputation, all new diagnoses should have a testing group assignment. The percentage of
new diagnoses with testing group assignments in the post-imputation column should be 100%.
3. Among cases with testing group assignments, the numbers of new and repeat testers should
increase due to imputation; however, the difference in percentages should not be dramatic.
Small changes are expected due to the fact that missing data are imputed multiple times
through independent draws from an imputation model; however, the fact that missing data are
imputed from observed data should preclude any dramatic changes in the percentages of new
and repeat testers.
4. Among repeat testers, all those missing T before imputation will have an imputed T after
imputation. In the post-imputation column, the percentage of repeat testers with T should be
100%.
5. AIDS and AIDS6 cases are excluded from STARHS result imputation. STARHS results from
concurrent HIV/AIDS cases are not used in the HIV incidence estimation model, and AIDS6 cases
are assigned BED-long-term results.
6. After imputation, all non-AIDS and non-AIDS6 cases should have STARHS results. Therefore, the
percentage of cases with STARHS results should be 100% in the post-imputation column.
7. Finally, among cases with STARHS results, Table 4 should show slight changes in the proportion
of BED-recent results, for the same reason described in #3 above.
If the output displayed in Table 4 does not match the description above, then the user should first check
the SAS log window for evidence of errors during program execution. If none are found, the user should
contact his or her assigned HICSB epidemiologist for assistance in identifying underlying data anomalies
or imputation errors.
Table 5. Determine Which Potential Strata Meet 20%/40/10 Requirement
The HIV incidence estimation model uses stratification to account for heterogeneity caused by the fact
that different groups of people have different testing behaviors. In the context of HIV incidence
estimation, a stratum refers to a group of cases with presumed similar testing behavior. The model
estimates incidence for each stratum individually; the total HIV incidence estimate is the sum of
15
estimates from all strata. To ensure stable estimates, each selected stratum should satisfy the
20%/40/10 rule or requirement5:
The number of non-AIDS cases with observed STARHS results must be greater than or equal to
40 and represent at least 20% of new diagnoses; and
The number of BED-recent results must be greater than or equal to 10.
Potential stratum
All New
Diagnoses
N
--
With STARHS Results
N
%
≥40
≥20
BED-Recent
N
≥10
Table 5 allows the user to assess, by year, whether a variety of potential strata meet the 20%/40/10 rule
and thereby indicates whether regrouping of strata may be necessary to generate stratified incidence
estimates. The potential stratification variables displayed in Table 5 include sex, race/ethnicity, age at
diagnosis, and imputed transmission category. (Note: Raw transmission category is also displayed but
only to show the impact of using the PERSON_MI imputed risk factor data set; imputed transmission
category should be used to assess the 20%/40/10 rule.)
Table 5 is designed to help the user prepare for any desired regrouping, which takes place at the
beginning of the next SAS program, PGM3_DETERMINE_STRATA. If the displayed numbers of new
diagnoses, observed STARHS results, and observed BED-recent results do not meet the 20%/40/10
requirement for a given potential stratum, the relevant numbers appear in red on the report, prompting
the user to regroup, or combine, certain strata to obtain the required case counts and/or percentages.
In order to compare strata-level HIV incidence estimates across years, the user must use the same
regrouping methodology for all years. Therefore, Table 5 output should be reviewed across all years,
and associated regrouping decisions should ensure that 20%/40/10 requirements are met for all years
whenever possible.
Consider, for example, a user who is estimating local HIV incidence for the years 2006 through 2008.
This user would like to generate HIV incidence estimates for injection drug users (IDU), as local
prevention efforts have been particularly focused on this group. When the user examines Table 5, he
notices that 2006 data are too sparse to allow IDU to meet the 20%/40/10 requirement as a separate
stratum; however, the IDU stratum does meet the 20%/40/10 requirement for 2007 and 2008. This
means that the user cannot generate an IDU-only HIV incidence estimate for 2006, but he could do so
for 2007 and 2008. At this point, the user has two options: (1) regroup the transmission category strata
to satisfy the 20%/40/10 requirement across 2006, 2007, and 2008—and lose the ability to generate an
IDU-only HIV incidence estimate for the latter years; or (2) regroup the transmission category
stratification variable to maintain a separate, IDU-only stratum, but note that the 2006 HIV incidence
estimate for IDU should be interpreted with extreme caution, recognizing the limitations of the 2006
data when strata are configured in this way. Either option is appropriate; the choice depends on the
needs of the local HIS area. (In either case, when regrouped in this way, the variable mode3 should not
5
Formerly known as the 200/40/10 rule or requirement.
16
be selected as the stratification variable for the 2006 HIV incidence estimate. Selection of the best
stratification variable will be discussed further in the description of Table 6 and in section 3.4 below.)
When making decisions about regrouping strata, the user should also incorporate knowledge about the
local HIV epidemic. Consider a scenario in which the “White,” “Black,” and “Hispanic” race/ethnicity
strata meet the 20%/40/10 requirement across all years, but the “Other” race/ethnicity stratum does
not meet the requirement for any year being analyzed. The user must then regroup strata in order to
combine “Other” with one of the remaining race/ethnicity groups. At this point, the user should rely on
what he or she knows about the local HIV epidemic to guide the regrouping decision. For example, if the
local epidemic among persons who are not White, Black or Hispanic is known to be most similar to that
among Whites, then the user may decide to regroup race/ethnicity strata as “Black,” “Hispanic,” and
“White/Other.”
Note: If a particular stratum does not meet the 20%/40/10 requirement but is of particular interest to
the local HIS area, then HIV incidence estimation can be performed for that stratum without regrouping.
In theory, a stratum can exist as long as it includes at least one BED-recent case. However, the user
should keep in mind that the smaller the number of cases in a selected stratum, the more unstable the
resulting estimate. For example, using a stratum that contains just one BED-recent case would
essentially mean estimating HIV incidence for an entire group of people based on a single case’s
information. Therefore, CDC does not recommend using any strata that do not satisfy the 20%/40/10
requirement.
The end result of reviewing Table 5 should be a decision as to the desired regrouping of potential strata.
The user will perform the actual regrouping in an interactive portion at the beginning of the third SAS
program.
3.3. PGM3_DETERMINE_STRATA
PGM3_DETERMINE_STRATA begins with an interactive section in which the user should regroup
potential strata based on the following instructions:
*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*;
*
ATTENTION: USER INTERACTION REQUIRED!
*;
*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*;
* Use Table 5 to regroup demographic variables below so that they satisfy the *;
* 20%/40/10 requirement across all selected years.
