passback

Response to OMB comments for Unreimbursed care

1. What is meant by "active member" of the AAFN? How many primary care physicians are not active members, and how do they differ from active members? Also, how do members in general differ from non-members?

The American Academy of Family Physicians (AAFP) is the national association of family doctors. It is one of the largest national medical organizations, with members in 50 states, D.C., Puerto Rico, the Virgin Islands, and Guam. Active members of the AAFP are those practicing physicians who have completed training and met eligibility, CME, and Chapter requirements outlined in the bylaws of the organization. The American Academy of Family represents more than 94,000 members of the AAFP and about 64,000 active members, the rest being medical residents, medical students, and retired physicians.

Over 90% of US family physicians are members of the AAFP. Non-members are more likely to be recently trained physicians and physicians in part-time practice.

AAFP active members represent about 25% of the 222,000 US primary care physicians actively caring for patients in the US.

2. Please clarify how the total cost to the government came to $129,956. This isn't clear from the materials provided.

The amount listed is the total contract value for the work being proposed and is being used by the AAFP-NRN to support the data collection and by the PBRN Resource Center for the analysis of the data collected.

3. Where will AHRQ present the standard PRA blurb, expiration date, and control number?

These materials will appear on the cover sheet for the survey. An updated copy of the survey with the cover sheet is attached.

AHRQ is not seeking any exemptions for this project.

A.17. Exemption for Display of Expiration Date

AHRQ does not seek this exemption.

A.18. Exceptions to Certification

There are no exceptions to the certification statement.

4. In the supporting statement, the distinction is drawn several times between private physicians and physicians in managed care. It is my understanding, however, that many physicians are both private physicians and also have contracts with managed care. Can AHRQ please clarify? Also, if this is an important distinction to capture, where in the instrument or the sampling will this distinction be captured?

The target respondent group is physicians in private office settings; physicians possessing contracts with managed care organizations will not be excluded from the survey. The demographic portion of the questionnaire asks several questions about the ownership of the practice and the physician’s relationship within the organizational structure of the practice. Within the survey questions also focus on pressures from both within and external to a practice and the type of employment relationship in which the physician works.

5. Will primary care physicians who work in the traditional "safety net" organizations (e.g. community care clinics, public hospitals, etc) be excluded from the study?

The survey is intended to capture responses from private physician practices and the survey will be directed to those physicians based on the AAFP membership data. Physicians in traditional safety net organizations will be excluded from the sample.

6. Under the information collection procedures it indicates that reminder phone calls will be used to decrease non-response. How many reminder phone calls are planned for weeks 10 - 12?

All non-respondents will receive one phone call. Additional calls may be made based on overall response rates and response rates within target recruitment cells.

7. Please provide more information on the results of the pilot project. How many members and physicians were surveyed? When did this occur? What were the results?

The pilot test which took place during September and October of 2006 and was conducted by the AAFP-NRN. The staff recruited separate groups of subjects to test the paper and Internet-based survey instrument. They attempted to maximize the diversity/variability of physicians assigned to each survey format. Variability was sought on physician and practice characteristics: gender, age, race, census region, practice type, and practice ownership. A total of 15 volunteers participated in the pilot test, 8 on the paper and 7 online.

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 is being completed by each pilot test physician or we was logged into 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 is either ambiguous or might suggest alternative meanings or interpretations. The 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 suggested revisions to both versions of the questionnaire, including revisions to: specific item wording, ordering of items and/or item response categories, formatting of the questionnaire, and instructions.

The revised survey is attached.

8. Does AHRQ have statutory authority to provide assurances of confidentiality? If not, all uses of the term "confidential" should be replaced with terms like "private to the extent permitted by law." "Confidential" carries a very strict legal definition and can only be used when an agency explicitly has the statutory authority to do so. If in doubt, please check with your GC.

Since our phone conference, AHRQ has decided to use our statutory authority to conduct this study. The recruitment letter has been revised accordingly. The revised cover letter is attached and contains the relevant statutory language.

9. Has this study gone before an IRB? If so, what were the results of the IRB? Were any changes made at their request? What were their comments regarding the "receipt control system" (B2)?

This study was approved by both the AAFP-NRN IRB and the NORC IRB Committees. Additional approval will be obtained after OMB clearance is obtained and documents are finalized.

