ICR Supporting Statement_Section A_7-12

ECONOMIC RESEARCH SERVICE OMB CLEARANCE PACKAGE

SECTION A. JUSTIFICATION

for

CLEARANCE TO CONDUCT THE SURVEY ON RURAL COMMUNITY WEALTH AND HEALTH CARE PROVISION

FROM FY2012 THROUGH FY2014

Prepared by

Farm and Rural Household Well-Being Branch

Resource and Rural Economics Division

Economic Research Service

U.S. Department of Agriculture

July 2012

TABLE OF CONTENTS

1. Need and authority for the Information Collection ……………………………………. 1

2. Use of the information to be collected ……………………………………………….... 4

3. Use of information technology/other advanced technology in information collection … 5

4. Efforts to identify and avoid duplication ……………………………………………… 6

5. Economic impacts on small businesses or other small entities ………………………… 9

6. Consequence to Federal program and policy activities if collection not conducted …... 9

7. Special circumstances ………………………………………………………………….. 9

8. Federal Register notice ……………………………………………………………….... 9

9. Incentives for respondents ……………………………………………………………. 11

10. Assurance of confidentiality ………………………………………………………….. 11

11. Sensitive questions…………………………………………………………………….. 12

12. Estimated hour burden on respondents ……………………………………………….. 12

13. Estimated annual cost burden on respondents ………………………………………... 18

14. Estimated annualized costs to the Federal Government ………………………………. 18

15. Reasons for program changes or adjustments …………………………………………. 21

16. Plans for tabulating and publishing survey results …………………………………….. 21

17. Reasons to not display the expiration date for OMB approval …………………………23

18. Exceptions to the certification statement ……………………………………………….24

Annex A. References

Annex B. Experts consulted in the design of the survey

Annex C. Letter from Thomas Morris, Associate Administrator for Rural Health Policy, HRSA, HHS

Annex D. Public comments on Notice of Intent to Request New Information Collection

Annex E. Report on cognitive interviews

Annex F. Health care provider questionnaire

Annex G. Community leaders/other stakeholders questionnaire

Annex H. Advance letter to respondents with disclosures and project brochure

Annex J. Script for sample development

1. Need and authority for the Information Collection

This survey will collect information on the assets and investments of rural communities and their influence on recruitment and retention of rural health care providers, and on the effects of rural health care provision on economic development of rural communities. This information will contribute to a better understanding of the roles that rural communities play in promoting or retarding the provision of health care services, and of how improved health care provision contributes to development of these communities. Such understanding is critical to develop effective policies to address the challenge of inadequate access to health care services in many rural communities, and to realize the opportunities offered by improved health care provision to attract and keep residents in rural areas, provide employment, and improve the quality of life.

Health care services is one of the largest and most rapidly growing industries in rural America, and adequate provision of health care services is critical for achieving economic development and improved well-being of rural people. In many rural communities, the health care services sector is the largest employer, and rapid growth in this sector is occurring and likely will continue, especially as the Baby-Boom generation retires.¹ Provision of adequate health care services may be a key factor attracting retirees and other migrants to rural areas, contributing to rural growth and prosperity.²

Despite recent growth and potential for continued growth in this sector, many rural communities suffer from poor access to health care services, especially because of the limited supply of health care professionals. For example, 85% of nonmetropolitan counties were entirely or partly classified as a primary care Health Professional Shortage Area (HPSA) at some time during 1996 to 2004, with 34% of nonmetropolitan counties classified as whole county persistent primary care HPSAs during this period.³ People living in these areas suffer from fewer primary care physicians and other health care providers, a smaller share of the population with access to a regular primary care provider, and a larger share of the population who do not receive needed care due to costs. Access to quality health services is the top rural health priority identified by state and local rural health leaders across the nation.⁴ Addressing these access problems likely will become increasingly important as the Patient Protection and Affordable Care Act is implemented.

