Responsiveness of five shoulder outcome measures at follow-ups from 3 to 24 months

BACKGROUND: To assess responsiveness of five outcome measures at four different follow-ups in patients with SLAP II lesions of the shoulder. METHODS: 119 patients with symptoms and signs, MRI arthrography and arthroscopic findings were included. The Western Ontario Shoulder Instability Index (WOSI), Oxford Instability Shoulder Score (OISS), EuroQol (EQ-5D3L), Rowe Score and Constant-Murley Score (CMS) were assessed at baseline, 3, 6, 12 and 24 months. The analysis contains both anchor-based and distribution-based methods, and hypothesis testing. RESULTS: Confidence intervals for ROC cut-off values, representing MID, for OISS, CMS and EQ-5D3L crossed zero at 3 months. Cut-off values were stable between 6- and 24-months follow-up. At 24-months ROC cut-off values (95% CI) were: Rowe 18 (13 to 24); WOSI 331 (289 to 442); OISS 9 (5 to 14); CMS 11 (9 to 15) and EQ-5D3L 0.123 (0.035 to 0.222). MID(95%limit) estimates were substantially higher than ROC cut-off values and MID(MEAN) at all follow-ups for all instruments. The reliable change proportion (RCP) values in the improved group were highest for WOSI and the Rowe Score (ranging from 68 to 87%) and significantly lower for CMS. EQ-5D3L had the lowest values (13 to 16%). We found a moderate correlation between mean change scores of the outcome measures and the anchor, except for the EQ-5D3L. CONCLUSIONS: In patients with SLAP II-lesions the patient reported OISS and WOSI and the clinical Rowe score had best responsiveness. Our results suggest that 3 months follow-up is too early for outcome evaluation.