OMB Unreimbursed Care Draft 10-23-06
OMB Clearance Application
Assessment of Unreimbursed Care among Community Primary Care Physicians.
November, 2006
Agency of Healthcare Research and Quality (AHRQ)
Table of contents
1. Circumstances of Information Collection - 3 -
2. Purpose and Use of Information - 4 -
3. Use of Improved Information Technology - 4 -
4. Efforts to Identify Duplication - 4 -
5. Involvement of Small Entities - 4 -
6. Consequences if Information Collected Less Frequently - 4 -
7. Special Circumstances - 5 -
8. Consultation outside the Agency - 5 -
9. Payments/Gifts to Respondents - 5 -
10. Assurance of Confidentiality - 5 -
11. Questions of a Sensitive Nature - 5 -
12. Estimates of Annualized Burden Hours and Costs - 5 -
13. Estimates of Annualized Respondent Capital and Maintenance Costs - 6 -
14. Estimates of Annualized Cost to the Government - 6 -
15. Changes in Hour Burden - 6 -
16. Time Schedule, Publication and Analysis Plans - 6 -
A. Purpose and Main Research Questions - 6 -
C. Tabulations and Statistical Analysis - 8 -
D. Time Schedule and Publication Plan - 10 -
17. Exemption for Display of Expiration Date - 10 -
18. Exceptions to Certification - 10 -
B. Collections of Information Employing Statistical Methods - 10 -
1. Respondent universe and sampling methods - 10 -
2. Information Collection Procedures - 11 -
The “Assessment of Unreimbursed Care Among Community Primary Care Physicians” is being conducted in response to a modification of an AHRQ Request for Proposals (RFP) entitled “Resource Center for Primary Care Practice-Based Research Networks (PBRNs)” (Issued under Contract 290-02-0008). Many current definitions of the “core safety net” do not include free or reduced price healthcare provided in private physicians’ offices, despite the fact that these clinicians deliver a substantial portion of care for uninsured and vulnerable populations.1 Prior research seeking to quantify the amount of safety net care in private physicians’ offices has found that approximately 70-80% of private physicians provide some type of unreimbursed care. The purpose of this study is to establish how primary care physicians view their obligations to provide unreimbursed care and what critical factors permit or deter them from providing unreimbursed care to patients in their offices.
Studies have shown that physicians in practices that derive a larger percentage of their revenue from managed care provide fewer hours of unreimbursed care than those who derive a low percentage of their revenue from managed care.2 Recent findings also indicate that physicians in small group practices and those in physician owned practices are more likely to provide unreimbursed care than those in practices with many doctors and non-physician owners respectively.2 Recognizing this, the study will use web and paper-based questionnaires distributed to an array of diverse practice environments to assess the level of unreimbursed care provided as well as the differences in attitudes of the physicians in their various practice settings.
Since 2000 AHRQ has supported the development of primary care Practice Based Research Networks (PBRNs). PBRNs design and conduct clinical research in real world practices and translate important research findings into the everyday practice of medicine. PBRN members are mainly non-academic, community-based primary care clinicians and practices that offer a potentially valuable resource for identifying genuine factors that enable or hinder private office safety net care. In 2002 AHRQ funded a PBRN Resource Center to assist in its mission to improve health care services to all, by facilitating, disseminating, and implementing, the highest quality of primary care research. The PBRN Resource Center was awarded to the collaborative team of Indiana University and the National Opinion Research Center (NORC) at the University of Chicago.
The Agency for Healthcare Research and Quality (AHRQ) is requesting Office of Management and Budget (OMB) approval to conduct a study through the American Academy of Family Physicians (AAFP) National Research Network (AAFP-NRN) of primary care physicians who are active members of the AAFP about the factors that permit or deter private primary care practicing physicians from providing unreimbursed or discounted care to their patients. The AAFP-NRN is a primary care Practice Based Resource Network (PBRN) registered with AHRQ’s PBRN resource center.
Current information about the factors that permit or deter private office physicians from providing unreimbursed care to their patients in the primary care setting is limited. Though research has been conducted to quantify the amount of unreimbursed care provided in private physicians' offices, we must now try to understand the critical factors that influence physicians when deciding whether or not to provide unreimbursed care. This assessment seeks to investigate these factors in conjunction with the current state of unreimbursed care and physician/practice demographic information. This information will be useful as the current trends toward managed care and large corporate practice continue to grow, which may consequently result in private physicians becoming less willing and able to provide unreimbursed care. This may subsequently place significant pressure on formal safety-net structures and have a potentially significant impact on the health care of vulnerable populations in the U.S. Therefore, this assessment will provide AHRQ with information that will be helpful in promoting and protecting the quality of healthcare in vulnerable populations.