Although substantial research has investigated the problems of attracting and retaining health care providers (especially physicians) in rural areas, very little of this research addresses the issue from the perspective of rural communities themselves. For example, a large number of studies have investigated the influence of “nature vs. nurture” on physicians’ decision to practice in a rural location; i.e., factors in physicians’ backgrounds, such as growing up in a rural area, demographic characteristics, or professional motivations and intentions vs. the location and nature of the medical school they attended, their residency experience, or Federal and State programs providing incentives to practice in rural areas.⁵ Much less research has focused on the factors affecting location choices by non-physician rural health care providers,⁶ or on the economic development impacts of rural health care provision in rural areas.⁷

Several large quantitative studies have investigated the effects of some community-level factors such as population level, proximity to an urban center, presence of a hospital or college, and median income and housing values on the practice location choice of physicians using available county-level data.⁸ However, these studies did not investigate the perspectives of community members or health care providers regarding the factors affecting recruitment and retention of providers in their communities, including less readily quantified assets such as community social capital or local efforts to integrate providers into the community. Some qualitative or small quantitative studies have investigated these issues in some depth, but only in a few communities with a small number of respondents, limiting the ability to draw conclusions applicable to broader rural regions.⁹ A few studies describe program approaches to link community development and provider recruitment efforts, but these lack empirical evidence regarding the factors affecting the success of these approaches.¹⁰

The proposed survey will address gaps in existing knowledge about the relationships between community development and rural health care provision by investigating these issues from the perspective of members of rural communities, including health care providers, community leaders and other stakeholders involved in recruiting and retaining health care providers. In contrast to the small number of communities included in the handful of previous studies, this study will be based on a sample of 150 communities in three major regions of the country representing nearly one-tenth of the U.S. rural population.

The survey will be conducted under the authority of 7 U.S.C. 2204(a), 7 U.S.C. 2204(b), and 7 U.S.C. 2661.

2. Use of the information to be collected

The information collected will be used for socioeconomic research on the influence of rural communities’ assets and investments on recruitment and retention of rural health care providers, and on the effects of rural health care provision on economic development of rural communities. This research will be led by the Economic Research Service (ERS) of the U.S. Department of Agriculture (USDA). The survey will collect information from individuals, including rural health care providers and community leaders, in 150 rural communities. The information will be collected via a telephone survey, with each respondent asked to provide information only once. Analysis of the survey information and related secondary information will be used to produce research reports and other peer reviewed publications on this research topic.¹¹ The data collected will be made publicly available to other research organizations using procedures to protect the strict confidentiality of the survey respondents, in accordance with 7 U.S.C. 2276, and with OMB Implementation Guidance, “Implementation Guidance for Title V of the E-Government Act, Confidential Information Protection and Statistical Efficiency Act of 2002 (CIPSEA)”, 72 FR 33362, June 15, 2007. As one of the statistical agencies of the Federal Government, ERS has established procedures for sharing survey data consistent with these laws and regulations, which will be followed in sharing the data collected by this survey.

3. Use of information technology/other advanced technology in information collection

The survey will be conducted by telephone by the Survey and Behavioral Research Services, Iowa State University. This method of survey was selected over alternatives based on consideration of cost (much higher for in-person interviews), response rate (higher for telephone interviews than mail or online surveys given the topic and sample characteristics), and quality control (better ability to monitor enumerators with a telephone survey than with in-person interviews; better ability to identify and answer respondents’ questions and provide clarifications than with a mail or online survey). The survey questions will be programmed using Blaise software for computer-assisted administration. Computer-assisted interviewing reduces interview time, assures proper skip sequencing and enables immediate range and consistency checks. All interviews will be conducted from the SBRS Computer-Assisted Telephone Interviewing (CATI) lab under the supervision of trained professional staff. A minimum of 10% of the interviews will be monitored for quality control using a silent audio/visual monitoring software system.

