Impact of response shift on change in patient-reported outcomes: a systematic review
摘要
The aim of this systematic review is to provide insight into whether response shift impacts the conclusions about change in patient-reported outcome measures (PROMs) in terms of statistical significance, magnitude, and decisions made.
MethodsResponse shift studies from Sawatzky et al. (2025; QLR) were analyzed: longitudinal quantitative studies that examined response shift using PROMs, published before May 2023. We determined whether: 1) impact of response shift was investigated, 2) information about change in the scores of a PROM (i.e., PROM-result) before and after taking response shift into account was provided, and 3) impact of response shift was evidenced in terms of statistical significance, magnitude, or decisions made.
ResultsA total of 173 response shift studies that included 943 PROM-results were evaluated. 55% of studies (N = 96) investigated impact of response shift and information about impact was available for 51% of studies (N = 89). The corresponding percentages based on PROM-results were 47% (N = 446) and 53% (N = 502), respectively. Impact of response shift was evidenced in 69% of studies (N = 61) and 41% of PROM-results (N = 207), where impact on statistical significance and/or magnitude of change were most often evidenced (49–53% of studies and 20–31% of PROM-results), whereas impact on decisions was evidenced less often (8% and 2% respectively).
ConclusionAbout half of the studies addressed impact of response shift and showed evidence of impact, however, impact on decisions was rarely addressed. Future research should focus on evaluating impact of response shift on conclusions about change in PROMs, especially in the context of healthcare decision-making.