OMB SUPPORTING STATEMENT
Year 2008 National Survey of Mental Health Treatment Facilities
A. Justification
1. Circumstances of Information Collection
The Substance Abuse and Mental Health Services Administration’s (SAMSHA) Center for Mental Health Services (CMHS) is requesting approval for a revision from the Office of Management and Budget (OMB) to conduct the 2008 National Survey of Mental Health Treatment Facilities (NSMHTF) (OMB No. 0930-0119), which expires on July 31, 2008. In previous survey cycles, this survey was entitled the Inventory of Mental Health Organizations (IMHO). The most recently completed survey was the 2004 IMHO.
In this request, SAMHSA is seeking OMB approval for a revised questionnaire, related correspondence, and information sheets. The details of the revised questionnaire are outlined in Section A.2.
The NSMHTF is authorized by Section 505(b) [42 USC 290aa--4] of the Public Health Service Act, which mandates the collection of statistical data on mental health programs and persons who receive care from them. It is also authorized by Section 520(a) and (b)(14) [42 USC 290bb--31], which establishes the Center for Mental Health Services and provides for the conduct of surveys with respect to mental health services.
The 2008 survey differs from earlier surveys by (a) shifting its focus from the mental health providers’ organization level to the providers’ point-of-service level, (b) introducing more client/patient-oriented questions, (c) including best practices questions, and (d) simplifying the questions to improve response rate. On average, each mental health organization has three service locations. It is this service location, or facility-level, that is the focus of the new survey. Changing the observation level from the organization to the facility more closely aligns this survey with SAMHSA’s National Survey of Substance Abuse Treatment Services (N-SSATS) (OMB No. 0930-0106). However, in order to provide continuity with the data collected in earlier IMHO surveys, it will still be possible to aggregate the data across facilities to produce organizational level data.
The NSMHTF has existed, in one form or another, for the last 38 years. From 1969 through 2004, the content and goals of the survey were very similar. In addition, DHHS collected similar information on mental health services from 1947 through 1968. Before this, the U.S. Bureau of the Census collected mental health services information from 1840 to 1946. This data collection effort, the longest continuous series in American public health, is the only mechanism for obtaining basic trend data on the type of mental health services provided, caseloads, bed counts, and admissions for specialty mental health facilities and for mental health services provided by general hospitals and Department of Veteran Affairs medical centers.
This survey also complements the Client/Patient Sample Survey (OMB 0930-0281), another survey sponsored by CMHS, which collects demographic, clinical, and service data on probability samples of client/patients in specific types of mental health facilities. The Client/Patient Sample Survey collects data on particular client/patients while the NSMHTF collects information at the facility level.
2. Purpose and Use of Information
The purpose of the survey continues to be the collection of information from mental health providers across the nation about the services they provide to persons suffering from mental problems and illnesses. The survey targets the specialty mental health industry sector. It purposely excludes individual and small group mental health practices. It also excludes military and institutionalized persons. For the 2008 NSMHTF, a clear effort is being made to understand the actual services provided and to collect information that is more helpful to the client/patients suffering or recovering from mental problems and illnesses. The selection of questions have changed from identifying structural components of the mental health facility (e.g., expenditures, staffing, and funding sources found in earlier IMHO questionnaires) towards a survey that targets the actual mental health services provided at a particular service location.
The survey will continue to be a 100 percent enumeration of all known specialty mental health facilities. The 2008 NSMHTF will use one questionnaire. The cross sectional data will provide a picture of services offered by mental health facilities to persons with mental conditions. The NSMHTF collects information on the facility’s mental health services, its caseload, accreditation, and service programs for specialized client/patient populations. The transmittal letter, instructions and inserts, and questionnaire for the 2008 NSMHTF are presented in Attachments 1.0 - 1.9.