Information technology and publicly available data sources are being used to collect secondary information on each study community and to identify the population of potential survey respondents. Web searches of community and local health system websites will be conducted to locate names and contact information for community leaders and health care professionals. Publicly available information for the sample frame will be collected from local organizations such as the Chamber of Commerce, economic development groups, local government offices, and health care provider institutions/agencies. State websites, such as the Department of Public Health, will be reviewed for relevant statistical data. National data sources such as the U.S. Census Bureau and the Dartmouth Health Atlas will be accessed to obtain both past and recent data.

4. Efforts to identify and avoid duplication

A review of relevant literature, publicly available data sources, and consultations with experts on rural health care and rural development issues was conducted and found limited research and data on the specific topic of this information collection. A list of the experts consulted is included in Annex B. Among the experts consulted were Thomas Morris, Associate Administrator for Rural Health Policy, Health Resources and Services Administration, U.S. Department of Health and Human Services (HHS), and several of his senior staff. As indicated in a letter from Thomas Morris to John Pender, dated March 13, 2012 (Annex C), none of the research centers funded by the Office of Rural Health Policy has done any survey work on health care and economic activity in the 15 years that he has served in this Office. The only research group that has investigated the link between rural health care and economic activity is the National Center for Rural Health Works at Oklahoma State University, and that research has been based upon input-output models of economic linkages, not upon a national survey.¹²

In addition to these individual consultations, the project leader conducted a webinar on the survey objectives and design on November 16, 2011 to obtain feedback from experts working on rural development and rural health care issues. The webinar was organized by the North Central Regional Center for Rural Development (NCRCRD) (based at Michigan State University), was advertised by the NCRCRD and on the Community Economics Network listserve of the Applied and Agricultural Economics Association (AAEA), and was open to all interested parties. 15 experts participated in the webinar, including seven researchers, five extension specialists, and three lecturers. The backgrounds of the experts included economics, health care and sociology.

All of the experts consulted agreed that the objectives of the proposed study are important and do not duplicate existing research. Several offered suggestions for improvement in the design of the approach and questionnaires, and these have been taken into account in the proposed design.

In addition to these consultations, a review of surveys related to health care conducted or supported by Federal Government agencies was conducted, and none was found that addressed the objectives of this Information Collection. Among the surveys reviewed were the National Health Care Surveys conducted by the National Center for Health Statistics (NCHS) of HHS (http://www.cdc.gov/nchs/dhcs.htm); the Medical Expenditure Panel Survey conducted by the Agency for Healthcare Research and Quality (AHRQ) of HHS (http://meps.ahrq.gov/mepsweb/); national and state databases of the Healthcare Cost and Utilization Project of AHRQ; the Medicare Current Beneficiary Survey conducted by the Centers for Medicare and Medicaid Services (CMS) (https://www.cms.gov/MCBS/); other data on hospital quality and health outcomes collected by CMS; and other national and state level data sources cited by the AHRQ’s website and reports (http://www.ahrq.gov/data/dataresources.htm). Although these data sources cover a wide range of topics related to health care access, costs, quality and outcomes, none of these sources collects information on economic determinants or outcomes of health care provision in rural areas, and none surveys community leaders involved in recruiting or retaining rural health care providers.

The most relevant national survey to the objectives of the proposed Information Collection is the Community Tracking Study (CTS), supported by the Robert Wood Johnson Foundation (http://www.icpsr.umich.edu/icpsrweb/content/HMCA/community-tracking-study.html). The CTS is a large longitudinal survey conducted in 60 sites, including 51 metropolitan areas and 9 nonmetropolitan areas. The CTS is investigating the ways in which health care and health insurance providers are restructuring their systems and the forces driving these changes, in addition to tracking changes in health insurance coverage, access to care, use of health services, health care costs, and the perceived quality of care. The CTS has included surveys of households and physicians since 1996-97 and an employer survey conducted in 1997. Although there is some overlap between the information collected by the CTS and the proposed Information Collection (for example, both surveys collect data on the basic characteristics of physicians and their practices, their satisfaction in practicing medicine, and factors affecting their ability to provide quality and affordable care), none of the CTS questions investigate how the assets of the community influence physicians’ ability to provide health care services or recruit and retain other providers, or how the provision of health care services affects the economic development of the community. No health care providers besides physicians and no community leaders (other than employers in one round) were included as respondents in the CTS surveys, and very few rural communities were included in the sample. Hence, it would not be possible to address the objectives of the proposed Information Collection using data from the CTS.

