SS Part B -- A Survey of Physicians in Solo Smaller PCPs 07 23 12

SUPPORTING STATEMENT

Part B

A Survey of Physicians in Solo and Smaller Primary Care Practices

Version July 23, 2012

Agency for Healthcare Research and Quality (AHRQ)

Submitted under AHRQ’s generic pretesting clearance 0935-0124

Table of contents

B. Collections of Information Employing Statistical Methods 3

1. Respondent universe and sampling methods 3

2. Information Collection Procedures 5

3. Methods to Maximize Response Rates 6

4. Tests of Procedures 6

5. Statistical Consultants 6

B. Collections of Information Employing Statistical Methods

1. Respondent universe and sampling methods

This application includes three stages of data collection to inform the final survey design of the survey of physicians in solo and smaller primary care practices. This section describes the sample frame and sampling methods for each of the three stages.

Cognitive testing:

The draft questionnaire will be subjected to cognitive testing with up to 25 physicians in solo and smaller practices (see Attachment A). This is a convenience sample of 25 physicians in solo or smaller practices. In addition these physicians include Internal Medicine (IM) and Family Practice (FP) physicians only. These physicians will be recruited through market research agencies with physician contacts, through the professional contacts of members of the National Integration Academy Council and associations such as the American Association of Family Practitioners (AAFP) and American Medical Association (AMA).

Pilot Survey:

For this pilot study, AHRQ is defining primary care physicians to include Internal Medicine (IM) and Family Practice (FP) physicians. AHRQ will use the National Plan and Provider Enumeration System (NPPES) maintained by the Centers for Medicare and Medicaid Services (CMS) that contains National Provider Identifier (NPI) records. A subset of the database, referred to as the NPI file, will be used as the source for the sample frame. Table 1 provides a count of FPs and IMs as of June 2010. This count is similar to the count from the AMA implying that the file provides a complete national coverage of physicians.

Table1: Numbers of FPs and IMs with an NPI

Physician Specialty	NPI File Count June 2010
Family Medicine	100,959
Internal Medicine	111,051

For the pilot study the sample will consist of physicians from states believed to be using varied care management models to support their efforts to integrate care and the delivery of behavioral health services. These are: Colorado, California, Maine, North Carolina, Texas, Maryland, Virginia, Louisiana, Illinois, and Kansas. This is a purposeful selection of States, based on recommendations from the National Integration Academy Council (NIAC).

In each of these States, AHRQ will select an equal probability sample of physicians from among all the physicians in the state, as identified by the address information associated with the physician available on the sample frame. Prior to sample selection, the NPI file will be sorted by a flag in the file indicating whether a physician is in a practice with the legal status of “sole proprietorship”. “Sole proprietorships” represent a subset of practices consisting of only a single physician. We will examine the percentage of physicians in each stratum that are flagged as “sole proprietorship”. If there is a non-negligible percentage, we may sort first on this flag, and then on ZIP code within “sole proprietorship” status. For sampled physicians not flagged as “sole proprietorships” the address information associated with the physician will be used to identify whether the physicians is in a solo, smaller (10 physicians or less) or larger practice. Any address associated with 11 or more physicians will be removed from the sample. This may still include physicians in larger practices located at different addresses. Those physicians will be screened out in the questionnaire.

FPs and IMs will be sampled as separate strata within each state in order to better assess possible differences in terms of eligibility and/or response rates between the two specialties. The target is 15 responding physicians per specialty per state; resulting in 300 completed interviews. The sample for initial release is 25 physicians from each specialty within each state, implicitly assuming the overall yield rate (accounting for both 75-80% eligibility and 75-80% response rate) to be about 60 percent. There will also be a reserve sample that can be released as needed, if the yield rates are somewhat lower than 60 percent. Table 2 provides the sample allocation by State.

Table2: Sample allocation by State

State	Sampled family practice physicians	Sampled internal medicine physicians	Total sampled physicians	Total physician yield
Colorado	25	25	50	30
California	25	25	50	30
Maine	25	25	50	30
North Carolina	25	25	50	30
Texas	25	25	50	30
Maryland	25	25	50	30
Virginia	25	25	50	30
Louisiana	25	25	50	30
Illinois	25	25	50	30
Kansas	25	25	50	30
Total	250	250	500	300

Follow-up interviews:

For the follow-up interviews physicians will be sampled from the population of physicians that completed the pilot survey. This sample of physicians will also be a convenience sample since the sample will be randomly selected from only those physicians that agreed to allow additional study contact. From among those who agreed to follow-up contact 30 physicians will be selected based on their responses to questions 16 and 19 of the pilot survey. The selection will be based on the following criteria:

1. Physicians who do not screen or assess behavioral health conditions- Select 4 Physicians who meet this criterion;

2. Physicians who treat with only medication- Select 10  physicians who meet this criterion; and

3. Physicians who treat with both medication and some form of counseling - Select 16 physicians who meet this criterion.

2. Information Collection Procedures

This section discusses the information collection procedures for the cognitive interview, pilot survey and follow-up interviews

Cognitive Interview:

These interviews will be conducted over the telephone. Prior to the interview each recruited physician will be sent a copy of the questionnaire by email. If the physician prefers a paper copy they will be sent a paper copy by first class mail to reach them a few days before the interview.

