CMS-10203_Supporting_Statement_Part_A_REVISED attachment 9

Supporting Statement – Part A

Supporting Statement For Paperwork Reduction Act Submissions
A. Background
The Centers for Medicare & Medicaid Services (CMS) has a responsibility to its Medicare
beneficiaries to require that care provided by managed care organizations under contract to CMS
is of high quality. One way of ensuring high quality care in Medicare Managed Care
Organizations (MCOs), or more commonly referred to as Medicare Advantage Organizations
(MAOs), is through the development of standardized, uniform performance measures to enable
CMS to gather the data needed to evaluate the care provided to Medicare beneficiaries.
Collected annually since 1998, the Medicare Health Outcome Survey (HOS) is the first outcomes
measure used in Medicare managed care. The goal of the Medicare HOS program is to gather
valid, reliable, clinically meaningful health status data in Medicare managed care for use in quality
improvement activities, plan accountability, public reporting, and improving health. All managed
care plans with Medicare Advantage (MA) contracts must participate.
The CMS, in collaboration with the National Committee for Quality Assurance (NCQA), launched
the Medicare HOS as part of the Effectiveness of Care component of the former Health Plan
Employer Data and Information Set, now known as the Healthcare Effectiveness Data and
Information Set (HEDIS®). This measure was initially titled Health of Seniors, and was renamed
the Medicare Health Outcomes Survey during the first year of implementation. This name change
was intended to reflect the inclusion of people with Medicare who are disabled and under age 65
in the sampling methodology.
The HOS measure was developed under the guidance of a Technical Expert Panel comprised of
individuals with specific expertise in the health care industry and outcomes measurement. The
measure includes the most recent advances in summarizing physical and mental health outcomes
results and appropriate risk adjustment techniques. In addition to health outcomes measures, the
HOS is used to collect the Management of Urinary Incontinence in Older Adults, Physical
Activity in Older Adults, Fall Risk Management, and Osteoporosis Testing in Older Women
HEDIS® measures
The Medicare HOS uses the Veterans 12-item Health Survey (VR-12) (a self-reported measure of
health status) to assess an MAO’s ability to maintain or improve the physical and mental health of
its Medicare members over time. Twelve hundred Medicare beneficiaries are randomly sampled
from each contract and surveyed every spring (i.e., a survey is administered to a new baseline
cohort, or group, each year). Two years later, these same respondents are surveyed again (i.e.,
follow-up measurement). For each member who completes a baseline and a follow-up survey, a
two-year change score is calculated and (taking risk adjustment factors into account) the member’s
physical and mental health status is categorized as better, the same or worse than expected.
1

(Members who are deceased at follow-up are included in the “worse” physical outcome category.)
Summary Medicare HOS results are calculated for each MAO based on aggregated member
outcomes.
Collected since 2005, the Medicare Health Outcomes Survey-Modified (HOS-M) is administered
to vulnerable Medicare beneficiaries at greatest risk for poor health outcomes. These beneficiaries
are enrolled in Program of All-Inclusive Care for the Elderly (PACE) programs. The HOS-M
instrument is a shorter, modified version of the Medicare HOS used by CMS to assess the frailty
of the population in these health plans in order to adjust annual capitated plan payments.
The chronology of all Medicare HOS OMB clearances is outlined below:
Beginning on December 24, 1997 the Medicare HOS was approved by OMB for collection
under HEDIS 3.0 (Health Plan Data and Information Set) and CAHPS (Consumer
Assessments of Health Plan Study) and Supporting Regulations 42 CFR 417.470. This
collection was cleared through December 31, 2000 under OMB number 0938-0701.
Due to a change in statutory authority as a result of the Balanced Budget Act of 1997,
HEDIS (Health Plan Data and Information Set) and CAHPS (Consumer Assessments of
Health Plan Study) and Supporting Regulations 42 CFR 417.470 and 42 CFR 417.126 was
submitted as a revised collection under OMB number 0938-0732 and was approved by
OMB on July 20, 1998 through January 31, 1999.
HEDIS (Health Plan Data and Information Set) and CAHPS (Consumer Assessments of
Health Plan Study) and Supporting Regulations 42 CFR 417.470 and 42 CFR 417.126 was
approved for extension by OMB on January 4, 1999 through January 31, 2002 under OMB
number 0938-0732.
Health Plan Employer Data and Information Set (HEDIS) and Health Outcomes Survey
(HOS) and Supporting Regulations at CFR 422.152 was submitted as a reinstated
collection requested under OMB number 0938-0701. The collection was approved on
March 29, 2002 through September 30, 2003.
Health Plan Employer Data and Information Set (HEDIS) and Health Outcomes Survey
(HOS) and Supporting Regulations as CFR 422.152 was approved by OMB on January 28,
2004 through January 31, 2007 under OMB number 0938-0701.
Medicare Health Outcomes Survey (HOS) and Supporting Regulations at 42 CFR 422.152
was approved by OMB on September 1, 2007 through August 31, 2010 under OMB
number 0938-0701.
Medicare Health Outcomes Survey (HOS) and Supporting Regulations at 42 CFR 422.152
was approved by OMB on February 1, 2011 through January 31, 2014 under OMB number
0938-0701.