CMHS will use the information from the 2008 NSMHTF to (a) update mental health care databases and directories for the United States, (b) provide a sampling frame of facilities for other surveys (such as the client/patient survey), (c) provide summary data for State and national health care reform efforts, and (d) study trends in the utilization and client/patient characteristics of mental health facilities. Data derived from the 2008 NSMHTF will be published by CMHS in Data Highlights, Mental Health, United States, the Mental Health Directory, and in professional journals such as Psychiatric Services and the American Journal of Psychiatry. In addition, SAMHSA will present the data in recurring agency publications, such as Mental Health: United States. Data from the NSMHTF will also be included in the National Center for Health Statistic’s (NCHS) Health: United States and in the Department of Commerce’s Statistical Abstract, which is published biennially. A current list of publications is contained in Attachment 2.0.
In addition to national tables, the NSMHTF will produce State-level tables that will be provided to each State Mental Health Agency (SMHA) as part of a longstanding, in-kind Federal-State partnership in mental health data collection and reporting. In conjunction with other surveys, results from the 2008 NSMHTF will be used to document selected characteristics of mental health services nationwide and by State.
Another Federal agency that uses these data is the National Institute of Mental Health (NIMH). It uses the name and address component to identify and mail questionnaires to mental health facilities in support of particular research studies. Other users include: (1) State governments, which utilize the information from the NSMHTF in their procedure for the licensing of facilities, in budgeting and planning, and in research; (2) mental health facilities themselves, which use the data in their program planning efforts; (3) evaluators supported by CMHS in State Governments, in universities, and in corporations; (4) the U.S. Congress, which requests periodic special reports on mental health services, such as services to elderly persons who have mental conditions; and, (5) the U.S. Bureau of the Census, which uses the names and addresses of facilities to help prepare the universe for the next decennial Census.
Finally, the general public uses the data via the Internet through the National Mental Health Information Center (NMHIC). The NMHIC, a service of CMHS and accessible from the CMHS home page, provides the public with information about mental health via a toll-free telephone number (800-789-2647), an Internet site, and more than 200 publications. The NMHIC was developed for client/patients of mental health services and their families, the general public, policy makers, providers, and the media. Using the NMHIC Internet site (http://mentalhealth.samhsa.gov/databases/), the public can search for mental health facilities by name, city, or State. The NMHIC is also used extensively by mental health professionals to locate referral locations for particular client/patients.
In addition, the Supporting Statement was reviewed by the Assistant Secretary for Planning and Evaluation (ASPE) and had no changes or comments.
3. Use of Information Technology
The 2008 NSMHTF will improve the appearance and usability of the respondent’s Internet data entry site (respondents will be offered the option of completing the questionnaire over the Internet) by taking advantage of improvements in Internet development tools since 2004. The quality of the Optical Character Recognition (OCR) software, used to scan mailed questionnaires, will also improve for the new survey. Respondents to the 2008 NSMHTF will have a choice of responding on the OCR questionnaires and mailing them, completing the questionnaires and faxing them, or responding entirely on the Internet.
To alleviate response burden, data will also be collected directly from the State mental health databases, which stores data already submitted by the mental health facility (usually to document the use of State grant funds) to the State. CMHS expects that data for approximately 5 percent of the facilities will come directly from States that maintain their own central databases. For these facilities, there will be no direct response burden from this survey since all of the information is collected centrally.
Respondents who choose to use the Internet will have more functionality than those completing the paper questionnaires. For example, Internet users will be able to (a) make multiple changes to their questionnaires before formally submitting their data, (b) produce lists of data entry exceptions that will help them to improve the quality of their responses, and (c) have direct access to some of the information from the previous survey (e.g., name and address information will be pre-filled). It is expected that 25 percent of all respondents will choose to use the Internet.
The 2008 NSMHTF will use the latest tools for the data collection applications. These include:
Microsoft SQL Server 2005 database to store the primary data tables
Visual Studio .Net 2005 for creation of the Internet pages
Blaise to create computer assisted telephone interviews
The SAS 9 for database management, imputations, summary statistics, cross tabulations
The latest Enterprise version (v10) of Teleform for the production of the questionnaires and OCR processing of responses
TPL Tables 6.0 to produce final Postscript-quality tables
The 2008 NSMHTF, will also, for the first time, use computer-assisted telephone interviews (CATI) to collect data from non-responders. The most current Blaise software will be used to collect missing responses (either missing responses for returned questionnaires or the collection of the entire questionnaire for non-respondents). In addition, data will be collected using Blaise to support the scheduling, monitoring, and documentation of all telephone calls made to the respondent.