*;
*
*;
* Note: If you are unable to create any strata that satisfy the 20%/40/10
*;
* requirement for a given year or years, then you will only be able to
*;
* generate an unstratified HIV incidence estimate for that year(s).
*;
*
*;
* Example: You are estimating HIV incidence for years 2006-2008. Your 2006
*;
* data cannot be stratified to meet the 20%/40/10 requirement, but for 2007
*;
* and 2008 you can regroup the agex4 variable to meet the requirement using
*;
* strata for those younger than 30 and those 30 and above. You should regroup *;
* the agex4 variable in the code below as necessary to meet the 20%/40/10
*;
* requirement for both 2007 and 2008, even if the strata will not satisfy the *;
* requirement for 2006--and understand that you will only be able to generate *;
* an unstratified estimate for 2006 (i.e., select 'all' as your 2006
*;
* stratification variable in the next program, PGM4_ESTIMATE_INCIDENCE). You *;
* will still be able to compare overall estimates across all 3 years, but you *;
17
* will only be able to compare stratified estimates for the two age strata
*;
* between 2007 and 2008.
*;
*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*;
%macro strata_group;
length sex2 race4 agex4 mode3 $15;
* Potential stratification variable: Sex (2 default levels)
if sex = 'M' then sex2 = '1Male';
else sex2 = '2Female';
*;
* Potential stratification variable: Race/ethnicity (4 default levels)
if race = '6' then race4 = '1White';
else if race = '4' then race4 = '2Black';
else if race = '1' then race4 = '3Hispanic';
else race4 = '4Other';
*;
* Potential stratification variable: Age at infection (4 default levels)
if agex <= 12 then delete;
else if agex <= 29 then agex4 = '13-29';
else if agex <= 39 then agex4 = '30-39';
else if agex <= 49 then agex4 = '40-49';
else agex4 = '50+';
*;
* Potential stratification variable: Transmission category (3 default levels) *;
if rf = '01' then mode3 = '1MSM';
else if rf in ('02' '03') then mode3 = '2IDU';
else mode3 = '3HET/OTH';
%mend strata_group;
*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*;
The default potential strata are the two sexes (variable sex2), four race/ethnicity groups (race4), four
age-at-infection groups (agex4), and three transmission categories (mode3). The age-at-infection
variable, agex4, is created by adjusting age at diagnosis based on imputed STARHS result and testing
group assignments. Since the HIV incidence estimation model applies only to adults and adolescents, all
cases that are calculated to have been infected at less than 13 years of age are deleted from further
processing. This element of the regrouping code above should not be changed—i.e., pediatric infections
should never be included in the HIV incidence estimation process.
Continuing with the earlier example of regrouping transmission category strata (see description of Table
5, page 15), the user may decide to modify the default code as follows in order to meet the 20%/40/10
requirement for 2006, 2007, and 2008:
* Potential stratification variable: Transmission category (3 default levels) *;
if rf = '01' then mode3 = '1MSM';
else mode3 = '2IDU/HET/OTH';
However, if the user would rather have the option of estimating HIV incidence among IDU for 2007 and
2008, then he could leave the default strata coding intact:
* Potential stratification variable: Transmission category (3 default levels) *;
if rf = '01' then mode3 = '1MSM';
else if rf in ('02' '03') then mode3 = '2IDU';
else mode3 = '3HET/OTH';
18
If the user selects the second option, he must remember that mode3 should not be used as a
stratification variable for the 2006 HIV incidence estimate, as the IDU stratum did not satisfy the
20%/40/10 requirement for 2006. Selection of the stratification variable to produce the most robust HIV
incidence estimate will be discussed further in section 3.4 below.
After making any necessary changes to the default strata, the user can submit
PGM3_DETERMINE_STRATA to generate the final HIV incidence estimation data set,
XX_FINAL_HIS_EST_YEAR(S), which includes the newly-defined potential stratification variables. The
program also produces the BED-recent heterogeneity review table, which should be used to identify the
best stratification variable to be used in estimating HIV incidence for each year.
Table 6. Determine Stratification Variable Based on BED-Recent Heterogeneity
PGM3_DETERMINE_STRATA ends by outputting Table 6 to the XX Year(s) BED-Recent Heterogeneity
Review PDF file in the “estlib” directory. This table is designed to help the user identify which potential
stratification variable—sex2, race4, agex4, or mode3—contains the most heterogeneity in percentage of
BED-recent cases for each year. Proportions of BED-recent results that differ across strata may indicate
that recency of infection differs by stratum, which would further indicate that HIV incidence should be
estimated separately for each stratum. Therefore, for each year, the variable with strata encompassing
the largest overall range in the percentage of BED-recent cases should be selected as the single “best”
stratification variable to produce the most robust HIV incidence estimation. Stratification variables
should be selected on a year-by-year basis and can differ from year to year.
Consider the following table as an example:
XX Year(s) HIV Incidence Estimation Post-Imputation Strata Checks:
Table 6. Select Stratification Variable by Year Based on BED-Recent Heterogeneity
Year of Diagnosis=YYYY
STARHS Result
1Recent
2Long-Term
N
%
N
%
Sex
1Male
296
29.8
2Female
83
25.8
Race/Ethnicity
1White
140
32.4
2Black
168
27.4
3Hispanic
55
23.9
4Other
14
42.2
Age at Infection
13-29
146
30.8
30-39
96
26.7
40-49
96
28.1
50+
41
29.8
Transmission Category (Imputed)
19
Total
696
237
70.2
74.2
992
320
293
445
177
20
67.6
72.6
76.1
57.8
433
613
232
34
329
264
246
95
69.2
73.3
71.9
70.2
475
359
342
136
1MSM
2IDU
3HET/OTH
Total
STARHS Result
1Recent
2Long-Term
N
%
N
%
252
34.4
481
65.6
41
21.2
150
78.8
86
22.1
302
77.9
378
28.8
934
71.2
Total
733
191
388
1,312
In this table, race/ethnicity strata encompass the widest range (42.2 – 23.9 = 18.3) in percentage of BEDrecent STARHS results for year YYYY. Therefore, race/ethnicity (i.e., SAS variable race4) should be
selected as the best stratification variable for use in estimating HIV incidence for year YYYY.
Note: All four of the aforementioned potential stratification variables—i.e., sex2, race4, agex4, and
mode3—are included in Table 6. The strata appear in their regrouped formats, which were determined
by the user at the beginning of PGM3_DETERMINE_STRATA. Only the strata that satisfied the
20%/40/10 requirement for a given year are recommended for use in stratifying that year’s HIV
incidence estimate. To continue with an earlier example, consider the user who really wants to estimate
HIV incidence among injection drug users (see description of PGM3_DETERMINE_STRATA, page 17).
