Psychometric evaluation of the PROMIS SD SF 8b and ISI for sleep disturbances associated with menopause

Abstract Background Sleep disturbances are common and impactful symptoms associated with menopause, negatively affecting women’s quality of life. Reliable and valid patient-reported outcome (PRO) instruments are needed to assess menopause-associated sleep disturbances in clinical trials. This study aimed to evaluate the psychometric properties of the PROMIS Sleep Disturbance Short Form 8b (PROMIS SD SF 8b) and Insomnia Severity Index (ISI) for this specific context of use. Methodology Measurement properties of the PROMIS SD SF 8b and ISI scores were assessed using data ( n = 110 participants) from a randomized, placebo-controlled, phase 2 study evaluating the efficacy and safety of elinzanetant for the treatment of sleep disturbances associated with menopause. Analyses included assessments of distributional properties, reliability (test-retest/score stability, internal consistency), validity (inter-item correlations, dimensionality, convergent/divergent, and known-groups), responsiveness to change, and minimal detectable change (MDC90). Results PROMIS SD SF 8b and ISI completion rates were generally high. No floor or ceiling effects were observed. For the PROMIS SD SF 8b T-score, test-retest reliability was excellent (intraclass correlation coefficients (ICCs) = 0.91), and acceptable internal consistency reliability was observed (Cronbach’s alpha = 0.91–0.94). The ISI total score showed good reproducibility among participants classified as stable between Weeks 4 and 12 (ICC = 0.82), and acceptable internal consistency reliability (Cronbach’s alpha = 0.86–0.88). Both scores demonstrated unidimensionality and mostly met convergent and divergent hypotheses for validity. Differences were observed between clinically distinct subgroups defined by Patient Global Impression of Severity items for severity of sleep problems and time spent awake ( p < 0.0001). Responsiveness analyses indicated both instruments could detect change, with large effect sizes in improved groups and correlations consistent with predefined hypotheses. MDC90 was 3.89 for the PROMIS SD SF 8b T-score and 3.99 for the ISI total score; individual changes exceeding these thresholds likely represent true change rather than measurement error. Conclusions Our findings support the reliability, validity and responsiveness of PROMIS SD SF 8b T-score and ISI total score for assessing sleep disturbances in women experiencing menopause. These findings support their use to assess clinical trial endpoints in this population and highlight the value of evaluating established PRO measures when applied in new contexts of use.