The data will be collected using a self-administered paper- and web-based questionnaire.
The information being collected is intended to document the current level of unreimbursed care provided along with factors that encourage or discourage private office physicians from engaging in the provision of unreimbursed care. Although there has been some research conducted to quantify the amount of unreimbursed care provided in private physicians offices, the information that we seek to collect is unique in the fact that it explores more in depth the deciding factors in implementing this care. This information is also unique because it has not been collected using a PBRN as a source for data collection. PBRNs are an ideal place to examine the current level of unreimbursed care because they have access to and command the respect of private physicians across the United States.
The collection of this information will involve a sampling of physicians; therefore small entities will be affected. However, because the survey will be both mail and web-based, burden will be kept to a minimum. Further, completion of the survey will require minimal time out of respondents’ work days to complete (approximately 15 minutes).
The design of this study requires only one data collection activity per respondent. Without collecting this data, AHRQ will not have access to a comprehensive assessment of the current state of unreimbursed care provided in the private primary care setting, as well as knowledge of the factors that encourage or discourage practices from providing unreimbursed care.
This request is consistent with the general information collection guidelines of 5 CFR 1320.5(d)(2). No special circumstances apply.
Attachment 1 is a copy of the Federal Register notice (July 12, 2006) soliciting comments on the information collection. AHRQ also consulted with members of the PBRN Resource Center staff at NORC as well as individuals from selected agencies outside of NORC. See Attachment 2 for a list of consulted individuals.
There will be no payments or gifts to the respondents
Personal identification information (i.e., respondent names) will not be collected in the surveys. Although the individual will be asked about practice demographics and physician demographics, this information will only be presented in aggregate form to AHRQ. The AAFP-NRN will use randomly assigned tracking numbers to ensure that the data collection is balanced across the selected strata and this information will not at any time be shared with the federal government. All potentially identifying information will be destroyed at the study’s conclusion.
Respondents will be asked about their experience in providing unreimbursed care to patients and factors that affect this decision. The survey will not include any questions of a more sensitive or personal nature.
In Exhibit 1, we provide estimates of the collection burden on participants for this effort. Study participants will participate in data collection one time only.
Exhibit 1. Estimate of Cost Burden to Respondents
Data Collection Effort |
Number of Respondents |
Estimated Time per Respondent in Hours |
Estimated Total Burden hours |
Average Hourly Wage Rate* |
Estimate Annual Cost Burden to the Respondent |
Primary care clinicians: |
780 |
.25 |
195 |
$57.90 |
$11,291.00 |
Total burden (hours): 195
Total imputed costs: $11,291.00
*Based upon the mean of the average wages for physicians, National Compensation Survey: Occupational wages in the United States 2004, “U.S. Department of Labor, Bureau of Labor Statistics.”
Data collection for this study will not result in any additional capital, start-up, maintenance, or purchase costs to respondents or record keepers. Therefore, there is no burden to respondents other than that discussed in the previous section (A.12).
The total cost to the government for this activity is estimated to be $129, 956.00
The only changes in the burden discussed above would result from feedback during pre-testing.
This section contains a detailed analysis plan for this study. In order to present a coherent plan, this section presents an overview of the study purpose and main research questions, reviews the data sources, discusses the types of results the study will produce, the statistical analyses that will be conducted, and the time schedule for completing the project, including publication of the results.
The American Academy of Family Physicians (AAFP) is the largest medical specialty organization devoted solely to primary care with more than 94,000 family physician, family medicine resident and medical student members. Family physicians provide the majority of care for America’s underserved rural and urban populations. A survey of these physicians will facilitate a better understanding of the current state of private office primary care, unreimbursed care and, more importantly, help assess the factors that encourage and discourage practices from providing unreimbursed care. This information will be of great value for promoting both AHRQ’s and PBRNs’ missions to promote the quality of health care in vulnerable populations.
The goal of this assessment is to create a comprehensive picture of the current state of unreimbursed care and specifically what factors influence the provision of unreimbursed care in private physician practices. The user group targeted for this survey is the AAFP active membership who report spending at least 50% of their time in direct patient care. The main research questions this study will address are outlined in Exhibit 2.