Finally, the 2008 NSMHTF will use interviewers in a telephone research center to prompt, remind, and encourage the respondents to complete the questionnaire. The telephone research center’s capability of quickly increasing or decreasing the number of callers will substantially improve the efficiency of effort employed in the data collection process.
4. Efforts to Identify Duplication
The information already available on mental health facilities, from other data collection efforts, cannot be used because the scope of coverage is limited and available data typically are outdated and not standardized across types of facilities. For example, the American Hospital Association (AHA) collects limited information on psychiatric hospitals in its annual survey of hospitals. However, neither the scope of coverage nor the data collected is in the detail required by CMHS to facilitate mental health planning, evaluation, congressional reports and other applications. No detailed data are collected by the AHA on ambulatory psychiatric services or on the availability of services in separate psychiatric inpatient units of non-Federal general hospitals, the nature of these services, caseloads, and client/patient characteristics. With regard to coverage, the AHA survey excludes many State, county, and private psychiatric hospitals, which are neither AHA members nor registered by the AHA. Although some basic information (e.g., provision of psychiatric services) is collected on the AHA survey, the NSMHTF will collect more detailed and standardized information regarding hospitals with separate psychiatric inpatient units in comparison with other types of mental health facilities.
There is no other national organization or Federal agency which collects standardized information on mental health services across particular mental health facility types such as outpatient community mental health clinics, residential treatment centers for severely emotionally disturbed children, and multi-setting community mental health facilities.
The Office of Applied Studies (OAS), SAMHSA, collects data on substance abuse treatment services (the N-SSATS). These data are restricted to substance abuse services provided by substance abuse or mental health facilities. The 2008 NSMHTF will complement and not duplicate the data collected by OAS since it will collect information only for mental health services.
The National Center for Health Statistics (NCHS) conducts a continuing survey of patient discharges, the National Hospital Discharge Survey (NHDS) (OMB No. 0920-0212), based on inpatient data from short‑stay hospitals as well as the National Hospital Ambulatory Medical Care Survey (NHAMCS) (OMB No. 0920-0278) that surveys hospital-based ambulatory services. However, neither of these surveys provides information at the facility level on mental health services and the utilization of those services. In addition, NCHS collects information on patients seen by physicians, including psychiatrists, in private practice in its National Ambulatory Medical Care Survey (OMB No. 0920-0234). In contrast to NCHS surveys, the NSMHTF does not survey treatment providers in private practice (e.g., psychiatrists and clinical psychologists). For these reasons, CMHS routinely shares data from the NSMHTF with NCHS.
Other databases, such as the Verispan (formerly known as SMG) Hospital Market Profiling Solution Database, do contain some facility-specific data for general hospitals. This file will be used in this survey to develop the universe of general hospitals with separate psychiatric inpatient units. In addition, the most recently completed AHA Annual Hospital Survey Database will be used to supplement the universe for the NSMHTF.
In conclusion, the proposed 2008 NSMHTF does not duplicate other data collection efforts and is the only national survey which will provide data for precise estimates on specialty mental health facilities and general hospital mental health services necessary to fulfill the minimal data needs of Federal, State, and local agencies, as well as private sector institutions.
5. Involvement of Small Entities
The 2008 NSMHTF involves small entities. The following methods will be used to minimize reporting burden for small entities in particular and for all respondents in general:
(a) For some State-operated and/or funded facilities, information being requested is already available from the State Mental Health Agency’s central electronic management information system and will be forwarded on a voluntary basis to CMHS in the form of an electronic file, thus eliminating the need for data collection from individual mental health facilities.
