Measurement Error, Minimal Detectable Change, and Minimal Clinically Important Difference of the Short Form-36 Health Survey, Hospital Anxiety and Depression Scale, and Pain Numeric Rating Scale in Patients With Chronic Pain

Abstract

In both pain research and clinical practice, patient-reported outcome measures are used to assess dimensions of health. Interpreting these instruments requires understanding their measurement error and what magnitude of change has subjective importance for patients. This study estimated the standard error of measurement, 1-year minimal detectable change, and 1-year minimal clinically important difference (MCID) for the Short Form-36 Health Survey physical component summary and mental component summary, the Hospital Anxiety and Depression Scale subscales for anxiety symptoms and depression symptoms, and the numeric rating scale for past-week average pain intensity. MCIDs for these instruments have not previously been estimated in a large sample of chronic pain patients participating in interdisciplinary pain rehabilitation. Data were drawn from the Swedish Quality Registry for Pain Rehabilitation (n = 8,854 patients). MCID was estimated as average change and change difference based on 3 different anchors. MCID estimates were 2.62 to 4.69 for Short Form-36 Health Survey physical component summary, 4.46 to 6.79 for Short Form-36 Health Survey mental component summary, .895 to 1.48 for numeric rating scale, 1.17 to 2.13 for anxiety symptoms in the Hospital Anxiety and Depression Scale, and 1.48 to 2.54 for depression symptoms in the Hospital Anxiety and Depression Scale. The common assumption of an identical standard error of measurement for pre- and post-treatment measurements was not always applicable. When estimating MCID, researchers should select an estimation method and anchor aligned with the study’s context and objectives. PerspectiveThis article presents estimates of MCID and minimal detectable change for several commonly used patient-reported outcome measures among patients with chronic pain. These estimates can help clinicians and researchers to determine when a measured health improvement is subjectively important to the patient and greater than measurement error.