Exhibit 2: Key Research Questions
1. Demographics? |
|
2. Current State of Unreimbursed Care? |
|
3. Factors Influencing the Provision of Unreimbursed Care |
|
The findings of this study will assist ARHQ in providing quality health care to vulnerable populations.
Data will be solicited from 1200 randomly sampled primary care physician respondents from the AAFP. A stratified sampling approach will be used to ensure appropriate representation from the four census regions, urban and rural areas, and small and large practices. Sampled physicians will receive a letter (Attachment 3) from the AAFP-NRN project team informing them of the purpose of the study, an invitation to participate, and a paper questionnaire. They will be asked to either complete the paper version and return it by a supplied postage-paid envelope to the AAFP-NRN office, or access the on-line version of the survey to complete (the URL will be provided in the letter). The questionnaire is estimated to take no more than fifteen minutes to complete. A second invitation to participate will be re-mailed to non-responders after three weeks. Are-mailing of the paper-based questionnaire will occur in week seven to all non-respondents. Reminder phone calls will be placed as needed in weeks 10 to 12 to non-responders.
The survey consists of three sections:
Demographic Information: Captures information about practice and physician demographics.
Current State of Unreimbursed Care: Captures data about level of unreimbursed care provided, patient population and location of services.
Factors Influencing the Provision of Unreimbursed Care: Captures data about enabling factors, disabling factors and level of vulnerability.
The survey is included as Attachment 4.
AAFP-NRN staff has conducted similar national mail and/or web-based surveys between 2001 and 2004. Across four such surveys with AAFP active members, response rates have ranged from 55% to 63% with either one or two mail follow-ups. With the addition of telephone follow-ups to be implemented as needed for the current survey, the likelihood of meeting the goal of a 65% response rate is anticipated.
This section details the tabulations and statistical analyses that will be conducted for this study. Data will be collected from primary care physicians who are current active members of the AAFP to address the main research questions. This study will use both univariate and, where possible, multivariate techniques to analyze this data.
Descriptive statistics will be provided for eligible as well as ineligible physicians and their offices. These statistics include physician characteristics such as gender, age, race-ethnicity, and the number of years since graduation. Additionally, eligible physician characteristics include the number of patients personally seen during a typical week. Questions concerning physicians’ provision of un-reimbursed care can be divided up into the following three subgroups:
Physicians who have never provided un-reimbursed care to patients,
Physicians who currently do not provide un-reimbursed care to patients, but who have provided such care in the past, and
Physicians who currently provide un-reimbursed care to patients.
Comparisons between these three subgroups using a categorical statistic such as, the Pearson Chi-square or Mantel-Haenzel Chi-square tests will be made by size of office (large vs. small); by type of ownership; and by other available characteristics, as needed.
Tabulation of characteristics related to the selected physician’s group practice (whether the physician is eligible or not) will include the presence or absence of offices in addition to the primary office located where the physician sees most of his/her patients. The urban/rural classification of the primary office location is determined based on physician self-reports as well as physician-reported 5-digit zip codes. Tabulations of the number of clinicians associated with the primary location and overall, type and ownership of the practice will also be provided.
Inferential data analysis will focus on identifying results of the following key research questions.
Research question I examines the physicians’ attitudes towards the provision of unreimbursed care. This objective attempts to assess the impact of physician attitudes towards unreimbursed care on their willingness to provide such care. The survey instrument identifies specific factors that may or may not affect a physician’s decision to provide unreimbursed care in their office setting and asks the participants to rate their importance. Reasons for willingness to offer and limitations to providing un-reimbursed care will also be summarized specifically for specialty care including: lab services, radiology or imaging services referrals, and access to medication.
Several statistical analyses will be conducted to investigate the question including basic descriptive statistics within each category such as univariate means and frequency tabulations. Comparisons between the three subgroups using a categorical statistic such as the Pearson Chi-square or Mantel-Haenzel Chi-square tests will be made by size of office (large vs. small), by type of ownership, and by other available characteristics, as needed.
Research question II seeks to determine the factors that permit private office physicians to provide un-reimbursed care. Alternatively, factors that lead to deterrence will be determined. Critical factors that could influence physicians that previously provided un-reimbursed care to provide un-reimbursed care again will also be determined.
Several statistical analyses will be conducted to investigate the question including basic descriptive statistics within each category such as univariate means and frequency tabulations. Comparisons between the three subgroups using a categorical statistic such as the Pearson Chi-square or Mantel-Haenzel Chi-square tests will be made by size of office (large vs. small), by type of ownership, and by other available characteristics, as needed.
