Supporting Statement Part B 0720-0034_2015-0921

OMB—SUPPORTING STATEMENT – PART B

DoD PATIENT SAFETY CULTURE SURVEY

B. Collection of Information Employing Statistical Methods

1. Description of the Activity

The proposed project will administer the patient safety culture survey as a web-based instrument to a census of all staff, both clinical and non-clinical, working in Continental United States (CONUS) and internationally (OCONUS) Military Health System (MHS) facilities. There are a total of approximately 154,000 MHS staff in CONUS and OCONUS (estimate from FY 2014 TRICARE Report to Congress). Of these staff, approximately 23,100 staff (about 15%) are contractors, local nationals, volunteers or other MHS staff who are not direct employees of the DoD. With the goal of an overall response rate of approximately 60%, responses would be received from approximately 14,022 U.S. and International MHS staff who are not direct employees of the DoD (contractors, local nationals, volunteers, other non-DoD MHS staff). The survey takes about 10 minutes to complete. All survey responses are voluntary and will be individually anonymous; only group-level results will be tabulated to protect individual anonymity. The response rate for the last administration of this survey (2011) was 43%.

While we recognize that OMB guidance requests a description of how non-response bias will be addressed, conducting a non-response analysis in this case is not feasible. Because no sampling frame of the population will be created, and the population universe is somewhat unknown, it is difficult if not infeasible to compare respondents to non-respondents. Additionally, because of the sensitivity of the survey topic, the Patient Safety Program places emphasis on anonymity; to ensure this anonymity, each MTF is responsible for distributing the survey link to staff members rather than the contractor receiving a sampling frame and distributing survey invitations. Each MTF reports the total number of staff eligible to take the survey so that a response rate can be calculated, but no demographic information or any other characteristics are provided. We will take the measures described elsewhere in this document to encourage a high response rate and will calculate response rates at the population level.

2. Procedures for the Collection of Information

The proposed project will administer the patient safety culture survey as a web-based instrument to a census of all staff, both clinical and non-clinical, working in CONUS and OCONUS MHS facilities. To standardize the process across MHS facilities, data collection will be simultaneously conducted at Army, Navy and Air Force MHS facilities. This survey will be administered on a voluntary-basis to all staff working in Military Treatment Facilities (MTF) hospitals and ambulatory and dental services. Responses and respondents will remain anonymous. There are two versions of the survey that may be administered, corresponding to the setting in which care is delivered, either Hospital (inpatient) or Medical Office (ambulatory/outpatient/clinic setting). All responses will be collected through the use of internet-based information technology and there will be no hard copy surveys administered. Using a web-based survey will accommodate the CONUS and OCONUS survey dissemination, shorten the field period necessary for data collection and decrease the need for data cleaning since only valid responses can be entered and automatic storage of responses can occur in an electronic database.

3. Maximizing Response Rates

The 60% response rate goal was estimated based off actual response rate from the last time the survey was administered in 2011 (43%). This study design will implement standard techniques to maximize response rates Potential respondents will receive a pre-notification letter followed by an email survey notification containing an embedded hyperlink to the internet location where the survey can be completed. Several reminder notifications will be sent, approximately one week to 10 days apart, so that the data collection field period will be approximately four weeks for each Service.

To maximize response rates, the following steps will be taken:

1) a pre-notification letter will be sent to participants detailing the purpose of the survey;

2) a first survey email notification will be emailed including a hyperlink to the survey web site where they can complete the web survey,

3) Reminder email notification(s) will be sent with hyperlink to the survey web site, and;

4) A final notification will be sent with hyperlink to the survey website before the survey closes.

To improve respondent trust of anonymity and increase the likelihood of individual response, no individual identifiers will be used. Therefore, all follow-up survey notifications will be sent to everyone; those who have already completed the web survey will be thanked for their participation and asked to disregard additional follow-up notices.

4. Tests of Procedures

The Survey on Patient Safety Culture, hospital and medical office versions, to be used in this data collection was previously developed and piloted tested by the creators, Agency for Healthcare Research and Quality (AHRQ). Pilot data on the hospital version were gathered from over 1,200 staff from 21 hospitals. After data analysis, the survey was revised and the final survey has demonstrated reliability and validity. This hospital survey has been used in three previous data collections (2005, 2008, and 2011) in MHS facilities. Similarly, the medical office version was pilot tested by AHRQ across 200 medical offices and data were gathered from over 4,100 staff. Based on pilot data, the medical office version was revised and has demonstrated reliability and validity. The medical office version was utilized in the MHS within the Air Force in 2011 and will be used again in this 2015 survey process.

5. Name of Contractor Collecting and Analyzing Information for Agency

Deloitte Consulting LLP, has been contracted to conduct this web survey data collection and analyze the survey feedback results. The individuals assigned to the project include: Patrick Koeppl, Senior Manager (pkoeppl@deloitte.com) and Sarah Godby, Staff Advisor (sgodby@deloitte.com). No other persons have consulted on statistical aspects of this survey nor will any other persons either collect or analyze the data.

3