Minimally important difference and responsiveness to change for numerical rating scale of menstrual pain severity: a psychometric study

Abstract

Background: Menstrual pain affects 45%-95% of reproductive-age females and increases the risk of other chronic pain conditions. Psychometrically sound measurement tools are essential for advancing research and clinical care in menstrual pain. Numerical rating scales (NRS) are widely used to measure pain severity. However, the minimally important difference (MID) and responsiveness to change of the NRS in the context of menstrual pain are not well understood. Understanding MID and responsiveness to change helps guide the evaluation of treatment efficacy and clinical decision-making. This study evaluated the MID and responsiveness to change in the NRS, ranging from 0 to 10, for menstrual pain severity. Methods: Participants who were menstruating (aged 14-42, N = 100) completed two surveys 24 h apart. In both surveys, we measured menstrual pain severity (worst, least, average menstrual pain in the past 24 h, and current menstrual pain) on a 0 (no pain) to 10 (extremely severe) NRS. MIDs were estimated using distribution-based approaches (standard error of measurement and effect size) and anchor-based approaches (using symptom interference and retrospective recall of change as anchors). Responsiveness to change was evaluated using standard response means and area-under-the-curve analysis. Results: The MID estimates were close to 1 point. The NRS of menstrual pain severity was responsive to menstrual pain improvement (standard response means ranged from 0.44 to 0.61, p < 0.001 for between-group comparisons). Area-under-the-curve estimates ranged from 0.66 to 0.70. Conclusions: The findings can inform the design and interpretation of studies testing interventions for menstrual pain, while also guiding clinicians in monitoring and adjusting treatment.