5. Economic impacts on small businesses or other small entities

This survey will not have a significant economic impact on small businesses or other small entities.

6. Consequence to Federal program and policy activities if collection not conducted

If the proposed information collection is not conducted, research and knowledge on the roles rural communities play in recruiting and retaining health care providers will remain limited. This lack of knowledge will continue to limit the effectiveness of Federal, State and local government efforts to improve provision of health care in rural areas and realize the potential for this to contribute to improved economic development and quality of life in these areas.

7. Special circumstances

There are no special circumstances.

8. Federal Register notice

A “Notice of Intent to Request New Information Collection” related to this ICR was published in the Federal Register on September 12, 2011 (76 FR, No. 176, pages 56141-56143). We received one public comment (Annex D). The comment asserted that Medicare has much of this information already and stated that the budget of this project should be cut to zero. We disagree with the assertion that Medicare has this information already. As indicated under items 1 and 4, a review of the relevant literature and data sources (including data available from Medicare) and consultations with experts on rural health care concluded that data are not available on the specific topics of this survey; namely, on the roles that the assets and investments of rural communities play in recruitment and retention of rural health care providers, or on the impacts that health care provision has on community economic development. We therefore made no changes in the planned survey in response to this comment.

With regard to efforts to consult with persons outside the agency to obtain their views on data availability, frequency of collection, clarity of instructions and recordkeeping, disclosure or reporting format, and on the data elements to be recorded, disclosed, or reported: The individuals consulted on the design of the survey are listed in Annex B, as noted previously. These people were able to comment on issues related to sampling design, data collection frequency (one-time survey), availability of the data requested of the respondents, and specific data elements. No concerns were raised related to the data collection frequency or the availability of the data, since the questions in the survey focus on individual characteristics or personal opinions that the respondents will readily know. None of the questions requires a review of the respondents’ records, or future recordkeeping. Concerning the clarity of instructions and interpretations of the specific questions, cognitive interviews were conducted with seven individuals representing different types of respondents. A report on these cognitive interviews in included in Annex E. Based on the comments of the experts consulted and those received from participants in the cognitive interviews, several revisions were made to the survey questionnaires, including rewording or deleting questions that were not clear or judged to be less necessary for the objectives of the survey. The survey questionnaires are provided in Annexes F and G. Annex F provides the questionnaire that will be used with health care providers, and Annex G provides the questionnaire for community leaders and other stakeholders.

9. Incentives for respondents

No monetary or non-monetary incentives will be provided to survey respondents.

10. Assurance of confidentiality

The confidentiality of the Rural Community Wealth and Health Care Provision Survey data is protected under the statutes of U.S. Code Title 18, Section 1905, U.S. Code Title 7, Section 2276, and Title V of the E-Government Act, Confidential Information Protection and Statistical Efficiency Act of 2002 (CIPSEA), (Public Law 107-347). Questionnaires include a statement that answers to the survey will be kept confidential, and that under no circumstances will identifying information about individuals be released to any unauthorized individuals, agencies, or institutions. It will assure respondents that only aggregated statistics will be reported, and that providing answers to any or all questions is strictly voluntary. Detailed disclosures regarding confidentiality, citing CIPSEA and providing the OMB Control Number for the Information Collection, will be provided in an advance letter to respondents (see Annex H), and enumerators will check to ensure that respondents have received and read the letter and disclosures prior to conducting the survey. ERS will use established procedures for survey storage and disposal to ensure that individual identifiers are protected from disclosure. ERS will also use statistical disclosure limitation methods to ensure that individual identifying information does not appear in any public data product.

