The minimum detectable measurement difference for the Scoliosis Research Society-22r in adolescent idiopathic scoliosis: a comparison with the minimum clinically important difference.

Background Context: The minimal clinically important difference (MCID) is the smallest change in an outcomes instrument deemed relevant to a patient. MCID values proposed in spine research are limited by poor discriminative abilities to accurately classify patients as "improved" or "not improved." Furthermore, the MCID should not compare relative effectiveness between two groups of patients, though it is frequently used for this. The minimum detectable measurement difference (MDMD) is an alternative to the MCID in outcomes research. The MDMD must be greater than the MCID for the latter to be of value and the MDMD can compare change between groups.

Purpose: The purpose of this study was to determine the MDMD for the Scoliosis Research Society-22r (SRS-22r) in adolescent idiopathic scoliosis (AIS) patients treated with surgery.

Study Design: Retrospective cohort study from multi-center registry.

Patient Sample: Patients treated surgically for AIS.

Outcome Measures: Self-reported SRS-22r.

Methods: An observational cohort of surgically treated AIS patients was queried for patients with complete baseline, 1-year, and 2-year SRS-22r data. The MDMD was calculated for SRS-22r domain and subscores. Effect size (ES) and standardized response mean were calculated to measure responsiveness of the SRS-22r to change. MDMD values were compared with MCID values. Research grants were received from DePuy Synthes Spine, EOS imaging, K2M, Medtronic, NuVasive, and Zimmer Biomet to Setting Scoliosis Straight Foundation.

Results: One thousand two hundred and eighty-one AIS patients (1,034 female, 247 male, mean age 14.6 years) were analyzed. MDMD values were between 0.23 and 0.31. SRS-Pain MDMD was 0.3, greater than the MCID of 0.2. SRS-Activity MDMD was 0.24, greater than the MCID of 0.08. SRS-self-image MDMD was 0.3, less than the MCID of 0.98. Sixty-four percent of those with baseline SRS-self-image>4.0 improved MDMD or more, whereas only 14% improved beyond the MCID. ES and standardized response mean were highest for subscore and self-image.

Conclusions: The MDMD can compare the relevance of change in SRS-22r scores between groups of AIS patients. SRS-pain and SRS-activity MDMD values are greater than the MCID and should serve as the threshold for clinically relevant improvement. MDMD may help evaluate change in patients with baseline self-image>4.0.