(b) The use of OCR and Internet technologies will increase the options available to the respondents (e.g., mail, fax, or Internet) and will decrease the time between data collection and error resolution (the Internet version will automatically check each response).
(c) If the facility was listed by an organization that responded to a previous organizational level survey, the name and address information will be included on the cover of the questionnaire (see Attachment 1.9). The facility will be asked only to correct and update the items on the cover rather than completing these items on a blank questionnaire. In addition, in the Internet version, name and address information will be pre-filled on the Internet page.
(d) The facility is requested to report only on components that it operates directly and not on services that it purchases from other facilities through contracts or agreements.
(e) All of the instructions for each question are included with the question rather than on a separate instruction page. This saves the respondent the time and trouble of turning pages between the questionnaire and an accompanying instruction manual.
(f) Contractor staff will be available to answer, via an 800 line, any questions that respondents may have regarding the 2008 NSMHTF.
6. Consequences If Information Collected Less Frequently
If the information is not collected, Federal program and policy activities will suffer in several ways. Data derived from the NSMHTF are used by CMHS for essential planning for State and national health care reform efforts and in support of policy formulation for Federal mental health programs. In addition, the information collected provides sampling frame data to conduct surveys of mental health providers and client/patients.
The CMHS surveys have a tradition of being fielded once every two to three years. An exception to this two-year series was in 1996 when it was not possible to issue a contract for the survey. In addition, because of the substantial changes made to this survey, there will be a four-year gap (2004 to 2008) instead of the normal two to three years. Future NSMHTF are planned to return to the two-year frequency of being fielded. If the data collection were conducted less frequently, essential cross sectional data, as well as trend data, would not be available for planning and research at the Federal, State, and local levels. Because of the rapid pace of change in the mental health service delivery system, a two year time frame is the minimum needed to document these changes. In addition, CMHS requires the 2008 NSMHTF to develop the sampling frame and sample weights for the CMHS Client/Patient Sample Survey Program, which typically occurs in years between the facility-level surveys. Finally, the 2008 NSMHTF will provide up-to-date information for the public utilizing the online services offered by the National Mental Health Information Center, including, the mental health services locator.
7. Consistency with the Guidelines in 5 CFR 1320.5(d)(2)
This data collection complies fully with 5 CFR 1320.5(d)(2).
8. Consultation Outside the Agency
A notice soliciting public comment on this data collection was published in the Federal Register on June 8, 2007(Volume 72, number 110, pages 31843-31844). No comments were received in response to this notice.
A draft of this OMB submission was sent to the following Federal, State and local professionals for their review (see Attachment 3.0 for a copy of the solicitation letter):
Federal Personnel:
Linda F. McCaig, MPH, Health Scientist, Division of Health Care Statistics, National Center for Health Statistics, (301) 458-4365.
Lyman Van Nostrand, Director, Office of Planning and Evaluation, HRSA,
(301) 443-1891.
Ira R. Katz, MD, Deputy Chief Patient Care Services Officer for Mental Health, Department of Veterans Affairs, (202) 273-8434.
Associations:
Ted Lutterman, National Association of State Mental Health Program Directors Research Institute, (703) 739-9333 extension 121.
Joy Midman, Director, National Association for Children’s Behavioral Health, (202) 857-9735.
State Mental Health Agencies:
Ms. Donna Hillman, Director, Division of Mental Health and Substance Abuse, Kentucky Department of Mental Health & Mental Retardation, (502) 564-4456.
James S. Reinhard, MD, Commissioner, Department of Mental Health, Virginia Mental Retardation & Substance Abuse Services, (804) 786-3921.
Mental Health Facilities:
Ms. Ann Robinson, Beacon Center, CO, (303) 761-6756.
Mr. Robert Tovey, Cascade Child Treatment Center, OR, (541) 548-6166.
Responses were obtained from NCHS, the VA, and both the Virginia and Kentucky State Mental Health Agencies. A summary of their comments is contained in Attachment 3.1.
9. Payments to Respondents
No payment or gifts are provided to respondents to participate.