ERS and ERS contractors comply with OMB Implementation Guidance, “Implementation Guidance for Title V of the E-Government Act, Confidential Information Protection and Statistical Efficiency Act of 2002 (CIPSEA), (Public Law 107-347)”. In conformance with existing law and Departmental regulations, it is the policy of the ERS that respondent identifiable information collected or maintained by, or under the auspices of, the ERS for exclusively statistical purposes and under a pledge of confidentiality shall be treated in a manner that will ensure that the information will be used only for statistical purposes and will be accessible only to authorized persons.

11. Sensitive questions

The survey will not ask any sensitive questions.

12. Estimated hour burden on respondents

The telephone survey will have a complex mixed survey administration to include telephone screening for sampling frame development, pre-notification letter with web access, multi-contact telephone interviewing, and follow-up mail/web questionnaires to non-respondents. The time required for respondents and non-respondents to read the notification materials, review instructions, discuss the project with a telephone interviewer, and decide whether to complete the survey is estimated to average 15 minutes per person. Completion time for each interview is estimated to average 20 minutes per respondent. In addition, the screening process used to develop the sampling frame will involve telephone conversations with knowledgeable people in each community. We estimate that this may require 15 minute conversations with up to 8 people per community.¹³

Pilot Study: A pilot test of the survey will be done in advance of the full implementation of the survey. The purpose of the pilot study is to evaluate the survey protocol and estimate the response rates for the survey as a whole and for individual questions. If the response rate is too low for individual questions or for the survey as a whole, some questions will be dropped in the full survey. The Economic Research Service will notify the Office of Management and Budget (OMB) of the results of the pilot study before proceeding with the full survey, and submit any changes to the study materials or survey questions to OMB for clearance. It is anticipated that any changes would be non-substantive, with no upward revision of burden hours, scope or design of the study.

The sample in the pilot study will be drawn from 10 of the 150 communities in the full sample, and the data from the pilot study will be combined with the full survey data in the analysis of results (except if questions are dropped or modified due to problems identified during the pilot study). The maximum sample size for the pilot study of the research will be 200 respondents (10 communities x 16 respondents maximum per community (including up to 8 health care providers and 8 community leaders)/80% expected response rate). The maximum total estimated burden for full respondents in the pilot study is 93 hours (160 respondents x 35 minutes per respondent), and for non-respondents is 10 hours (40 non-respondents x 15 minutes per non-respondent). In addition, we estimate a maximum burden of 20 hours on non-sample individuals interviewed during the sampling frame development process for the pilot study (10 communities x 8 interviewees/community x 15 minutes per interviewee), and 1 hour for individuals who are contacted for these interviews but who refuse to participate (10 communities x 2 non-respondents/community x 3 minutes per non-respondent).

Full Study: The maximum sample size for the full study (in addition to those sampled in the pilot study) is 2,800 respondents (140 communities x 16 respondents maximum per community/80% response rate). The maximum total estimated response burden for all of those participating in the study is 1,307 hours (2,240 respondents x 35 minutes per respondent¹⁴) and for the non-respondents is 140 hours (560 non-respondents x 15 minutes per non-respondent¹⁵). In addition, we estimate a maximum burden of 280 hours on non-sample individuals interviewed during the pre-sample screening process for the full study (140 communities x 8 interviewees per community x 15 minutes per interviewee) and 14 hours for contacted individuals who refuse to participate (140 communities x 2 non-respondents per community x 3 minutes per non-respondent).