Assume that the user opted to maintain IDU as a separate stratum, which would enable him to generate
IDU-specific HIV incidence estimates for each of the three years. The user should be aware that the 2006
HIV incidence estimate for IDU should not be interpreted (or should be interpreted with extreme
caution, recognizing the limitations of the estimate for that stratum). Further, when examining Table 6
output for 2006, the user should ignore the transmission category breakout, as that variable should not
be chosen as the stratification variable to produce the most robust 2006 HIV incidence estimate
(regardless of any apparent heterogeneity).
3.4. PGM4_ESTIMATE_INCIDENCE
PGM4_ESTIMATE_INCIDENCE represents the final step of the HIV incidence estimation process. The
program begins with a brief interactive portion in which the user must enter the stratification variable(s)
selected during review of Table 6:
*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*;
*
ATTENTION: USER INTERACTION REQUIRED!
*;
*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*;
* Enter list of desired stratification variables (one for each incidence
*;
* estimation year) based on review of Table 6. For each year, among those
*;
* variables that met the 20%/40/10 requirement (either by default or after
*;
* regrouping), select the one with the most heterogeneous distribution of
*;
* percent BED-recent. Stratification variables can differ by year. Enter one *;
* stratification variable for each year, in order by year, separated by
*;
* spaces. End the list with a semicolon.
*;
*
*;
* Example: You are estimating HIV incidence for 2006 and 2007. All potential *;
* stratification variables were successfully regrouped to meet the 20%/40/10 *;
* requirement. Sex strata had the most heterogeneity in percent BED-recent
*;
* for Dx year 2006, but transmission category had the most heterogeneity in
*;
* percent BED-recent for Dx year 2007. Set the strata_list variable as:
*;
*
*;
*
%let strata_list = sex2 mode3
*;
20
*
*;
* Note: Selecting "all" for a given year indicates that no available strata
*;
* met the 20%/40/10 requirement for that year and will generate only an
*;
* overall, unstratified estimate.
*;
*
*;
* Example: When estimating HIV incidence for 2006-2008, your available data
*;
* did not meet the 20%/40/10 requirement for any potential stratification
*;
* variable for Dx year 2006. You were, however, able to regroup variables to *;
* meet the 20%/40/10 requirement for 2007 and 2008. In 2007 the race strata
*;
* showed the most heterogeneity in percent BED-recent, but in 2008 the
*;
* transmission category strata showed the most heterogeneity in percent BED- *;
* recent. Set the strata_list variable as:
*;
*
*;
*
%let strata_list = all race4 mode3
*;
*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*;
%let strata_list = ;
*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*~*;
As mentioned in the instructions above, the stratification variable all prompts the program to generate
an unstratified estimate. The user should select all as the stratification variable for any years for which
none of the potential stratification variables could be appropriately regrouped to meet the 20%/40/10
requirement.
After entering the desired stratification variable(s), the user can submit the program, which
accomplishes the following for each year:
For each of the 20 imputed data sets, within each stratum:
o Calculates the number of BED-recent results for both repeat and new testers
o Estimates the probability of being detected as BED-recent among repeat testers from
intertest interval T
o Estimates the probability of being detected as BED-recent among new testers based on
the proportion of concurrent AIDS diagnoses
o Estimates HIV incidence separately for repeat and new testers based on underlying
formula (i.e., HIV incidence equals the number of BED-recent results divided by the
probability of being detected as BED-recent), then sums to generate an overall HIV
incidence estimate
o Calculates 95% confidence intervals of HIV incidence estimate
Sums HIV incidence estimates across strata to generate a single estimate for each imputed data
set
Averages the 20 estimates to generate the final overall HIV incidence estimate
Averages the 20 estimates for each stratum to generate final stratified HIV incidence estimates
Outputs the HIV incidence estimate results tables containing the final overall and stratified HIV
incidence estimates for the selected year(s). (Note: Unlike the first three programs,
PGM4_ESTIMATE_INCIDENCE generates a separate PDF file for each year.)
PGM4_ESTIMATE_INCIDENCE creates many intermediate calculated variables that are used throughout
the program and summarized in the table below:
21
Variable Name
aids
g0p
inc
inc_g0
inc_g1
n_g0
n_g0_0
n_g0_a
n_g0_b
n_g1
n_g1_0
n_g1_a
n_g1_b
notest
q
q1
recent
reported
tested
v_g0p
v_g1p
v_inc
v_incg0
v_incg1
Description
Total number of concurrent AIDS cases
Probability of being detected as BED-recent among new
testers
Total HIV incidence estimate
HIV incidence estimate for new testers
HIV incidence estimate for repeat testers
Total number of new testers
Number of BED-recent new testers
Number of new testers with concurrent AIDS
Number of BED-long-term new testers
Total number of repeat testers
Number of BED-recent repeat testers
Number of repeat testers with concurrent AIDS
Number of BED-long-term repeat testers
Total number of new testers
Percentage of concurrent AIDS among new testers
Percentage of concurrent AIDS among repeat testers
Total number of BED-recent cases
Total number of reported cases
Total number of repeat testers
Adjustment for calculating variance among new testers
Adjustment for calculating variance among repeat testers
Variance of total HIV incidence estimate
Variance of HIV incidence estimate for new testers
Variance of HIV incidence estimate for repeat testers
PGM4_ESTIMATE_INCIDENCE ends by generating a XX Year HIV Incidence Estimate Results PDF file for
each year included in the estimation. Each file includes overall and stratified HIV incidence estimates
(calculated variable inc) for a single year, along with the following additional information:
Number of BED-recent results (recent)
Number of reported diagnoses of HIV disease (reported)
HIV incidence estimate among new testers (inc_g0)
HIV incidence estimate among repeat testers (inc_g1)
Standard deviation of the HIV incidence estimate (sd)
Lower (low95) and upper (high95) bounds of 95% confidence interval for HIV incidence estimate
Each yearly output file contains HIV incidence estimates for all potential strata included in the
regrouping process; however, estimates based on the selected “best,” most heterogeneous
stratification variable (entered at the start of PGM4_ESTIMATE_INCIDENCE) appear first and should be
considered the most robust overall and stratified HIV incidence estimates. Estimates for the other, lessheterogeneous strata can be calculated by applying the stratum-level proportions of total estimated HIV
incidence based on the less heterogeneous stratification variable to the overall most robust HIV
incidence estimate. Confidence intervals can be similarly calculated by applying an adjustment factor
22
(overall most robust HIV incidence estimate divided by overall HIV incidence estimate based on lessheterogeneous stratification variable) to the stratum-level confidence intervals based on the lessheterogeneous strata.