2

B. Justification
1 . Need and Legal Basis
A. Need
The collection of Medicare HOS is necessary to hold Medicare managed care organizations
accountable for the quality of care they are delivering. This reporting requirement allows
CMS to obtain the information necessary for proper oversight of the Medicare Advantage
program. It is critical to CMS’ mission that the Agency collect and disseminate valid and
reliable information that can be used to improve quality of care through identification of
quality improvement opportunities, assist CMS in carrying out its oversight responsibilities,
and help beneficiaries make an informed choice among health plans.
B. Statutory and Regulatory Basis
Section 722(a)(3)(A)(i) of the Medicare Prescription Drug, Improvement, and Modernization
Act of 2003 mandates the collection, analysis and reporting of health outcomes information in
the Medicare Advantage program. This legislation also specifies that data collected on quality,
outcomes and beneficiary satisfaction to facilitate consumer choice and program
administration must utilize the types of data collected prior to November 1, 2003. This
provision was enacted by the agency at 42 CFR 422 Subpart D. Collected since 1998, the
Medicare HOS has collected outcome measures in the Medicare managed care setting and
remains a critical part of assessing health plan quality. In addition, CMS includes the
Medicare HOS results as part of the Plan Ratings and is included in the Quality bonus
Payment ratings for MA contracts.
These regulatory requirements are also contained within Chapter 5, Section 40 of the Medicare
Managed Care Manual.
2.

Information Users
The primary end users of HOS data are CMS, MAOs, and Quality Improvement
Organizations (QIOs). The data are used by CMS to monitor health plan performance, to
construct a frailty adjuster for payment purposes, and to inform other agency programs and
priorities (e.g. disabled, poor, chronically ill, etc.). MAOs and QIOs use HOS results to target
quality improvement activities and resources. Other stakeholders (i.e. other federal agencies,
advocacy groups, health policy scholars, and health services researchers) use HOS data to
monitor the health of the Medicare population and vulnerable subgroups, and to evaluate
treatment outcomes and procedures.

3.

Use of Information Technology
The Medicare HOS collects self-reported information through a combination of mail and
computer assisted telephone interviewing (CATI) techniques for survey administration.
3

Other than the manual reply necessary for mail survey instruments, there are no barriers or
obstacles that prohibit the use of improved technology for this information collection activity.
The Medicare HOS instrument is distributed to beneficiaries by independent thirty party
survey vendors and the resulting data are aggregated electronically. Beneficiaries complete
the survey either manually (accounting for roughly 84% at baseline and 90% at follow-up), as
this is the most cost effective means to collect information from them, or through CATI
(accounting for roughly 16% at baseline and 10% at follow-up). Mail surveys are processed
using scanned image readers to enhance coding accuracy and increase production speed. The
CATI program records collected information, which reduces respondent burden by
minimizing the potential for double reporting and inconsistent responses. CATI enables the
interviewer to move through skip patterns quickly, which reduces respondent burden by
shortening the interview and eliminating the need for call backs to correct errors. This
collection does not require a signature from the respondent.
4.

Duplication of Efforts
This information collection does not duplicate any other effort and the information cannot be
obtained from any other source.
The Medicare HOS is the first outcomes measure used in Medicare managed care. The
Medicare HOS measure was developed under the guidance of a Technical Expert Panel
comprised of individuals with specific expertise in the health care industry and outcomes
measurement. CMS and its data collection contractor have not been able to identify another
source of data, which would be an effective substitute for the Medicare HOS. Continuing
interagency collaboration insures against the likelihood of duplicative data collection
processes now and in the future.