10. Assurance of Confidentiality
No assurance of confidentiality is pledged to the respondents because the questions pertain to their professional role as a representative of the facility with knowledge of this information.
11. Questions of a Sensitive Nature
There are no questions of a sensitive nature in the 2008 NSMHTF.
12. Estimates of Annualized Hour Burden
The response burden and hourly wage were estimated by (a) using the current burden of 40 minutes for the 2007 N-SSATS and (b) consulting with the nine-member pilot group described in Attachments 6.0-6.3. In addition, the OMB Review Group assessed the validity of the response burden and the $40 hourly cost estimate.
Estimates of the response burden are shown in Table 1. The estimated time for response to the questionnaire is one hour. Each facility will be asked to respond to one questionnaire. The NSMHTF will involve 13,381 facilities. These 13,381 facilities are administered by approximately 5,120 mental health organizations. The total national hourly cost will be approximately $553,240 using $40 as the median wage plus fringe hourly estimate. Unlike the previous IMHO surveys, respondent burden variation across facility is expected to be small given the change in the type of data to be collected. However, when seen from an organizational perspective, mental health organizations with many facilities will have a higher additive response burden.
13. Estimates of Annualized Cost Burden to Respondents
There are no capital, startup costs, operations, or maintenance costs to respondents associated with this project.
14. Estimates of Annualized Cost to the Government
The 2008 NSMHTF will be executed under a modified four-year contract for which the total cost is $2,447,928. Years one and two concentrated on frame and questionnaire development. The third year will focus on data collection followed by a fourth year of effort directed toward data cleaning, summarizing, and publishing. The estimated cost for Federal staff time is $40,000, which represents one-tenth time annually for a government project officer over the four years of the contract. Thus, the total government cost estimate is $2,487,928 with an annualized cost of $621,982.
Table 1: Response Burden for the 2008 National Survey of Mental Health Treatment Facilities |
||||||
|
(1) |
(2) |
(3) |
(4) |
(5) |
(6) |
Facility Type |
Number of Respondents |
Average Hours per Response |
Responses Per Respondent |
Total Hour Burden |
Hourly Wage Plus Fringe ($) |
Total Hourly Cost ($) |
Public Psychiatric Hospital |
502 |
1 |
1 |
502 |
$40 |
$20,080 |
Private Psychiatric Hospital |
557 |
1 |
1 |
557 |
$40 |
$22,280
|
General Hospitals with Separate Psychiatric Services |
1,599 |
1 |
1 |
1,599 |
$40 |
$63,960 |
Department of Veterans Affairs Medical Centers |
150 |
1 |
1 |
150 |
$40 |
$6,000 |
Department of Veterans Affairs Community-Based Outpatient Clinics |
810 |
1 |
1 |
810 |
$40 |
$32,400 |
Residential Treatment Centers for Sev. Emot. Dist. Children |
1,456 |
1 |
1 |
1,456 |
$40 |
$58,240 |
Outpatient Clinics (including hospital based) |
3,493 |
1 |
1 |
3,493 |
$40 |
$139,720 |
Multi-Setting Community Facilities |
5,264 |
1 |
1 |
5,264 |
$40 |
$210,560 |
Total |
13,831 |
|
|
13,831 |
|
$553,240 |
3-Year Average |
4,610 |
|
|
4,610 |
|
$184,413 |
15. Changes in Burden
Although the 2008 NSMHTF is now at the facility level (13,831 facilities compared with 4,160 organizations), the total hour burden is 632 hours less than the 14,463 hours estimated for the 2004 IMHO. The decrease is due to the reduction in the complexity of questions used in the organization level survey and the use of a single questionnaire with easier questions that require less effort (resources) to answer.
16. Time Schedule, Publication and Analysis Plans
Table 2 shows the time schedule for the 2008 NSMHTF. The survey mailing is scheduled for one month after OMB approval. The generation of tables from the 2008 NSMHTF Survey is scheduled for 14 months following OMB approval.
