BACKGROUND: Optum conducts direct-to-patient mailed surveys among a claims-identified population, often combining patient-reported outcomes with claims-derived healthcare outcomes. Literature shows that response rates can vary greatly depending on methods of sample identification, administration mode, and response rate calculation. With growing interest in online data collection, it is valuable to evaluate the importance of mailed surveys and examine response rates within a claims-identified population to provide a benchmark for comparison with other recruitment strategies.

OBJECTIVES: To examine claims-identified direct-to-patient mailed survey response rates by insurance type and survey characteristics; and evaluate changes in response rates over time.

METHODS: A total of 30 direct-to-patient mailed survey studies fielded between 2007 and 2022 were examined, averaging 0 to 5 surveys annually across therapeutic areas. Patients were identified from the Optum Research Database (a large, national administrative claims database) using study-specific criteria and were recruited directly by mail for participation. Most surveys were 7-12 pages and utilized a post-paid $25 incentive. Insurance type (Commercial vs. Medicare Advantage), survey year, length, and other factors were assessed for impact on response rates.

RESULTS: Overall mean response rate was 27.0%; ranging from 19.2% (2021) to 33.3% (2018) with all studies meeting sample size targets. Response rates varied year-over-year (2007-2022). A small upward trend was seen over time but was statistically consistent with no change (p-value=0.598). Highest rates were observed in later years: 2018 (33.3%) and 2022 (30.4%). Response rates were higher among Medicare Advantage vs. Commercial patients (31.2% vs. 20.4%, respectively; p-value=0.025). Shorter surveys had higher response rates: <10 pages (31.11%), 10-12 pages (25.3%) and 13+ pages (22.0%) (p-value=0.032).

CONCLUSIONS: Response rates to direct-to-patient mailed surveys remained consistent over time demonstrating continued value in the methodology. More research is needed to understand how these response rates compare to other recruitment strategies within a claims-identified population.