Establishing the minimal clinically important difference for the UCLA functional scale in patients undergoing clinical treatment for adhesive capsulitis.

Abstract

The minimal clinically important difference (MCID) helps interpret whether changes in patient-reported outcomes reflect meaningful benefit. For adhesive capsulitis, an MCID for the University of California, Los Angeles (UCLA) Shoulder Score has not been established for patients treated nonoperatively. We performed a retrospective observational cohort study of consecutive patients with idiopathic adhesive capsulitis treated between 2020 and 2024 at a single outpatient center. A standardized conservative program (supervised physiotherapy with home exercises; analgesic optimization) was used; 42% additionally received serial suprascapular nerve blocks as part of a separate RCT protocol. The UCLA Shoulder Score and pain measured by visual analog scale (VAS) were collected at baseline and 2, 4, and 6 months (primary endpoint). MCID was estimated using (1) an anchor-based method with the UCLA satisfaction item (satisfied vs not satisfied), (2) distribution-based indices (0.5 SD; minimum detectable change [MDC] assuming r=0.84), and (3) receiver operating characteristic (ROC) analysis with Youden index. Multivariable logistic regression explored predictors of 6-month satisfaction. A total of 226 patients completed 6-month follow-up (mean age 54.2±9.2 years; 66.7% female). Median UCLA improved from 18.0 to 35.0 (p<0.001); VAS decreased from 8.0 to 0.0 (p<0.001). Range of motion improved significantly in all planes (all p<0.05). At 6 months, 200/226 (88.5%) were satisfied. Median UCLA change was 17 points (Interquartile Range - IQR 4) in satisfied vs 8 points (IQR 6.5) in unsatisfied patients; the between-group median difference was 9 points (95% CI 6-10; p<0.001), defining the anchor-based MCID. ROC analysis identified a 14-point improvement as optimal (AUC 0.900; sensitivity 90.5%; specificity 84.6%), whereas distribution-based estimates were smaller (0.5 SD=1.41; MDC=3.12). Integrating anchor-based and ROC findings, we recommend an MCID of 11.5 points for the UCLA score in nonoperative adhesive capsulitis. In multivariable analysis, higher baseline UCLA was the strongest independent predictor of patient satisfaction (OR = 1.37, 95% CI 1.13-1.66; p=0.001) CONCLUSION: In patients undergoing conservative management for adhesive capsulitis, we recommend a UCLA Shoulder Score improvement of 11.5 points as the clinically meaningful threshold, with anchor-based (9 points) and ROC (14 points) analyses providing convergent validity. This benchmark can guide clinical interpretation, sample-size planning, and comparative effectiveness studies in frozen shoulder.