Don’t middle your MIDs: regression to the mean shrinks estimates of minimally important differences
Minimal important differences (MIDs) for patient-reported outcomes (PROs) are often estimated by selecting a clinical variable to serve as an anchor. Then, differences in the clinical anchor regarded as clinically meaningful or important can be used to estimate the corresponding value of the PRO. Although these MID values are sometimes estimated by regression techniques, we show that this is a biased procedure and should not be used; alternative methods are proposed.
KeywordsMinimally important difference Clinical significance Quality of life Patient-reported outcomes Regression to the mean
Minimal clinically important difference
Minimal important difference
- NEI VFQ-25
Eye Institute Visual Function Questionnaire-25
Ron D. Hays was supported in part by grants from the NIA (P30-AG021684) and the NIMHD (P20MD000182).
- 2.US Food and Drug Administration. (2009). Patient-reported outcome measures: Use in medical product development to support labeling claims. Guidance for industry. http://www.fda.gov/downloads/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/UCM071975.pdf. Accessed March 20, 2013.
- 9.Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum.Google Scholar
- 11.Fayers, P. M., & Hays, R. D. (2013). Linking should replace regression when mapping from profile to preference-based measures. Value in Health (submitted).Google Scholar