5.

Small Businesses
The burden on small MCOs is reduced by requiring a standardized and commonly accepted
measure set in the managed care industry, with which MAOs can meet requirements of
Medicare and some private purchasers for reporting performance. In order to help control
costs, CMS only surveys a sample of beneficiaries from each Medicare managed care plan.
There is no way to further reduce the burden and still ensure the reliability of the information
collected.

6.

Less Frequent Collection
CMS collects the Medicare HOS data annually. Contracts between CMS and MCOs are
renewable on an annual basis, so the Agency needs these performance data for program
management and contracting decisions.

7.

Special Circumstances
Any publicly reported data that CMS makes available is aggregated and does not identify
4

beneficiaries in any way. For example, 2-year HOS performance measurement data are
included at the MAO-level in Medicare Plan Finder on the Medicare.gov web site. The
Medicare HOS individual level file is available only to requesters who, for confidentiality
reasons, must sign a Data Use Agreement with CMS and must meet CMS's data policies and
procedures that include, but are not limited to, submitting a research protocol and study
purpose. For information about Data Use Agreements, contact the Division of Privacy
Compliance, Enterprise Architecture and Strategy Group, within CMS's Office of Information
Services.
8.

Federal Register/Outside Consultation
A 60-day Federal Register Notice was published on April 27, 2012. A copy of the notice is
included as Attachment 1, and can be found on Page 25181 of the FR Notice. Since this
collection is not new, CMS has not gone out to solicit outside consultation; however, during
the 60-day Federal Register Notice the public is free to comment.

9.

Payments/Gifts to Respondents
There are no provisions to provide any payment/gift.

10. Confidentiality
All Contractors and HOS survey vendor staff directly involved in HOS data collection and/or
analysis activities are required to sign confidentiality agreements. Furthermore, all HOS
patient-level data are protected from public disclosure in accordance with the Privacy Act of
1974, as amended.
11. Sensitive Questions
CMS is not asking questions which would be considered of a sensitive nature. The Medicare
HOS (a functional status survey) does request information about one's capability to perform
certain physical and mental activities. However, the core component of the HOS instrument,
the VR-12, is a standardized instrument that has been used in both clinical practice and
research for a number of years. Demographic information in the survey, including income
and race, is requested for the purpose of risk adjustment.
12. Burden Estimates (Hours & Wages)
The following managed care organization types, with a minimum of 500 members, that had a
Medicare contract in effect on or before January 1 of the year prior are required by CMS to
administer the current baseline survey: (1) all coordinated care contractors, including health
maintenance organizations (HMOs), local preferred provider organizations (PPOs) and
regional PPOs; (2) private fee-for-service (PFFS) contracts; (3) medical savings account
(MSA) contracts; and (4) continuing 1876 cost contracts with open enrollment.
5

Organizations eligible to report also include MA contracts with exclusively special needs plan
benefit packages, regardless of institutional, chronically ill, or dual-eligible enrollment.
Furthermore, all MCOs with contracts in place on or before January 1 of three years prior,
and which administered a Baseline Survey two years prior are required to administer the
current Follow-Up Survey. In addition, all Program of All Inclusive Care for the Elderly
(PACE) organizations with a contract effective date of on or before January 1 of the previous
year are required by CMS to administer the HOS-M.
The HOS sampling strategy is designed to reduce burden on survey respondents. Because of
variations in health plan population size, three sampling approaches are used. (MAOs with
fewer than 500 members are exempt from HOS reporting.)
1) MAOs with populations of 500 to 1,200: all eligible members are included in the sample;
2) MAOs with populations of 1,201 to 2,999: 1,200 members are randomly selected for the
Baseline Survey. To ensure a sample size of 1,200, members who were sampled for and
returned a completed survey the previous year are not excluded from sampling in the
following year.
3) MAOs with populations ≥3,000: 1,200 members are randomly selected for the Baseline
Survey. To reduce burden on survey respondents, members who were sampled for and
returned a completed survey the previous year are excluded from sampling.
The HOS-M sample size is 1,200 for programs with at least 1,400 participants. For PACE
organizations with less than 1,400 enrollees, the entire eligible enrollment is included.
Table 12 A.

