AHRQ memo 5-8-09

AHRQ/OMB passback – Data Invetory

Below are AHRQ’s responses to your pass back.  I have also attached the introductory letter that was requested in one of the comments, as well as a revised version of the questionnaire incorporating  OMB’s suggestions.

 

OMB Comment 1:  “Supporting statement (SS) page 4 says that interviews will be conducted to collect information not readily available in the public domain.  Is there an interview schedule/script that Econometrica plan(s) to use? What is the sample for these interviews?” 

 

AHRQ response: The interview/script is that submitted with the OMB Supporting Statement as attachment C.  The respondent pool for this survey is not a sample.  The respondent pool is a census of all state and federal databases found to have outpatient data.  This identification was made in Phase I of this project. 

 

OMB Comment II:  “SS page 5: what will the initial respondent contact look like?” 

 

AHRQ response: The initial contact will be the notification signed by Dr. Clancy which will be sent to all respondent:   A copy is attached to this email.

 

OMB Comment III:  “SS page 6:  where on the survey is respondent identity information collected?  And does privacy act really apply here?  Also, the statutes that give AHRQ the authority to provide assurances of confidentiality should be cited on the cover letter that initially recruits respondents to participate in the study.  The word ‘confidential’ should not be used if this is not iron clad (e.g., if there are situations where AHRQ would be compelled by law to disclose the responses, these should be clearly stated).”

 

AHRQ response: Individual respondents will not be identified, only organizations.   We do not believe AHRQ can be compelled by law to disclose the respondent’s identify. 

 

OMB follow up: Please confirm with HHS general counsel whether AHRQ has statutory authority to provide assurances of confidentiality. If AHRQ does have such statutory authority, please cite the statute in the cover letter as well as in the recruitment letter.

 

Otherwise, please remove all assurances of confidentiality and replace with something to the effect of “AHRQ will keep responses private to the extent permitted by law.”

 

AHRQ response: As you have suggested, we have removed all assurances of confidentiality and have replaced it with  “AHRQ will keep responses private to the extent permitted by law,”  as in Section A.10 of the Supporting Statement Part A.

 

OMB follow up: There is still one reference to “confidentiality” on page 1 of the questionnaire. Can we change that as follows? If so, OMB can conclude review on this ICR. Thanks.

The information you provide will be aggregated with information from other respondents and kept private to the extent permitted by law. However, your specific responses will not be identified separately in any reports nor will any information about your responses be shared with AHRQ or otherwise identified as provided by you. If you have any questions or concerns about the survey or confidentiality, please contact Richard Hilton at [email protected].

AHRQ response: Fine. We will change

OMB Comment IV:  “SS page 6:  Who are the 80 respondents and why is the sample limited to 80?  Has some kind of sampling plan been undertaken (e.g., snowball sample) to identify further respondents?”

 

AHRQ response: As stated above the respondent pool is a census.  Snowball sampling was considered as a project component but not included because of budget considerations.

 

OMB Comment V:  “Academy Health public comment:  Academy health submitted the following suggestion: “We believe the quality and utility of this activity could be enhanced if AHRQ were to expand the analysis to determine the extent to which the data are currently available to the research community, and if not, whether the data could be made available” This does appear to be (a) salient comment.  How ill AHRQ determine the extent the current data is adequate and available to the needs of the health services research community? 

 

AHRQ response: The Inventory of Outpatient Data, which is the end product of this project, will be a resource the research community can learn about available sources of outpatient data.  Our project’s Final Report will identify “gaps” in currently available data and any barriers to making more such data available.

 

OMB Comment VI:  “Survey page 3:  Question #4 seems a bit off and out of context.  Does AHRQ currently review the information in various databases?  If so, it seems like AHRQ should be able to know what data is currently in the databases without asking respondents to tell AHRQ what’s in the database.”

 

AHRQ response: To our understanding our survey of public use data sets was the first rigorous effort on the part of AHRQ to identify such data.  Also, the purpose of Question #4 is primarily to confirm that the information already obtained is accurate. 

 

OMB follow up: With the way this question is worded, this is still not clear. Is it more direct to simply ask “how often do you update your outpatient data?” or something to that effect?

 

AHRQ response: >As you have suggested, we have replaced this question with “How often do you update the information in your database? This will help AHRQ to make sure that your agency’s data collection and reporting information presented in its Inventory is accurate, complete, and up to date.”

 

OMB Comment VII:  “Survey page 7:  to follow up on Academy Health’s comment, we would suggest adding a question about whether the respondent makes their data available to the health research community and if so, under what circumstances, what are the requirements, etc.  If they don’t make it available, it would be helpful to better understand why(e.g., are there regulations in place either at the federal or state level that prohibit data sharing, is it the institution’s own concerns about privacy, etc.)?”

 

AHRQ response: AHRQ agrees with this suggestion.  We have revised the questionnaire and include these questions on page 8 as questions 13a and 13b.  The revised instrument is attached.

 

OMB Comment VIII:  “Survey Page 8:It would be worth clarifying that these questions are soliciting general questions about the availability and accessibility of outpatient data as a whole, and not specifically about the respondents’ own databases.”

 

AHRQ response: AHRQ agrees with this suggestion. We have revised the questionnaire and include this clarification in the first paragraph that appears on page 9.

 

OMB Comment IX:  “Survey Page 8: are the respondents the best people to address the question whether the data available is sufficient for purposes of policy makers, clinicians, and consumers?  Are policy makers, clinicians, and consumers part of the respondent pool?”

 

AHRQ response: AHRQ’s project design is based on the knowledge that, currently, there is no centralized information source from which policy makers, clinicians and consumers can find out about the availability of outpatient data.  The purpose of the survey is to maximize the usefulness of the planned Data Inventory by making the first iteration of the Inventory as complete and comprehensive as possible.  The survey will also enable AHRQ to establish effective procedures to update and expand the Inventory in the future. 

 

OMB follow up: While OMB thinks this is a very worthwhile purpose, it is still unclear how the question will get you the information you need to maximize the usefulness of the inventory. What percentage of the respondents are going to be policy makers, what percentage will be clinicians, and what percentage will be consumers? Let’s take a hypothetical example: if the majority of the respondents are, say, statisticians, what is the utility of asking statisticians whether they think the data is sufficient for policy makers, clinicians, and consumers?

AHRQ response: >We do not interview individuals, but data organizations. We will interview some organizations that represent consumers, some that represent clinicians, and some that represent policy makers. In fact, most of the organizations that we will interview will have all three—policy-makers, consumers, and clinicians---as their stakeholders. Thus, it is in their general interest and mission to know what policy-makers, consumers, and clinicians really need in terms of data. All of the organizations interviewed supply data to policy-makers, consumers, and clinician.