Table 2: Proposed Schedule
Activity Planned Start Time
Printing of Questionnaires, Letters, and Instructions |
OMB Approval |
Mailing |
OMB Approval + 1 month |
Data Collection |
OMB Approval + 2 month |
Non-Response Prompting Calls |
OMB Approval + 3 months |
Core and Missing Data CATI |
OMB Approval + 4 months |
Anomaly Production |
OMB Approval + 6 months |
Data Collection Cutoff Date |
OMB Approval + 7 months |
Production of Un-imputed File |
OMB Approval + 10 months |
Production of Imputed File |
OMB Approval + 11 months |
Table Production |
OMB Approval + 14 months |
Final File Delivery |
OMB Approval + 18 months |
Submission of Project Documentation |
OMB Approval + 19 months |
The planned data analysis is designed for the following purposes:
To develop national tables for use in CMHS publications including Mental Health, United States, and Data Highlights, and to answer information requests.
To develop State level tables, including national totals, for distribution to the States.
To provide detailed facility level utilization data by service settings (e.g., inpatient, residential and outpatient providers) at both the national and State level to answer information requests and for use in CMHS publications.
Tables for Mental Health US will include all mental health organizations in the US and the District of Columbia aggregated from the facility to the organization level. Tables include:
Number of mental health organizations, by type of organization: United States, selected years, 1970-2008
Number, percent distribution, and rate of 24-hour hospital and residential treatment beds, by type of mental health organization: United States, selected years, 1970-2008
Number, percent distribution, and rate of 24-hour hospital and residential treatment admissions, by type of mental health organization: United States, selected years, 1969-2008
Number, percent distribution, and rate of less than 24-hour care admissions, by type of mental health organization: United States, selected years, 1969-2008
An example of a Mental Health US Table is presented in Attachment 4.0.
State Tables 2008 will include all mental health organizations in the US and the District of Columbia aggregated from the facility to the organization level. Specific tables will include:
Number of Mental Health Organizations by State According to Organizational Type: 2008 NSMHTF
Patient Census in 24-hr Hospital or Residential Care Settings By State, According to Patient Census Indicators: 2008 NSMHTF
Number and Rate per 100,000 Population in a 24-Hr Hospital or Residential Setting By State, and According to Caseload Statistics: 2008 NSMHTF
An example of a table from State Tables 2004 is presented in Attachment 5.0.
Tables for Mental Health Facilities: 2008 will include all mental health facilities in the US and the District of Columbia at the facility level. These tables will be new since this will be the first time that facility data, instead of organization data, are collected. Tables to be created include:
Primary Focus (MH or SA) of Mental Health Facilities in the US and the District of Columbia [see question 3 in Attachment 1.9]
Mental Health Services in the US and the District of Columbia [question 6]
Treatment Approaches by Facility Type for the US and the District of Columbia [question 7]
Facility Ownership [question 19]
Facility Licensing, Certification, or Accreditation by State [question 26]
Number of Inpatient Beds by State and the District of Columbia [question 29c]
The State Tables and the new tables from the 2008 NSMHTF will be distributed to the States and made available to the public via the National Mental Health Information Center.
17. Display of Expiration Date
An exemption for the requirement to display the expiration date is not requested.
18. Exceptions to Certification Statement
This collection of information involves no exceptions to the Certification for Paperwork Reduction Act Submissions.
Collection of Information Employing Statistical Methods
1. Respondent Sampling Frame and Sampling Methods
A full enumeration of all mental health facilities is planned; therefore, no formal sampling plan is needed. The estimated number of mental health facilities by type is presented in columns 1 and 2 of Table 1. A total of 13,381 mental health treatment facilities will be surveyed.
Historically, response rates for the IMHO have been in the 80 to 90 percent range. However, since 1998, the response rates declined to the 70 to 80 percent range. Because the 2008 NSMHTF questionnaire contains substantially fewer questions than the past organization level surveys and has more easily answered questions, the response rate is expected to increase back to the 90 percent level.