Pilot Survey:

Physicians are known to be challenging respondents. They are, as expected, busy with patients and do not have time to complete a survey. Based on survey methods research experience, a mail methodology works best with this population. The mailing, addressed to the sampled physician, will include a cover letter, the questionnaire, and a postage paid reply envelope.

This will be followed in ten days by a reminder/thank you postcard. And ten days later by a second survey packet to non-responders. A third survey packet will be sent ten days later to remaining non-responders. If 300 surveys are not returned in the course of the four mailings a telephone call and a mailing by Express (Overnight) mail will be used as a prompt to get responses. Physicians with a telephone number on file will receive a telephone call and those with no telephone number will receive another copy of the questionnaire by Express (Overnight) mail.

Follow-up Interviews:

These interviews will be conducted over the telephone. At the time of scheduling the interviews respondents will be told that they will receive a copy of their completed questionnaire either by mail, fax or email. They will be sent a copy in the mode of their choice.

3. Methods to Maximize Response Rates

Cognitive Interviews:

Due to the nature of the convenience sample, the physicians selected for the cognitive interviews will be those committed to responding. Further the recruitment effort will continue until 25 interviews are completed. However, as the time commitment for the interviews is significant, Physicians will be sent an incentive of $150 for cognitive interviews.

Pilot Study:

The methods to maximize response rates for the pilot study include:

1. Short survey instrument: about 10 minutes in length.

2. A large 9x10 survey envelope with an appealing design.

3. Non-responders are being followed-up with at least three additional reminders or re-mailed surveys.

4. A pre-paid $25 gift card for the physician’s practice that will be sent out with the first mailing.

Follow-Up Interviews:

Due to the nature of the convenience sample, the physicians selected for the follow-up interviews will be those committed to responding and therefore maximizing response rates is not applicable to this population. However, as the time commitment for the interviews is significant an incentive of $75 for follow-up interviews is being provided.

4. Tests of Procedures

The purpose of this submission is to conduct rigorous testing of the survey instrument and methodology. To prepare a draft questionnaire for submission, the attached draft data collection instrument (Attachments A and B) has been tested with five physicians. While this testing provided direction for the attached draft, it is insufficient to provide robust information to design the questionnaire.

5. Statistical Consultants

The National Integration Academy Council and Westat are being consulted on statistical aspects of the study design. Below are all those persons involved with this study who are affiliated with either Westat or the National Integration Academy Council who have provided input on the statistical aspects of the study design.

Westat	Garrett Moran	Phone: 301-294-3821 email: [email protected]
	Rebecca Noftsinger	Phone: 240-453-5636 email: [email protected]
	Vasudha Narayanan	Phone: 301-294-3808 email: [email protected]
	Paul Weinfurter	Phone: 714-262-1856 email: [email protected]
	Ralph DiGaetano	Phone: 301-294-2062 email: [email protected]
Project Staff	Benjamin F. Miller University of Colorado School of Medicine	Phone: 303-724-9706 email: [email protected]
	Deborah Cohen Oregon Health and Science University	Phone: 503-494-7840 email: [email protected]
NIAC Members	Frank DeGruy University of Colorado School of Medicine	Phone: 303-724-9753 email: [email protected]
	Teresa Chapa Office of Minority Health, HHS	Phone: 240-453-6904 email: [email protected]
	Macaran A. Baird University of Minnesota Medical School	Phone: 612-624-4641 email: [email protected]
	Dave DeBronkart e-patient Dave	Phone: 603-459-5119 email: [email protected]
	Alexander Blount University of Massachusetts Medical School	Phone: 774-443-2147 email: [email protected]
	Michael Hogan New York State Office of Mental Health	Phone:518-474-4403 email: [email protected]
	Roger Kathol Cartesian Solutions, Inc.	Phone: 952-426-1626 email: [email protected]
	Ned Calonge The Colorado Trust	Phone: 303-837-1200 email: [email protected]
NIAC Members continued	Parinda Khatri Cherokee Health System	Phone: 865-765-0304 email: [email protected]
	C.J. Peek University of Minnesota	Phone: 612-626-3860 email: [email protected]
	Neil Korsen Maine Health	Phone: 207-675-3515 email: [email protected]
	Jürgen Unützer University of Washington	Phone: 206-543-3128 email: [email protected]
	Stephen P. Melek Milliman	Phone: 303-672-9093 email: [email protected]
	Steven E. Waldren American Academy of Family Physicians	Phone: 913-906-6000 x4100 email: [email protected]
	Kavita K. Patel The Brookings Institution	Phone: 202-797-2475 email: [email protected]