For example, consider the HIV incidence estimation process for year YYYY in which the selected “best,”
most heterogeneous stratification variable was mode3. The user would like to calculate HIV incidence
estimates by age at infection (agex4), too. To do so, the user must refer to output from the mode3 and
agex4 pages of the XX YYYY HIV Incidence Estimate Results PDF file. The first step is to use the “By
agex4” table to calculate the proportion of total estimated HIV incidence represented by each level of
agex4:
By agex4
Strata
13-29
30-39
40-49
50+
Number
of BEDrecent
146
96
96
41
378
Number
of
reported
HIV
diagnoses
558
459
412
162
1,592
HIV
incidence
estimate
among
new
testers
218
229
149
73
670
HIV
incidence
estimate
among
repeat
testers
286
239
252
92
869
HIV
incidence
estimate
(0.33) 504
(0.30) 468
(0.26) 401
(0.11) 166
1,539
Standard
deviation
for HIV
incidence
estimate
108
106
90
58
Confidence
interval for
HIV
incidence
estimate:
lower
bound
289
260
222
51
Confidence
interval for
HIV
incidence
estimate:
upper
bound
719
677
580
280
These proportions are based on an estimate stratified by agex4, which was not the most heterogeneous
variable and therefore does not represent the most robust estimate. The calculated proportions should
then be applied to the overall most robust HIV incidence estimate from the mode3 output page:
XX YYYY HIV Incidence Estimate: Stratified by mode3
** Estimate Based on Most Heterogeneous Strata **
Overall
Number of
BED-recent
378
Number of
reported
HIV
diagnoses
1,592
HIV
incidence
estimate
among
new
testers
643
HIV
incidence
estimate
among
repeat
testers
860
HIV
incidence
estimate
1,503
Standard
deviation
for HIV
incidence
estimate
215
Confidence
interval for
HIV
incidence
estimate:
lower
bound
1,074
Confidence
interval for
HIV
incidence
estimate:
upper
bound
1,932
The results of applying these proportions are revised stratum-level HIV incidence estimates for agex4
(based on the more robust mode3 estimate):
Strata
13–29
Revised HIV
Incidence Estimate
0.33 × 1,503
= 496
23
Revised HIV
Incidence Estimate
0.30 × 1,503
= 451
0.26 × 1,503
= 391
0.11 × 1,503
= 165
Strata
30–39
40–49
50+
Next the user would like to calculate confidence intervals for the newly calculated HIV incidence
estimates by age-at-infection strata. The first step is to calculate an adjustment factor, which in this case
equals the overall HIV incidence estimate based on mode3 divided by the overall estimate based on
agex4: 1,503 ⁄ 1,539 = 0.98. The user should then apply this proportion to the stratum-level confidence
intervals in the “By agex4” table above to obtain more robust confidence intervals for the age-atinfection HIV incidence estimates:
Strata
Revised HIV
Incidence Estimate
13–29
496
30–39
451
40–49
391
50+
165
Revised C.I., Lower
Bound
0.98 × 289
= 283
0.98 × 260
= 255
0.98 × 222
= 218
0.98 × 51
= 50
Revised C.I., Upper
Bound
0.98 × 719
= 705
0.98 × 677
= 663
0.98 × 580
= 568
0.98 × 280
= 274
4. Interpreting HIV Incidence Estimates
The fifth and sixth CDC-provided SAS programs, PGM5_COMPARE_ESTIMATES and
PGM6_EVALUATE_DIFFERENCES, are designed to assist the user in comparing and interpreting one or
more HIV incidence estimates.
4.1. PGM5_COMPARE_ESTIMATES
PGM5_COMPARE_ESTIMATES should be used when the local HIS areas has generated multiple singleyear HIV incidence estimates and should be run immediately following programs 1–4, as it uses the
same initial setup completed in PGM1_PREPARE_DATA. Program 5 uses the initial and imputed HIV
incidence estimation data sets, XX_HIS_EST_DATA_YEAR(S) and XX_IMPUTED_YEAR(S), to calculate various
frequencies and percentages associated with factors that may affect HIV incidence estimates across
years. The program ultimately outputs a single table to the PDF file XX Year(s) HIV Incidence Estimate
Comparison in the “estlib” directory. The table is designed to assist the user in understanding
differences that may exist across two or more single-year HIV incidence estimates. More specifically, the
table addresses the underlying factors that impact each estimate. All frequencies and percentages
displayed in the estimate comparison report are based on the imputed data that were used for the final
HIV incidence estimates. Rows shaded in grey are purely informational and are less important than
other rows.
24
Before reviewing the estimate comparison report, the user should take a moment to revisit the
foundations of the HIV incidence estimation model. The basic formula is N = R ⁄ P, where N is the
number of estimated new HIV infections (i.e., HIV incidence), R is the number of BED-recent STARHS
results, and P is the estimated probability of being detected as BED-recent at HIV diagnosis. In a
simplistic scenario in which P is held constant, an increase in R will lead to a higher HIV incidence
estimate, and a decrease in R will lead to a lower HIV incidence estimate. Because HIV incidence is
estimated separately for new and repeat testers and summed to generate a final estimate, the formula
can also be written as N = Nnew + Nrep = (Rnew ⁄ Pnew) + (Rrep ⁄ Prep).
By definition, new testers are tested less frequently than repeat testers and thus have a lower likelihood
of being detected as BED-recent at HIV diagnosis. Therefore, each BED-recent new tester carries more
weight in the estimation process than each BED-recent repeat tester. Furthermore, because new and
repeat testers have different testing behaviors, their probabilities of being detected as BED-recent at
HIV diagnosis are estimated differently:
The most important parameter in estimating the new tester’s probability of being detected as
BED-recent at HIV diagnosis (Pnew) is the proportion of concurrent HIV/AIDS diagnoses (q). Pnew is
inversely related to q: The higher the proportion of concurrent HIV/AIDS diagnoses, the lower
the probability of a new tester being detected as BED-recent at HIV diagnosis. If the number of
BED-recent cases (Rnew) is held constant, then the HIV incidence estimate for new testers (Nnew)
will increase when q increases.