Research question III focuses on those physicians who currently provide un-reimbursed care to patients and the type and frequency of care provided. In particular, frequency of care will be established by patient age group and specified service. Statistics concerning who determines eligibility for un-reimbursed care and verification of such eligibility will be displayed. Correlations between these factors and physician characteristics (and the characteristics of their offices) will be established. Standard errors will be provided for these estimates. Non-sampling errors arising from unit and item non-response will be dealt with through weighting and imputation where appropriate.
Additionally, the questionnaire will allow the PBRN Resource Center to gather data on the demographic characteristics of the physicians’ practices in order to compare the characteristics of the three subgroups (e.g. staffing levels, number of clinicians, resources available). Results from the research questions will be examined to identify enabling and prohibitive factors affecting the physicians’ ability to provide unreimbursed care. Using hierarchical modeling, the relationships between office demographics and decisions/ability to provide unreimbursed care will be explored.
Exhibit 3. Timetable for Data Collection, Analysis, and Publication
Activity |
Expected Date of Completion |
Survey sent to respondents |
1-4 months following OMB approval |
Data preparation |
4 months following OMB approval |
Analyze Findings |
5 months following OMB approval |
Prepare Draft Reports |
6-7 months following OMB approval |
Final Report |
7-8 months following OMB approval |
AHRQ does not seek this exemption.
There are no exceptions to the certification statement.
Surveys are being administered to collect information about the current status of unreimbursed care in private physician practices and factors that affect physicians to provide or not provide unreimbursed care.
The results will be generalizable to the respondent universe, which consists of private physicians from the AAFP active membership of those who report seeing patients at least 50% of the time. Based on current information from the AAFP Member Master Database of active U.S. members who have reported on time spent in direct patient care (N=35,461), 91.3% report spending from 50% to 100% (mean=84% and median=90%).
The sample will include 1200 AAFP active physicians from the four census regions (50 states and District of Columbia), metropolitan and non-metropolitan areas, and practice arrangement. The unit of analysis for the sample will be private physicians from the AAFP who report at least 50% of their time in direct patient care. This sample of private physicians will be able to provide the type of data necessary to evaluate the current state of unreimbursed care in private physician practices, along with factors that affect the provision of this care.
We will draw a stratified random sample based on these sampling specifications:
Four census regions (Northeast, Mid-west, South, and West)
Geographic location (metropolitan & non-metropolitan based on zip code specifications)
Practice arrangement (Solo and 2-physician practice; family medicine and multi-specialty group; and
Other.
These current data show:
|
Anticipated Sample Size |
Anticipated Response Rate |
Anticipated Respondents |
|
Census region |
|
|
|
|
Northeast |
14% |
168 |
65% |
110 |
Mid-west |
29% |
348 |
65% |
227 |
South |
35% |
420 |
65% |
273 |
West |
22% |
264 |
65% |
172 |
Practice Arrangement |
|
|
|
|
Solo |
18% |
216 |
65% |
141 |
2-physician |
8% |
96 |
65% |
63 |
FM Group |
41% |
492 |
65% |
320 |
Multi-specialty |
21% |
252 |
65% |
164 |
Other |
12% |
144 |
65% |
94 |
Geographic location |
|
|
|
|
Metropolitan |
76% |
912 |
65% |
593 |
Non-metropolitan |
24% |
288 |
65% |
188 |
Project staff will access the most recent AAFP Member Master Database and select the stratified random sample from it two weeks prior to the first mailing of the study questionnaire. The cover-letters will be personalized and inserted into an outgoing envelope that will also include the paper questionnaire and the postage-paid return envelope. These documents will be printed by AAFP duplicating services, and the documents will be collated and inserted into the mailing envelopes by AAFP mailing services. These same services are used by staff for all NRN projects.
The sample will include 1200 physician members from the AAFP. The unit of analysis for the survey will be private physicians from the AAFP, and no physician will be asked to complete more than one survey. Fielding of the survey will entail mailing cover letters explaining the purpose of the survey to members of the AAFP along with paper surveys and return postage-paid envelopes. The letter will invite the sampled physicians to complete either the enclosed paper survey or complete the online version with the URL furnished. Then after three weeks, non-respondents will receive a second paper questionnaire. Non-respondents will also be mailed a third questionnaire in week seven. Reminder phone calls will be placed in weeks 10-12 to all non-responders.
Project investigators will use a receipt control system using case ID numbers to track the initial survey mailing, address updates, completion of the questionnaire online, mailing follow-up questionnaires, phone call reminders and complete and incomplete questionnaire returns.