The total respondent burden, including the pilot and full study, is estimated at 1,865 hours (Table 1). The total estimated cost of the hour burden, based on estimated mean hourly salary/wage rates for the different types of respondents involved, is $91,550 (Table 2). We estimate that 50% of the hourly burden of the survey of health care providers will involve physicians and dentists and that 50% will involve other health care providers (physician assistants, nurse practitioners, nurse midwives, and pharmacists). The survey of community leaders and other stakeholders involved in recruiting and retaining health care providers (e.g., mayors, county public health officials, county economic development officials, and leaders of businesses and non-profit organizations) will involve a roughly equal number of respondents as the survey of health care providers. In addition, administrative staff of local government, business and non-profit organizations will be involved in the screening calls used to identify the list of potential respondents. The mean hourly salary/wage rates assumed for these groups – $100 for physicians and dentists, $40 for other health care professionals, $40 for community leaders and other stakeholders, and $20 for administrative staff – are based on the May 2010 Metropolitan and Nonmetropolitan Area Occupational Employment and Wage Information from the Bureau of Labor Statistics website (http://www.bls.gov/oes/current/oessrcma.htm), using the data for nonmetropolitan regions of the States included in this study.

Table 1. Estimated Respondent Hour Burden

Instrument	Estimated Number of Respondents	Responses Annually per Respondent	Total Annual Responses	Estimated Average Number of Hours per Response	Estimated Total Annual Hours of Response Burden
Pilot Study
Community screening
Completed interviews	80	1.00	80	0.25	20.00
Attempted interviews	20	1.00	20	0.05	1.00
Health care providers questionnaire
Completed interviews	80	1.00	80	0.58	46.67
Attempted interviews	20	1.00	20	0.25	5.00
Community leaders questionnaire
Completed interviews	80	1.00	80	0.58	46.67
Attempted interviews	20	1.00	20	0.25	5.00
Full Study
Community screening
Completed interviews	1120	1.00	1120	0.25	280.00
Attempted interviews	280	1.00	280	0.05	14.00
Health care providers questionnaire
Completed interviews	1120	1.00	1120	0.58	653.33
Attempted interviews	280	1.00	280	0.25	70.00
Community leaders questionnaire
Completed interviews	1120	1.00	1120	0.58	653.33
Attempted interviews	280	1.00	280	0.25	70.00
Total Burden	4500	1.00	4500	0.41	1865.00

Table 2. Estimated Cost of Respondent Hour Burden

Instrument	Type of respondent	Burden hours	Unit cost	Total cost
Pilot study
Community screening	Administrative staff	21.00	$20.00	$420.00
Health care providers questionnaire	Physicians and dentists	25.84	$100.00	$2,583.50
	Other health care providers	25.84	$40.00	$1,033.40
Community leaders questionnaire	Community leaders	51.67	$40.00	$2,066.80
Full study
Community screening	Administrative staff	294.00	$20.00	$5,880.00
Health care providers questionnaire	Physicians and dentists	361.67	$100.00	$36,166.50
	Other health care providers	361.67	$40.00	$14,466.60
Community leaders questionnaire	Community leaders	723.33	$40.00	$28,933.20
Total Cost		1865.00	$49.09	$91,550.00

13. Estimated annual cost burden on respondents or record keepers

There are no capital/start-up or ongoing operation/maintenance costs associated with this information collection.

14. Estimated annualized costs to the Federal Government

The estimated total and annualized costs of the Information Collection to the Federal Government are shown in Table 3. The estimated total cost is $266,473, including $23,498 in Fiscal Year (FY) 2011 and $242,975 in FY 2012.

These costs include salaries, wages and associated fringe benefits of the Principal Investigators (PIs) at Iowa State University’s Center for Statistics and Survey Methodology (CSSM), Project Director (PD) at Iowa State University’s Survey and Behavioral Research Services (SBRS), and other professional personnel affiliated with CSSM and SBRS ($196,073 in total). The salary and wage costs are based on time commitments of 160 hours of the CSSM PI, 3,118 hours of the SBRS PD and other professional staff, 60 hours of graduate student research assistants, and 2,350 hours of telephone interviewing staff.