The most important parameter in estimating the repeat tester’s probability of being detected
BED-recent at HIV diagnosis (Prep) is the time interval between the last negative test and the first
positive test (T). Prep and T are inversely proportional as well: The longer the time interval
between last negative and first positive tests, the smaller the probability that a repeat tester will
be detected as BED-recent at HIV diagnosis. If the number of BED-recent cases (Rrep) is held
constant, then the HIV incidence estimate for repeat testers (Nrep) will increase when T
increases.
Two factors that are related the numbers of BED-recent results among both new and repeat testers
(Rnew and Rrep, respectively) are:
Number of new HIV diagnoses: Typically, an increase in true HIV incidence for a given year will
result in a commensurate increase in HIV diagnoses; therefore, the more new diagnoses of HIV
disease occur in a given year, the higher the estimated HIV incidence for that year will be.
Proportion of BED-recent: The proportion of BED-recent results determines the number of
observed recent infections. Typically, the higher the proportion of BED-recent results, the higher
the HIV incidence estimate will be.
In the absence of evidence suggesting changes in testing behavior (i.e., changes in q and/or T), a change
in the proportion of BED-recent results may indicate a real change in HIV incidence. On the other hand,
if testing behavior does change suddenly, then an accompanying change in the HIV incidence estimate
does not necessarily represent a true change in HIV incidence. One of the underlying assumptions of the
25
estimation model is that testing behavior has not changed in recent years; therefore, the model is
sensitive to sudden changes in testing, and a change in estimated HIV incidence that is accompanied by
sudden changes in testing may not represent a change in true HIV incidence.
With this understanding of how the number of new HIV diagnoses, proportion of concurrent AIDS
diagnoses (q), intertest interval (T), and proportion of BED-recent results interact, the user can examine
the XX Year(s) HIV Incidence Estimate Comparison PDF output to evaluate differences across two or
more single-year HIV incidence estimates. The comparison report should be reviewed along with the HIV
incidence estimate results tables generated by PGM4_ESTIMATE_INCIDENCE. Consider the sample
2006–2008 HIV incidence estimates for area XX and the accompanying HIV incidence estimate
comparison report:
Year
2006
2007
2008
HIV Incidence Estimates
New
Repeat
Testers
Testers
Total
893
686
1,579
643
860
1,503
1,024
679
1,703
XX 2006-2008 HIV Incidence Estimate Comparison:
Comparing Potential Factors Affecting HIV Incidence Estimates Across Years
2006
Factor
All new HIV diagnoses
New testers
AIDS at HIV Dx
AIDS Dx 1-6 mos. after HIV
With STARHS results
Among cases with STARHS results, BED-recent
New testers
Among new testers with STARHS results, BED-recent
Among new testers, AIDS at HIV Dx
Repeat testers
Among repeat testers with STARHS results, BED-recent
Among repeat testers, AIDS at HIV Dx
Among repeat testers, average T (in months)
Among repeat testers, median T (in months)
N
1,600
648
331
176
1,093
342
2007
2008
%
N
%
N
%
. 1,600
. 1,600
.
40.5
685 42.8
659 41.2
20.7
285 17.8
295 18.4
11.0
196 12.3
187 11.7
68.3 1,119 69.9 1,118 69.9
31.3
378 33.8
370 33.1
89 24.0
177 27.3
79 19.8
177 25.8
111 27.7
172 26.1
252 34.9
154 16.2
32
.
13
.
299 41.6
108 11.8
39
.
24
.
260 36.3
123 13.1
33
.
24
.
A few points of interest:
The number of new diagnoses of HIV disease stayed constant across all three years. The
percentage of new testers also stayed roughly the same.
26
The proportion of concurrent AIDS diagnoses was slightly higher in 2006, but the number of
BED-recent results was lower in 2006 than in other years; therefore, the proportion of
concurrent AIDS diagnoses should not have a big impact on the HIV incidence estimates.
The overall proportion of BED-recent results was about the same across all three years, but it
differed quite a lot between new and repeat testers. The 2008 percentage of BED-recent results
among new testers was higher than in 2006 and 2007, and this explains the similar increase in
the 2008 HIV incidence estimate among new testers.
Because proportions of concurrent AIDS diagnoses did not change much across the years,
testing behavior should not have changed much, either; however, the 2008 percentage of BEDrecent results among new testers increased, so the higher 2008 HIV incidence estimate among
new testers may represent a true increase in HIV incidence.
The 2007 percentage of BED-recent results among repeat testers was much higher than the
same percentage for either 2006 or 2008. The average T for repeat testers in 2007 was longer
than in 2006 and 2008, and the percentage of new testers was slightly higher in 2007,
suggesting either that some of the newly-diagnosed repeat testers had not been tested in quite
awhile or that repeat testers tested less frequently. The cumulative effect of increased
percentage of BED-recent results and longer inter-test interval (T) resulted in an increased 2007
HIV incidence estimate among repeat testers. However, because the data indicate that there
might have been a change in testing behavior, it is difficult to determine whether the increased
estimate represents a true increase in HIV incidence for repeat testers in 2007.
4.2. PGM6_EVALUATE DIFFERENCES
PGM6_EVALUATE_DIFFERENCES provides additional tools to assist the user in interpreting local HIV
incidence estimates. This program requires the user to enter various numbers by hand but does not
refer to any input data sets; therefore, PGM6_EVALUATE_DIFFERENCES can be run at any time. The sixth
and final program is designed to allow for future additions; at present, the program allows the user to
(1) calculate confidence intervals for HIV incidence rate estimates and (2) determine whether two HIV
incidence estimates are significantly different. The program requires user input as described in the
following instructions:
*******************************************************************************;
* 1. To calculate confidence intervals for HIV incidence rate estimates,
*;
* values for the following macro variables: inc (HIV incidence estimate),
*;
* se_inc (standard error of HIV incidence estimate), pop (population
*;
* estimate), and se_pop (standard error of population estimate).
*;
*******************************************************************************;
%let inc = ;
%let se_inc = ;
%let pop = ;
%let se_pop = ;
*******************************************************************************;
* 2. To determine whether two HIV incidence estimates are significantly
*;
* different, enter values for inc1 (HIV incidence estimate #1), se1 (standard *;
* error of HIV incidence estimate #1), inc2 (HIV incidence estimate #2), and *;
* se2 (standard error of HIV incidence estimate #2).
*;
*******************************************************************************;
%let inc1 = ;
27
%let se1 = ;
%let inc2 = ;
%let se2 = ;
Results of both calculations are printed to the SAS Output window.
5. Contact Information
If the user has any questions or concerns about this document or local HIV incidence estimation in
general, he or she should contact the HICSB epidemiologist assigned to provide technical assistance to
his or her HIS area. The main HICSB telephone number is (404) 639-2050.