All data from the completed questionnaires will be keyed (data entered) to create the analytic data file. All paper questionnaires will be keyed a second time (double entry). AAFP-NRN project staff will also conduct a 10% sample of returned and keyed surveys to obtain an estimated rate of data-entry errors. The double keying verification process will ensure that data-entry errors are minimized, and the 10% review will allow researchers to report the rate of accuracy to the Project Officer. The questionnaires will be edited and keyed as they are received. Data entry of the final batch will be completed within two weeks of the close of data collection.
The investigators will use a number of proven methods to maximize participation in the study. First, the instrument itself is designed to maximize response rates. The style of the survey is inviting and user friendly, with a maximum of 47 questions. The instructions for the survey are straightforward, and there are a limited number of skip patterns. Second, the questionnaire will be pilot tested with physicians from the sampling frame, and questions will be amended to reflect suggested improvements from these respondents. Third, each surveyed physician will receive a cover letter encouraging participation in the survey (Attachment 3). The cover letter will 1) convey the importance of the survey to AHRQ and to the AAFP, and 2) will indicate that the respondent will not be identified to any government agency. It also will be signed by either the AAFP President, AAFP Board Chair, or AAFP Executive Vice-President. Finally, the investigators will mail out a follow-up paper questionnaire to non-responders after three and seven weeks. Reminder phone calls will be placed in weeks ten to twelve to all non-responders after the second follow-up mailing during week seven.
In order to create an accurate survey tool, interviews were conducted with several private physicians about their experience in providing unreimbursed care. After vetting through the interview notes, questions for this survey were created to elicit a response that would accurately paint a picture of the state of unreimbursed and the factors that contribute to the decision to provide unreimbursed care by private physicians. We discovered that it was important to reiterate our definition of unreimbursed care throughout the survey due to conflicting definitions of unreimbursed care among physicians we sampled. Therefore we have included our definition of unreimbursed care several times throughout the survey.
The instruments were pilot tested with members of the AAFP NRN recruited to take the Internet-based survey instrument and to take the paper-based survey. Variability was sought on physician and practice characteristics: gender, age, race, census region, practice type, and practice ownership. For both the on-line and paper pilot tests, an AAFP NRN project member was either on the phone at the time when the questionnaire was being completed by each pilot test physician or via the WebEx web conferencing program that allows users to view each other’s work on the computer in real time. The project staff member asked the physician to comment on any questionnaire item that was either ambiguous or might suggest alternative meanings or interpretations. Pilot test physicians were also asked to comment on other aspects of the questionnaire, including instructions, language, item and section ordering, formatting etc.
The NRN project team analyzed and summarized the result of the pre-test for each item and shared the pilot test results with the rest of the project team.
Dan Gaylin. MPP
Executive Vice President for Health Research, NORC
1350 Connecticut Ave, NW, Suite 500
Washington, DC 20036
Claudia Schur
Associate Director, Health Policy and Evaluation, NORC
7500 Old Georgetown Rd
Betheseda, MD 20814
Caitlin Oppenheimer, MPH
Senior Research Scientist, NORC
1350 Connecticut Ave, NW, Suite 500
Washington, DC 20036
Janella Chapline
Senior Statistician, NORC
Benjamin Hamlin, MPH
Senior Research Analyst, NORC
1350 Connecticut Ave NW, Suite 500
Washington, DC 20036
Wilson D. Pace, MD
Network Director
AAFP National Research Network
11400 Tomahawk Creek Parkway
Leawood, KS 66211
James M. Galliher, Ph.D.
Research Director
AAFP National Research Network
11400 Tomahawk Creek Parkway
Leawood, KS 66211
Aaron J. Bonham, MS
AAFP National Research Network
11400 Tomahawk Creek Parkway
Leawood, KS 66211
References:
1. Kinney ED, Tai-Seale M, Greene JY, Murray R, Tierney W. Three political realities in expanding coverage for the working poor: one state's experience. Health Aff (Millwood). Jul-Aug 1999;18(4):188-192.
2. Cunningham PJ, Grossman JM, St Peter RF, Lesser CS. Managed care and physicians' provision of charity care. Jama. Mar 24-31 1999;281(12):1087-1092.
| File Type | application/msword |
| File Title | OMB Clearance Application |
| Author | hamlin-ben |
| Last Modified By | wcarroll |
| File Modified | 2006-11-21 |
| File Created | 2006-11-21 |