Objectives
To test methods to optimize mail survey response and authorization rates (i.e., allowing one's survey to be linked to medical records) when the survey is sensitive and the targeted population has a mental disorder.
Study Design and Setting
Randomized controlled trial of 1,099 veterans treated for posttraumatic stress disorder (PTSD) at a Midwestern Veteran Affairs Medical Center. Subjects were randomized to one of three groups: (1) Health Insurance Portability and Accountability Act (HIPAA) authorization form embedded in the survey (checking “yes” for consent); (2) HIPAA form requiring signature for consent sent after the survey; or (3) HIPAA form requiring a signature and social security number (SSN) for consent sent after the survey.
Results
The “embedded” strategy yielded the lowest survey response rate (67.5%) but highest authorization rate (59.1%). Requiring respondents' signatures and SSNs after the survey was returned generated the highest survey response rate (74.2%) but lowest authorization rate (48.7%). However, the response rate difference was not statistically significant. Reported emotional upset was low (1.5%) and primarily directed to the survey, not the HIPAA form. Older age and greater trustfulness were associated with higher authorization rates.
Conclusion
Even with our most optimized strategy, authorization rates were disappointingly low.