Materials and Supplies consist of office supplies for use by the PIs and are estimated at $500. Travel expenses are estimated to be $3,350, including one trip for three people (the CSSM PI and two professional staff of SBRS) during FY 2012 to Washington, DC to collaborate with ERS personnel on project development and survey design. Automated data processing costs are $28,550, calculated as 12% of all other direct charges. All other direct costs are estimated at $38,000 and include:

• 100,000 minutes of telephone calls (including conference calls among the study team, sample frame and development for 150 communities, and calls to non-responders and interview participants) at $0.18 per minute ($18,000);

• Miscellaneous project supplies including copy paper, printer cartridges, notebooks, labels, and material organizers ($3,000);

• Printing costs for letterhead, envelopes, project information sheets, project brochures, reminder postcards, and interviewer training materials ($10,000);

• Postage for mailing letters, postcards, brochures, and follow-up information ($7,000).

Table 3. Estimated Total and Annualized Costs of the Project to the Federal Government

	TOTAL		Year 1 (10/10-9/11)		Year 2 (10/11-9/12)
A. Salaries and Wages
1. PI(s) / PD (s)	$12,800		$4,160		$8,640
2. Other Professional Personnel	$137,744		$10,932		$126,812
Total Salaries and Wages	$150,544		$15,092		$135,452
B. Fringe Benefits	$45,529		$5,208		$40,321
C. Total Salaries, Wages, and Fringe Benefits (A plus B)	$196,073		$20,300		$175,773
D. Nonexpendable Equipment	$0		$0		$0
E. Materials and Supplies	$500		$125		$375
F. Travel – Domestic	$3,350		$0		$3,350
G. Publication Costs	$0		$0		$0
H. Automated Data Processing (computer) Costs	$28,550		$2,518		$26,032
I. All Other Direct Costs
1. Telephone	$18,000		$180		$17,820
2. Interviewer and Project Supplies	$3,000		$375		$2,625
3. Printing	$10,000		$0		$10,000
4. Postage	$7,000		$0		$7,000
Total All Other Direct Costs	$38,000		$555		$37,445
J. Total Direct Costs	$266,473		$23,498		$242,975

15. Reasons for program changes or adjustments

This is a new information collection.

16. Plans for tabulating and publishing survey results

Data collection for the survey is expected to begin in August 2012 and to be completed by December 2012. An initial technical report tabulating and discussing the results will be completed by March 2013. The tabulated results in the initial technical report will include the following:

• Descriptive statistics (means or proportions and standard errors) of all closed-form response variables (mostly binary (e.g., yes/no) and ordered responses), reported for the entire study region/respondent population, for each region and hospital stratum (with/without hospital), and for each respondent group stratum (physicians, dentists, other health care providers, and community leaders/other stakeholders).

• Descriptive statistics of selected open-ended response variables, reported for the entire population and for the subpopulations mentioned above. Responses to open-ended questions such as “What was the most important factor that drew [a health care provider] to practice in this town” will be post-coded, and where there is significant variation and useful clustering of these responses, the results will be tabulated.

• Statistical tests of comparisons between means or proportions of variables across the different subpopulations mentioned above, estimated as explained in Part B, item 2.

• Estimation and comparisons of descriptive statistics of selected variables for other subpopulations. The set of subpopulations and variables for which tabulations will be included in the report will be decided after investigation of the data. The subpopulations that will be investigated include:

  • communities with good access to a broad range of assets vs. communities with less favorable access;

  • communities with good access and quality of health care services vs. communities with less favorable health care services;

  • communities in which health care services are improving vs. communities in which these services are stagnant or declining;

  • communities that have made significant efforts to recruit or retain health care providers vs. communities that haven’t;

  • communities whose efforts to recruit or retain health care providers have been more or less successful.