28
HIV Incidence Surveillance Completeness Report 1/20/2010
Appendix 7- Documentation for HIV Incidence Completeness Report
Purpose
The HIV incidence completeness report provides a snapshot of the proportion of HIV cases diagnosed
within a three year period that have required TTH data elements and BED results. The HIV incidence
completeness report should serve as a data quality monitoring tool for CDC epidemiologists and
state/local coordinators to measure HIS data completeness, identify problems with data transfer and
monitor progress toward meeting target performance levels and data quality standards. Even though
the criteria used in the HIV incidence completeness report are not necessarily the criteria used for the
annual national evaluation, high completeness rates will support the HIV incidence surveillance
programs to meet the outcome standards. If issues are identified CDC will provide technical assistance
to improve the completeness of the data. The HIV incidence completeness report will be generated on
a monthly basis starting in January 2010.
Description
The HIV incidence completeness report is composed of three sections: (1) completeness of TTH
required data elements; (2) completeness of cases having a sample with STARHSID and BED result;
and (3) proportion of data contributing unique information for HIV incidence analysis.
Section 1: Completeness of TTH required data elements
This section reflects an HIV incidence surveillance program’s efforts in collecting data for the
following TTH data elements:
TTH date of first positive HIV test
Ever tested HIV negative
Number of negative tests in the 2 years before first positive (including the first positive –
following eHARS implementation only negative test before first positive tests will be
included, the first positive test will not be included)
Ever taken ARVs or ARV use dates
TTH date of last negative HIV test
ARV use dates
Completeness of TTH
Completeness is defined as the value for the data element is not blank; the field is deemed complete as
long as there is a response to the variable, including responses of “Refused” or “Unknown”. For date
variables, the value can be 9999, as long as incidence surveillance uses the IVR database. After the
transition to eHARS, that value will not be available. Completeness of each TTH data element is
calculated separately.
A complete TTH is defined as having a value for least one of the six TTH data elements. The six data
elements are: (1) TTH date of first positive; (2) ever tested HIV negative; (3) number of negative tests
in the 2 years before first positive; (4) TTH date of last negative HIV test; (5) ever taken ARVs; (6)
ARV use dates..
The completeness of TTH date of first positive, ever tested negative, number of negative tests in the
past 2 years before first positive, ever taken ARVs or dates of ARV use, and completeness of TTH are
based on all HIV cases diagnosed in a specific year. The completeness of TTH date of last negative is
Cleared by HICSB Management 10/15/2012
HIV Incidence Surveillance Completeness Report 1/20/2010
based on cases that have a “Yes” for ever tested negative. The completeness of ARV dates is based on
cases that have a “Yes” for ARV use.
Section 2: Completeness of cases having a sample with STARHSID and BED result
This section measures the completeness of cases
With a STARHSID
Without a STARHSID but having a reason for not testing
With BED result of recent or long term
All measures are based on all HIV cases diagnosed in a specific year.
The measure of cases with a STARHSID determines the proportion of cases having a sample that has
been identified by the incidence program as “to be tested” from a test/toss list or by the state’s public
health laboratory. The second measure (cases without STARHSID but having a reason for not testing)
determines the proportion of cases that do not have a sample for testing, but a reason for not testing has
been identified by the incidence program. The third measure (cases with BED result) determines the
proportion of cases having a sample with a BED test result either “Recent” or “Long term”. Cases with
a BED result are a subset of cases having a sample with STARHSID. Therefore, from the difference of
the two measures (cases with a STARHSID and cases with BED result), sites may identify cases that
have pending samples for BED results or did not have a test result. Significant differences between the
two measures need to be investigated.
Section 3: Proportion of data contributing unique information for HIV incidence analysis
This section measures the proportion of data that contributes unique information for HIV incidence
estimation. Data contributing unique information for analysis include responses that help to classify a
case as a repeat or new tester or having a recent vs. long standing infection that would only be obtained
as a part of HIV incidence surveillance.
The proportion of TTH data contributing unique information for analysis is based on all HIV cases for
a specific year. TTH data contributing unique information for analysis is defined as having at least one
valid answer for any of the six required data elements:
1. TTH date of first positive HIV test has a valid year and a date that is earlier than the eHARS
diagnosis date.
Note: For HIV incidence analyses, the eHARS diagnosis date is selected when TTH date of
first positive is the same as eHARS diagnosis date. However additional information to reclassify a case as recent or long term infection is available if the TTH date of first positive is
earlier than the eHARS diagnosis date. If the TTH date of first positive is 6 months or earlier
than the eHARS diagnosis date, the BED results of the case will be set to “Long term”. If the
TTH date of first positive is 1-6 months earlier than eHARS diagnosis date, the BED results
will be set to missing.
2. Answer to ever tested HIV negative is “Yes” or “No”.
Note: If the answer to ever tested negative is “Yes”, the case is classified as a repeat tester; if
the answer is “No”, the case is classified as a new tester. Surveillance staff should never enter
“No” if previous testing status is not known.
3. Answer to number of HIV negative tests in the 2 years before first positive is not blank,
“Refused” or “Don’t know”.
Cleared by HICSB Management 10/15/2012
HIV Incidence Surveillance Completeness Report 1/20/2010
Note: If the number of HIV negative tests in the 2 years before first positive is >=2, the case is
classified as a repeat tester, and if the answer is “0”, the case is classified as a new tester.
Currently CDC is unable to use the value “1” because it has been inconsistently collected.
4. Date of last negative test year must have a valid year.
Note: If there is a valid date of last negative test, the case is classified as a repeat tester.
5. Answer to ever taken ARVs is “Yes” or “No”
6. Dates of ARV use exist and are valid.
Note: If answer to ever taken ARVs is “Yes” and if the ARV use is during the six months prior
to collection of the specimen used for the BED test, the BED result of the case is set to missing.
The proportion of BED data contributing unique information for analysis is defined as having BED test
results of “Recent” or “Long term”. The measure is calculated in two ways. The first measure excludes
cases with AIDS diagnosed within 6 months of HIV diagnosis from the denominator. In the analysis,
the BED result for cases with AIDS diagnosed within 1- 6 months of HIV diagnosis is set to “Long
term”, regardless of their original BED result. This measure is useful for assessing the completeness
for cases diagnosed in the previous years, but because it does not allow time for reporting delay, it is
not as helpful in assessing completeness in the current year.
The second measure of BED completeness is useful to assess cases diagnosed in the current year. In
the second measure, cases with AIDS diagnosed in the same month of HIV diagnosis are excluded
from both numerator and denominator.
