Porthole
How do supplemental items affect HCAHPS response rates and scores? (306658)
Megan K Beckett, RANDJulie Brown, RAND
*Marc Nathan Elliott, RAND Corporation
Laura Giordano, HSAG
Elizabeth Goldstein, CMS
Katrin Hambarsoomian, RAND
William G Lehrman, CMS
Keywords: response rate, survey length, HCAHPS
We estimated the response rate (RR) effect of supplemental items (SIs) and whether it differs by survey mode, patient characteristics, or hospital characteristics; we then assessed whether any RR reductions result in changes to HCAHPS scores. We used data from 10,099 adult discharged Q1 2016 randomized within 51 national hospitals to 4 combinations of survey mode [Mail Only, Mixed Mode (MM; mail with telephone follow-up of nonrespondents)] and survey version (each with 32 core items plus 0 or 27 SIs]. Regression models predicted patient responses and HCAHPS scores from study arm, patient characteristics, and hospital characteristics. MM RRs were 21.8 percentage points higher than Mail Only with no SIs. 27 SIs were associated with a 5.6pp reduction in MM and a 2.9pp reduction in Mail Only. The RR reductions from SIs had no effect on standard patient-mix adjusted HCAHPS scores for either mode. HCAHPS users may want to consider RR trade-offs when adding many SIs; using MM may offset the RR effects of many SIs. Reducing RR within the same hospitals does not change patient-mix adjusted HCAHPS scores.