The utility of classifying these subpopulations cannot be pre-specified, because this depends on the distribution of these types that will be revealed by the survey. The response variables that will be investigated for these different subpopulations will be based upon hypotheses about how responses are expected to vary across subpopulations. For example, the ability to recruit health care providers, provider satisfaction, retention, and changes in the availability and quality of health care services are hypothesized to depend upon community assets. Hence, response variables reflecting these outcomes will be compared across subpopulations of communities having different levels of community assets.

The initial technical report will not be published, but the results will be presented in technical seminars at ERS and elsewhere. After obtaining feedback from peers on the initial findings, the report will be revised and submitted for review as an ERS research publication, including further analysis of the data. The tabulations in the initial report will suggest relationships between the factors used to define the subpopulations being compared and the outcome variables investigated. Further investigation of such relationships will be conducted using multivariate statistical methods, including ordinary least squares (OLS) regression and maximum likelihood models such as simple probit (for binary response variables) and ordered probit models (for ordered response variables). Multivariate methods will enable the analysis to account for the influence of other factors (including those for which secondary data are available, such as population level, access to urban centers, infrastructure, natural amenities or other factors) on the response variables being investigated. These methods are discussed further in Part B, item 2.

The ERS research report discussed above is expected to be submitted for review by June 2013, and to be published during 2013. The results of this research will suggest hypotheses for further exploration. For example, if certain types of community assets are found to be associated with communities having greater success in recruiting and retaining health care providers, more in-depth research into how such assets are contributing to this success and the prospects for replicating successes in other communities, could be very useful. To address this research need and opportunity, we plan to conduct further case study research in a subsample of the communities included in this survey (in up to 40 of the communities) in 2013, using field interviews in these communities. A separate Information Collection Request will be prepared for this phase of the research.

17. Reasons to not display the expiration date for OMB approval

The OMB approval expiration date will be displayed.

18. Exceptions to the certification statement

There are no exceptions to the certification statement for this information collection.

1 For example, analysis of employment growth trends by industry and county from 2002 to 2007 revealed that two of the top six industries experiencing rapid employment growth in the largest number of non-metropolitan counties were in the health care sector: hospitals and nursing care facilities.

2 For example, the mean number of active non-Federal medical doctors per 100,000 population in non-metropolitan retirement destination counties was 103 in 2007, compared to 84 in other non-metropolitan counties.

3 Doescher, et al. (2009). References are included in Annex A.

4 Gamm, et al. (2003).

5 Excellent reviews of this literature are provided by Crandall et al. (1990); Brooks et al. (2002); Gamm et al. (2003); Laven and Wilkinson (2003); Rabinowitz et al. (2008); Ballance et al (2009); and Rosenblatt et al. (2010). See Annex A for the list of references.

6 Examples of research on factors affecting recruitment and retention of health care providers other than physicians include Lin et al. (1997); Anderson and Hampton (1999); Fairbanks et al (2001); and Daniels et al. (2007).

7 Most of the studies investigating the economic impacts of rural health care provision in the United States are based on predictions of input-output models rather than data on actual ex post impacts; e.g., see Doeksen et al. (1998); Doeksen and Schott (2003); St. Clair, Doeksen, and Schott (2007); and St. Clair and Doeksen (2009).

8 Newhouse et al. (1982); Dionne et al. (1987); Langwell et al. (1987); and Carpenter and Neun (1999).

9 Cutchin (1997); Hancock et al (2009); Hanlon et al. (2010); Quinn and Hosokawa (2010).

10 Felix et al. (2003); Shannon (2003).

11 Detailed information on the analysis to be conducted and the research reports to be produced is provided below in the response to question 16.

12 See the citations given in footnote 7 for examples of the research led by this research center.

13 The script used for sample development is provided in Annex J.

14 The 35 minutes per respondent includes 15 minutes to review the materials, participate in the screening interview, and decide whether to participate, and 20 minutes to complete the questionnaire.

15 The 15 minutes per non-respondent is to review the materials, participate in the screening interview, and decide whether to participate.