How to use the HIV incidence completeness report:
1. The denominator includes cases reported from each HIS area and residing in areas funded for HIS.
2. The proportion of data contributing unique information to analysis is added to emphasize the content
of the data rather than the collection and reporting of the data.
3. The completeness of each data element and completeness of TTH (Section1) reflects the overall
efforts in data collection; all values (as long as there is a response in the field, including responses of
“Refused” or “Unknown”) are accepted. The proportion of TTH data contributing unique information
for incidence analyses (Section 3) only counts answers that provide unique information for analysis,
therefore, it is usually lower than the completeness proportion of TTH or each data element. Among
the TTH data elements, completeness of TTH date of first positive is particularly more than the TTH
date of first positive contributing unique information for incidence analysis. As the eHARS diagnosis
date variable is used for incidence analyses, the TTH date of first positive only provides unique
information for analysis if it has a valid date that is earlier than the eHARS diagnosis date. It is
important to highlight that TTH variable values of “Refused” or “Unknown” do not provide additional
information for analysis.
4. The proportion of BED results contributing unique information for incidence analysis for cases
diagnosed in the previous years is higher than for cases diagnosed in the current year due to delays in
reporting time.
Cleared by HICSB Management 10/15/2012
HIV Incidence Surveillance Error Report 06/08/2012
Documentation for HIV Incidence Error Report
Purpose
The HIV incidence error report is a tool to monitor data entry errors and inconsistencies between variables and
improve the quality of HIV incidence surveillance (HIS) data. The purpose is to support the efforts of state/local
incidence coordinators in improving the validity and accuracy of the data and monitoring progress toward meeting
target performance levels and data quality standards. If errors are checked and fixed routinely, HIS professionals can
detect problems with data entry and data collection, and improve data quality. The HIV incidence error report is
generated quarterly (June, September, December, and March). Errors should be resolved prior to submission of
incidence datasets in August, November, February and May.
Description
The HIV incidence error report is an excel file titled XX_error_yyyymm.xls (where xx is the 2-letter or 3-letter as
appropriate abbreviation for the state/local incidence area) with an error summary (worksheet 1) and an error details
(worksheet 2).
Worksheet 1: Error Summary
The error summary is a summary report of all the errors, including types and the frequencies of each error. The error
summary has three columns: Errormsg, Errortype and Count. Errormsg is the description of each error. Errortype
identifies whether the error is related to TTH, laboratory or core surveillance, and Count is the column for frequency, or
number of records with that error (see example below). Based on the frequency of the errors, program areas can
identify the most frequently occurring errors and prioritize corrective actions.
Errormsg
3. TTH date of last neg after eHARS date of diagnosis of HIV infection
Errortype
TTH
Count
10
Worksheet 2: Error Details
The error details sheet provides case-level error information. Errors are sorted by diagnosis year, stateno/cityno, and
entered year. The error detail worksheet has 11 columns:
1
Column Name
PID
2
3
4
5
State_city_No
Errortype
Entered_yr
Dx_yr
6
CheckingID
7
VariableA
8
VariableB
9
10
Errormsg
Fixable
11
Comment
Column Description
A Pseudo ID variable was created in order to link the error report and the error resolution file
records together. This is important because the State_city_No column contents are cleared
from the resolution file before the files is sent to CDC.
State_city_No is a unique identifier that helps in the identification of the case.
Determines whether the error is related to TTH or laboratory.
Provides the year that the case was entered into eHARS
Provides the year the case was diagnosed as HIV infection that is the earlier date between
HIV diagnosis date and AIDS diagnosis date.
Helps with the identification of the case using document_uid or starhsid. HIS data in eHARS
is document based with multiple TTH and Lab documents. Therefore, this new variable has
been added to the error report to provide additional help in identifying which document
within a case is potentially problematic.
Displays the variable names and values with errors. VariableA is used in combination with
VariableB to identify data entry errors and inconsistencies between variables.
Displays variable names and values with errors. VariableB is used in combination with
VariableA to identify data entry errors and inconsistencies between variables.
Provides a description of the error. The table below displays each error message.
Sites should record whether the error can be resolved or not using the values “YES” or
“NO”.
Sites should indicate the reason the error is not fixable using the codes A-E or specify any
other comments described below.
Cleared by HICSB Management 10/15/2012
HIV Incidence Surveillance Error Report 06/08/2012
The first 9 columns (PID, State or city no, Errortype, Entered_yr, Dx_yr, CheckingID, Variable A, Variable B and
Errormsg) are completed by CDC.
The columns Fixable and Comment are for program areas to use. Fixable is used to record whether the error can be
resolved or not; values for Fixable field should be “YES” or “NO”. If the error is fixable the program area should indicate
“YES” in the Fixable field and correct the error in eHARS; fixed errors will be reflected in the monthly SAS datasets and
will therefore not reappear in subsequent error reports. If the error is not fixable (the data must remain as entered) the
program area should indicate “NO” in the Fixable field and provide additional comments in the Comment field.
Likewise, the data inconsistencies will also not reappear on subsequent error reports. Use the list below to enter the
code for the most commonly used comments:
A. Self-reported information:
Data entered correspond to the original data collection form provided by the patient, provider or DIS.
B. Unable to determine from chart abstraction
DIS or incidence staff re abstracted data from chart, but they were not able to fix the error based on the information
obtained.
C. Unable to follow up
DIS or incidence staff was unable to conduct further investigation to fix the error. Some possible reasons include that
the patient was lost to follow up (e.g. diagnosed previously in another state, or patient is unable to be re interviewed),
or the chart was not available for re-abstraction.
D. Ineligible client
Patient is not a new diagnosis or is from a jurisdiction not participating in HIS.
E. Data verified
This is specifically for specimens with an obtained date that is not within 0-3 months of eHARS date of diagnosis of
HIV infection. Due to the nature of the data collection and reporting process available specimens may not have been
drawn within 0-3 months of the eHARS date of diagnosis of HIV infection and program areas can do nothing about it.
For example, a specimen was sent for STARHS and an earlier diagnosis date was subsequently found.
If there are other comments beyond this list, please briefly describe in the Comment field.
The error details worksheet is protected. As a result, program areas cannot delete rows or columns or make changes
to the worksheet except for four columns: State_city_no, CheckingID, Fixable and Comment. Fixable and Comment
are used to record resolutions for errors and the content of State_city_no and CheckingID columns must be cleared
before the resolution file is sent to CDC.