Sample Size

1

Estimated Response Rate
Number of Estimated Respondents
Survey Timing by Instrument
Time Burden in Hours
Number of Participating Plans 2
Total Time Burden in Hours
Average Hourly Wage Rate
Cost to Respondents

3

HOS

HOS

Baseline

Follow-Up

HOS-M

Total

1,200

840

1,200

70%

80%

70%

840

672

840

0.33

0.33

0.33

277.20

221.76

277.20

471

330

58

130,561

73,181

16,078

$ 21.35

$ 21.35

$ 21.35

$ 2,787,481.62

$ 1,562,410.08

$ 343,256.76

219,820
$ 4,693,148.46

1 The HOS Follow-up sample size is based on a 70% response rate at Baseline.
2 The number of participating plans is based on the 2010 participating plan list.
3 Based upon the average mean wages, “National Occupational Employment and Wage Estimates in the United
States", May 2010, U.S. Department of Labor, Bureau of Labor Statistics.

Tests have shown that the average time to complete the survey is 20 minutes. The average
6

response rate over the past five years is 63% at baseline, 82% at follow-up, and 75% for the
HOS-M; however, for the purpose of this clearance package we will use the following figures
so as to illustrate the greatest burden potential. The calculated response rate for HOS baseline
and HOS-M is 70% and the expected HOS follow-up response rate is 80%. Based on a plan
sample size of 1,200 and total plan participation in 2010, the total estimated burden upon the
beneficiaries is .33 hours x [(471 Baseline MCOs x 840 average baseline response rate) +
(330 Follow-up MCOs x 672 average follow-up response rate) + (58 PACE organizations x
840)] = 219,819.60 hours.
MCO’s will be contracting directly with an HOS-certified third party vendor to administer the
Medicare HOS. The MCOs will not experience an hourly burden, but will absorb the cost of
the CMS contracted third party administrator. The average cost borne by an MCO or PACE
organization for a completed survey is $20. Therefore, the estimated cost burden to an MCO
for HOS is: $20 x (840 baseline + 672 follow-up completed surveys) = $30,240. Similarly,
the estimated cost burden to a PACE organization for HOS-M is: $20 x 840 completed
surveys = $16,800.
13. Capital Costs
There are no capital costs.
14. Cost to Federal Government
There are costs to the Federal government in terms of its contracts with NCQA to administer
the Medicare HOS data collection and with HSAG to provide data file preparation, analysis,
and dissemination. Average contract costs for Medicare HOS activities are $2,415,000 per
year: $1,207,500 for each of two contracts, respectively. CMS personnel involved in
Medicare HOS include approximately 2.0 FTEs at the GS 13/14 level.
Grade
GS13 step 5
GS14 step 5
Travel
Total

FTE
1.5
0.5

2012 Annual Salary
$100,904
$119,238
$2,000

Cost to Government
$151,356
$59,619
$2,000
$154,034

15. Changes to Burden
The goal of the Medicare HOS is to gather valid, reliable and clinically meaningful data that
have many uses. With the introduction of the Five Star Rating system by the CMS, the
Medicare program now requires an ever increasing ability to provide accurate and
measureable, side-by-side comparisons between each MAO. Consequently, several questions
have been added to meet this need, while other questions have been removed to offset the
increase in burden.
Additionally, to meet the new requirements stipulated in Section 4302 of the Affordable Care
7

Act, the race, ethnicity, sex, language and disability questions outlined by the OMB have been
adopted. While in some instances this translates into rewording of existing questions and thus
does not impact the overall burden, there are instances in which the question is new to the
instrument and by adding it to the survey will marginally increase the time burden placed on
the respondent and on the cost for collection.
For a crosswalk of all survey instrument changes to include new questions being added,
current questions being replaced or removed see Attachment 4. In some instances we are
adding a question where a similar question will remain. An example of this is can be seen
with current question 10 (Because of a health or physical problem,…) and recommended
question 19 (Do you have difficulty dressing or bathing?). In this case we want to retain
question 10 while complying with Section 4302 requirements to include wording found in
question 19. It is important for the HOS to collect information at a more granular level than
what some Section 4302 questions permit. The Survey Notes column of Attachment 4 will
specify if the question recommended is simply being added or if it is replacing a current
question and if the current HOS question is being removed.
16. Publication/Tabulation Dates
Three major types of analyses are planned: descriptive, explanatory and predictive. A
number of published studies have already been conducted (Attachment 2). In addition, data
files will continue to be prepared over the course of the survey, see (Table 16 A).
Table 16 A. Schedule for information collection and dissemination
06/2012
07/2012
08/2012
09/2012
11/2012
04/01/2013
05/01/2013
06/2013
07/2013
08/2013
09/2013
11/2013
04/01/2014
05/01/2014
06/2014
07/2014
08/2014
09/2014
11/2014
04/01/2015