10. Respondents have the choice of completing either a paper or an electronic version. It is possible that some respondents may do both - what measures will you use to unduplicate the survey responses?

The AAFP-NRN will assign a unique identifier number to each of the requests for participation. The electronic survey will ask the respondent to enter this number before proceeding. The paper survey will be pre-marked with the unique identifier. If both an electronic and paper version is returned, the electronic version will be included and the paper discarded. The identification number will be used by the AAFP-NRN to track responses and conduct follow-up mailings and phone calls. The link between number and physicians will be destroyed at the end of data collection and not shared with AHRQ at any time.

11. Definition of unreimbursed care-how about care that is not reimbursed but where the cost was offset by donations, fund-raising, or cost-shifting? What about benefits that were excluded from coverage but were provided anyway? What about consultations that occur by phone or by email?

The study authors are attempting to capture the amount of unreimbursed care provided in private physician offices. Efforts to raise funds to cover unreimbursed care are not included in the definition. We believe the likelihood of this type of activity in private physician offices is small. Using free text fields respondents can report this type of activity.

The survey addresses benefits outside of coverage in options offered in Q4. The study does also include consultations via phone and email as an option in Q4.

12. Aren't physicians who do not spend at least 50% of their time caring for patients excluded?

The AAFP-NRN databases will be used to select the study sample. All physicians who report to the AAFP they practice less than half-time will be excluded. Recognizing that some physicians may have changes their practice arrangements or not shared this information with the AAFP, an early survey question asks about time in practice. The AAFP defines a full-time patient care provider week as containing 32 hours so anyone reporting less than 16 hours of patient care, will be excluded from the study. The study power calculations account for a small loss due to this exclusion.

13. Since the first question about unreimbursed care occurs here at question 3, it might be a good idea to remind respondents what "unreimbursed care" means for purposes of this survey. Otherwise, it's my hunch that physicians will tend to lump in care for which reimbursement was expected but not received.

The definition now appears at the beginning of the survey and again on the second page before question #4 the first question that refers to unreimbursed care.

14. Why does the question begin, "In the past"; it may be clearer to just begin with "Have you provided."

The study authors intended the language to represent unreimbursed care provided by a physician at a previous practice or someone who might have offered such care in the past but no longer does due to the factors outlined in the previous questions.

15. Q11 and 11a - What do the * represent?

The asterisks on question 11 were meant to link to Q11a and Q11b and skip patterns existed in the electronic survey to navigate the respondent based on their answers to question 11 (Q11, Q11a, & Q11b now exist as consecutively numbered questions 12-14). During pilot testing, this was found to be confusing. The questions have been renumbered and reworded to exclude any skip patterns.

16. Shouldn't there be a skip pattern so that only those people who are in group practices fill out page 15?

All physicians are expected to complete this section. The word group was initially used to distinguish between the full practice, which may include multiple locations, and the individual site in which a physician practices. In response to recent feedback, we have removed the word group and highlighted the explanation that follows the section header.

17a. A proposed response rate of 65% is a little low for this type of survey.

The anticipated response rate of 65% was based on a power calculation to obtain the minimum number of respondents for adequate analysis given the study questions. We expect a higher response rate because the respondents are all voluntary members of the AAFP organization and are being asked by this organization to take part in this project. The AAFP is employing a series of techniques to maximize response rates. The membership will be contacted through a variety of means, including follow-up phone calls, to enhance participation. We are currently estimating the actual response rate to be closer to 70% based on estimates from the AAFP which are the result of previous studies conducted within this organization.

17b. Are there plans for analysis of non-response bias? Please explain what those plans are.

We plan to conduct an analysis for non-response bias. We will use the demographic variables available from the AAFP database including age, sex, and census region. We may also consider components of care (child health, maternity care, hospital care, surgery), practice ownership, and practice type (solo, FM group, multispecialty).

Additional question:

OMB staff members asked if AHRQ would be willing to incorporate additional questions into the survey to assist HRSA in collecting information related to physicians knowledge and attitudes towards FTCA malpractice coverage available when volunteering at free clinics.

AHRQ shared the request with NORC and AAFP-NRN team members, who agreed to add three HRSA-proposed questions to the survey. The questions have been incorporated in the attached revised instrument.

5