Data Error Checks and Recommended actions
General comments:
Data errors are errors associated with data entry or inconsistencies between variables. There are three categories of
errors: TTH-related, laboratory-related and core surveillance related. Based on the categories of the errors, program
areas can refer to relevant documents or personnel to investigate. All errors identified in the error report should be
investigated by verifying the source documents or conducting additional follow-up activities. Data entry errors should
be corrected after investigation.
Data inconsistencies reflect problems associated with data collection and reporting process that should be
investigated. Inconsistencies in the data may require additional follow up activities including re-abstraction of data,
contacting the reporter for clarification or re-interview of the patient if a short time has elapsed between the original
interview and the time the error is identified. If systematic errors or inconsistencies are identified re-training of data
collection staff and reporters may be necessary. Training may need to include review of interviewing skills and data
abstraction procedures. CDC has developed Job Aids to assist all HIS sites in conducting these training activities.
The following table contains error checks, action priority and suggestions for how they may be addressed. Astute
observers will note that two errors on the previous error report are no longer reported.
Error #6 has been deleted and replaced by error #5 which is a combination of the former Errors #5 and #6
from the previous error report.
Cleared by HICSB Management 10/15/2012
HIV Incidence Surveillance Error Report 06/08/2012
Error #8 “BED result exists, specimen collection date before eHARS date of diagnosis of HIV infection” has
been deleted in order to avoid redundancy as Error 8 is a subset of Error #13.
Priority
Error (errormsg)
BED result exists, but STARHSID missing
Type
Lab
Comments
Check all STARHS laboratory reports related to
the case to obtain the STARHSID.
BED result exists, but specimen collection date
missing
Lab
Check all laboratory reports to obtain the specimen
collection date.
TTH date of last negative after eHARS date of
diagnosis of HIV infection
TTH
3
Refer to original source of report for the TTH and
laboratory reports to verify.
TTH date of last negative after TTH date of first
positive
TTH
4
Refer to original source of report for the TTH and
laboratory reports to verify.
If “ever tested negative” is “refused”, “unknown”
or blank but number of negative tests >=1.
TTH
Verify TTH source for data entry errors. If not data
entry errors, further investigation may be required.
If a systematic error is detected re-training of the
staff and reporters may be required.
6
Blank
Blank
Error #6 has been deleted and replaced with
Error #5
7
BED result exists, specimen collection date after
STARHS test date
8
Blank
1
2
5
9
10
11
12
13
14
15
Lab
Blank
18
Error #8 been deleted as it is a subset of
Error #13.
TTH date of first positive before 1975
TTH
Check for data entry or record error.
TTH date of first positive before birth date from
eHARS
TTH
Verify birth date and verify TTH date of first
positive.
If TTH last negative date within 24 months of
TTH first positive date, and “number of negative
tests* within 24 months before first positive“ < 1
TTH
If “ever tested negative” is “no”, “refused”,
“unknown” or blank, and “date of last negative
test “ exists
TTH
BED result exists, specimen collection date not
within 0-3 months of eHARS date of diagnosis of
HIV infection*
Lab
Ever tested positive is Yes, first positive test
date is missing
TTH
First positive date after eHARS date of HIV
diagnosis
TTH
Invalid self-reported last negative date
TTH
Invalid specimen collection date
Lab
Invalid SOD values
Lab
Verify TTH source for data entry errors. If not data
entry errors, further investigation may be required.
If a systematic error is detected re-training of the
staff and reporters may be required.
Verify TTH source for data entry errors. If not data
entry errors, further investigation may be required.
If a systematic error is detected re-training of the
staff and reporters may be required.
Check the accuracy of the specimen collection
date. This error is very likely due to the availability
of the specimen and the nature of data collection
and reporting process.
Verify TTH source for data entry errors. If not data
entry errors, further investigation may be required.
If a systematic error is detected re-training of the
staff and reporters may be required.
Verify TTH source for data entry errors. If not data
entry errors, further investigation may be required.
If a systematic error is detected re-training of the
staff and reporters may be required.
Verify TTH source for data entry errors. If not data
entry errors, further investigation may be required.
If a systematic error is detected re-training of the
staff and reporters may be required.
Check the accuracy of the specimen collection
date.
Check for data entry or record error.
16
17
Check laboratory related reports and STARHS
result excel files.
Cleared by HICSB Management 10/15/2012
HIV Incidence Surveillance Error Report 06/08/2012
How to use the error report:
The error report is generated quarterly at the beginning of June, September, December and March. Error Reports are
uploaded to the SDN for areas to retrieve with a password of XXDQ (where XX represents the site name abbreviation).
CDC epidemiologists will discuss progress on resolving the incidence error report as a part of the routine monthly call.
Project areas will investigate and determine the resolutions for errors.
(1) Fixable errors should be corrected in eHARS and the corrections will be reflected in monthly SAS datasets.
(2) For errors that are not fixable, record “No” in the Fixable column, and provide reasons in the Comment column
using the comment list provided. Any error marked “No” will not appear on future error reports.
(3) When finished, clear the content of the State_City_No and CheckingID columns in the Error_details worksheet,
save the file as XX_error_yyyymm_resolution.xls (where XX is the project area abbreviation) and email it to CDC
incidence data manager (Mona Doshani – [email protected] or [email protected]) by Aug 15, Nov15, Feb15 and
May15.
Because the error details worksheet is protected, columns cannot be deleted, but the contents of the columns can be
cleared. To clear the content of the State_city_no and the CheckingID columns prior to returning the resolution file to
CDC, right click on the State_city_no and the CheckingID column letter and select Clear Contents from the popup
menu. The resolution file must be sent back to the CDC incidence data manager by August 15, November 15,
February 15 and May 15, because the resolutions of the errors will be incorporated in the subsequent error reports.
Errors that have been fixed or investigated but with no possible resolution will not appear in the subsequent error
reports.
It is recommended that HIS program areas keep a copy of the resolution file with State_city_no and CheckingID in
case consultation with the CDC epidemiologist or data manager is needed regarding certain data problems.
Note: Following the NCHHSTP Data Security and Confidentiality Guidelines for HIV, Viral Hepatitis, STD, TB
programs: Standards to Facilitate Sharing and Use of Surveillance Data for Public Health Action, absolutely NO
State_city_no and CheckingID column should be sent in emails. Sending confidential data through email will result in
having to use SDN to send the resolution file back to CDC in the future.
Cleared by HICSB Management 10/15/2012
| File Type | application/pdf |
| Author | Karen Whitaker |
| File Modified | 2012-12-19 |
| File Created | 2012-10-17 |