Disseminate Cohort 14 Baseline QIO and MA Report
Disseminate Cohort 14 Baseline QIO Data
Disseminate Cohort 12 QIO and MA Performance Measurement Report
Disseminate Cohort 12 Performance Measurement QIO and MA Data
Disseminate 2011 HOS-M PACE Report and Data
Data collection begun for Medicare HOS Cohort 16 Baseline and HOS-M
Data collection begun for Medicare HOS Cohort 14 Follow-Up
Disseminate Cohort 15 Baseline QIO and MA Report
Disseminate Cohort 15 Baseline QIO Data
Disseminate Cohort 13 Performance Measurement QIO and MA Report
Disseminate Cohort 13 Performance Measurement QIO and MA Data
Disseminate 2012 HOS-M PACE Report and Data
Data collection begun for Medicare HOS Cohort 17 Baseline and HOS-M
Data collection begun for Medicare HOS Cohort 15 Follow-Up
Disseminate Cohort 16 Baseline QIO and MA Report
Disseminate Cohort 16 Baseline QIO Data
Disseminate Cohort 14 Performance Measurement QIO and MA Report
Disseminate Cohort 14 Performance Measurement QIO and MA Data
Disseminate 2013 HOS-M PACE Report and Data
Data collection begun for Medicare HOS Cohort 18 Baseline and HOS-M
8

05/01/2015 Data collection begun for Medicare HOS Cohort 16 Follow-Up
B. Medicare HOS research data files
Several types of Medicare HOS data files are available for research purposes.
Medicare HOS data files are available as public use files (PUFs), limited data sets
(LDSs), and research identifiable files (RIFs). Medicare HOS PUFs contain the
majority of the survey items collected on the Medicare HOS instrument (excluding
beneficiary identifying information) as well as selected additional administrative
variables. Medicare HOS PUFs are constructed to prevent the identification of any
single beneficiary or plan and only respondents to the survey are included in the files.
Medicare HOS PUFs are available at no cost and can be downloaded directly from the
CMS website.
Medicare HOS LDSs and RIFs are comprised of the entire national sample for a given
cohort (including both respondents and non-respondents), and contain all of the
Medicare HOS survey items. The Medicare HOS LDSs include plan identifiers as well
as several additional variables describing plan characteristics. They also contain
protected beneficiary-level health information such as date of birth; however, specific
direct person identifiers (i.e. name and health insurance claim number) are not included
in LDSs.
The RIFs contain all of the variables in an LDS as well as specific direct person
identifiers (i.e. name and health insurance claim number) that are not included in an
LDS file. A signed Data Use Agreement with CMS is required to obtain either LDS or
RIF data files. (Table 16 B) summarizes data collection year and availability of
baseline, follow-up and analytic research files for the past three and next three cohorts.
Table 16 B: Data Collection and Availability
HOS Cohort
Baseline Data
Follow-Up Data
2006
2008
9
Summer 2007
Fall 2009
2007
2009
10
Summer 2008
Fall 2010
2008
2010
11
Summer 2009
Fall 2011
2009
2011
12
Summer 2010
Expected Fall 2012
2010
2012
13
Summer 2011
Expected Fall 2013
2011
2013
14
Expected Summer 2012 Expected Fall 2014

9

Analytic Data
2006-2008
Fall 2009
2007-2009
Fall 2010
2008-2010
Fall 2011
2009-2011
Expected Fall 2012
2010-2012
Expected Fall 2013
2011-2013
Expected Fall 2014

17. Expiration Date
CMS would like an exemption from displaying the expiration date as the collection of
Medicare HOS is an ongoing endeavor. Therefore, an expiration date is not practical.

10