Information Collection Procedures
The survey will include all 50 States, the District of Columbia, and all US Territories (for a total of 50 States, one District, and eight Territories). The 2008 survey will collect information at the facility, or point-of-service level, while, at the same time, retaining the relationship of the service location to the organizational level. The survey defines a mental health facility as one that (a) has a stated mission to ameliorate the effects of mental conditions, (b) has staff specially trained to treat persons with mental problems and illnesses, (c) has a known caseload of client/patients, (d) provides services specifically designed to help persons suffering from mental conditions and (e) has a clear organizational structure. This survey collects information from organized facilities and purposely excludes private or small group practices. The survey collects information only on the civilian population and excludes military client/patients, those served by the Indian Health Service, and those residing in a local, State, or national correctional facility.
Before the questionnaire is fielded, contact will be made with the State Mental Health Agencies to determine their level of involvement in the data collection. This involvement may take the form of actual completion of some questionnaires or only furnishing a letter of endorsement.
There will be one mailing for the NSMHTF. One week after the mailing of the survey packets, calls will be made to the respondents to make sure that they have received the packet and are aware of the deadlines. After the deadline has passed, we will contact the non-respondents and encourage them to complete and return the questionnaire. For respondents who still have not returned the questionnaire, a computer-assisted telephone interview (CATI) will be conducted to collect core data questions (Sections A and B of the questionnaire in Attachment 1.9).
Mailed questionnaires will be processed using Enterprise v10 of Teleform (the most current version of the software). The questionnaires will be scanned, read by the Teleform Reader, and verified by staff members. For faxes, the questionnaires will be sent directly to the Teleform Reader. Responses via fax and mail will then be added to the master database following staff visual verification. For Internet responses, the data will already be saved in the master database.
All questionnaires, irrespective of the response mode, will be reviewed for consistency. This means that questionnaires returned in the mail, questionnaires that are faxed, electronic databases for States, and questionnaire completed using the Internet application will be checked to ensure consistency and to check for item non-response. Questionnaires with inconsistencies or missing responses will be included in the CATI and the facility called to gather the missing information or to resolve the inconsistency. It is hoped that this will improve question response rates. For facilities that failed to return a questionnaire, we will use a CATI to collect Sections A and B of the questionnaire.
For responses that are still missing after these data collection methods have been concluded, we use one of the imputation techniques described below. For item non-response (i.e., they had an incomplete response), if data cannot be obtained from alternative sources, then the values will also be obtained through imputation procedures. The main imputation method will be "hot decking" from a stratified data base. Hot decking is the process of "borrowing" data for a non-responding facility from one that did respond. When possible, the known characteristics of the non-responding facility will be matched to the donor facility characteristics. This procedure will be used only when data are not available for a non-respondent facility from a returned questionnaire or core data collection. Prior to hot decking, stratification will be done on three characteristics: (a) region of the U.S., (b) facility type/ownership, and (c) size. Other imputation methodologies will include multiple linear regression and use of State population proportions.
3. Methods to Maximize Response Rates
The first method used to maximize the response rate will be to handle undelivered mail. On the mailing envelope, we will request a Return Address Service. This request instructs the United States Postal Service to return undeliverable mail with a sticker showing the correct address. We will then re-mail the packet. For other returned mail, the contractor will attempt to determine the correct address (using the telephone and the Internet) and if found, update the database and send the packet to the right address.
A second method is to simplify the questions. In previous IMHO surveys, questions such as those asking for the breakdown of expenditures and staffing within multiple categories, were difficult to answer and therefore resulted in a low item response rate. These “difficult questions” have been dropped from the 2008 NSMHTF (see Attachment 1.9).
A third method will be the use of business reply mail (BRM) envelopes, which will be included in the packet so that the respondent will be able to easily return the questionnaire without paying for postage. A fax number will be included in the instructions so that the respondents may use that medium for return of the questionnaires if they wish. In addition, the questionnaires will be incorporated into an Internet site so that facilities may also use that mode of response. Therefore, the respondent has three modes of responding to the 2008 NSMHTF: (1) mail; (2) fax; or (3) Internet. Increasing the modes of return is intended to help improve the response rate.
A fourth method is the involvement of each State Mental Health Agency (SMHA) in the survey. These State contacts coordinate with CMHS to determine the best data collection plan for their own State. The State contacts may provide an endorsement letter from the State mental health director (over 90 percent provided letters for the 2004 IMHO), decide to provide data from their central database, or help in non-response call follow-up. The development of data collection plans can help the response rate by (a) exempting data that is already in the central database, (b) providing credibility for the survey via local endorsements, and (c) increasing motivation for non-responders. Survey information for an estimated five percent of facilities, will be collected through State mental health agencies. Collection of these data from central databases eliminates the need to collect information directly from the facilities.
A fifth method will be a computer-assisted telephone interview (CATI) conducted for every facility that did not return a questionnaire and for questionnaires with missing or inconsistent data. For non-responders, the CATI will include only the core questions (all questions in Sections A and B in Attachment 1.9). The software package Blaise will be used to create the CATI, and the calls will be made from a telephone research center. It is expected that all facilities that do not complete a questionnaire will complete the CATI.
4. Tests of Procedures
The similarity between the questionnaire used by N-SSATS and the one proposed for the 2008 NSMHTF is substantial. Since the N-SSATS questionnaire has been successfully fielded for many years with an average 95 percent response rate, it is reasonable to assume that NSMHTF questionnaire will also be successful. However, because this is the first time that the questionnaire will be completed by staff at the mental health facility level, a nine-member pilot was conducted. Attachment 6.0-6.3 describes the pilot administration, the cover letter, the questionnaire feedback form, and a summary of the pilot study’s results.
5. Statistical Consultants
The primary contractor, Social & Scientific Systems, Inc. (SSS), will have overall responsibility for implementation and execution for the project and for the preparation of all tables. The subcontractors will help in frame development (NRI) and statistical imputations methods (Westat).
The primary contractor for this project is:
Social & Scientific Systems, Inc.
James Maedke, Project Manager
8757 Georgia Avenue, Suite 1200
Silver Spring, Maryland 20910
(301) 628-3252
The subcontractor working with State mental health representatives is:
NASMHPD Research Institute
Mr. Ted Lutterman
66 Canal Center Plaza, Suite 302
Alexandria, VA 22314
(703) 739-9333 ext. 121
The subcontractor for statistical imputation methods is:
Westat
David Morganstein, PhD
1650 Research Blvd
Rockville, MD 20850
(301) 251-8215
Daniel J. Foley, M.S., Epidemiologist, Survey, Analysis and Financing Branch, DSCSD, CMHS, Room #2-1100, 1 Choke Cherry Road, Rockville, MD 20857, telephone (240) 276-1766, email [email protected], is the Project Officer for the Federal government.
List of Attachments
Attachment 1.0 – CMHS Cover Letter
Attachment 1.1 – Criteria Used to Determine Facility Eligibility
Attachment 1.2 – Service Descriptions
Attachment 1.3 – Mental Health Therapeutic Approaches
Attachment 1.4 – Supportive Services
Attachment 1.5 – Seclusion and Restraint Definitions
Attachment 1.6 – Mental Health Facilities Type Definitions
Attachment 1.7 – Service Settings
Attachment 1.8 – Response Modes
Attachment 1.9 – Questionnaire
Attachment 2.0 – Current List of Publications
Attachment 3.0 – Solicitation Letter for Outside Reviewers
Attachment 3.1 – Summary of Outside Reviewer’s Comments
Attachment 4.0 – Sample Mental Health US Table
Attachment 5.0 – Sample State Table
Attachment 6.0 – Pilot Design
Attachment 6.1 – Pilot Cover Letter
Attachment 6.2 – Pilot Feedback Form
Attachment 6.3 – Summary of Pilot Result
| File Type | application/msword |
| File Title | OMB SUPPORTING STATEMENT |
| Author | JMaedke |
| Last Modified By | James Maedke |
| File Modified | 2007-12-24 |
| File Created | 2007-12-24 |