Rehabilitation pathways after arthroscopic rotator cuff repair: a perspective from real-world multicentric cohort study data

Background The varied goals of postoperative management (PM) after arthroscopic rotator cuff repair (ARCR), including medication, immobilization, and rehabilitation, make recovery multifactorial and complex, which challenges the identification of prognostic factors for poor outcomes. The primary objective was to describe PM after ARCR and analytically identify groups with similar management pathways that could serve as prognostic strata in future analyses. The secondary objective was to compare the clinical outcomes of these groups at baseline, six and twelve months postoperatively. Methods Data were drawn from a Swiss-German multicentre ARCR cohort of 973 patients from 19 clinics. Standardized follow-ups at six weeks, six and twelve months captured medication use, immobilization practices, and rehabilitation details. Polytomous latent class analysis was applied to each PM domain to identify patient groups with similar management profiles. Group differences in Constant Scores (CS) up to twelve months were analysed using linear mixed models adjusted for relevant confounders. Results Two distinct groups were identified for both medication and immobilization management, and four for rehabilitation. Baseline severity and management differed markedly between groups. For example, patients immobilized in abducted positions with delayed mobilization had more severe tears and lower baseline CS. Despite these differences and some divergence in the rate of improvement, absolute Constant Scores converged across all groups by twelve months, with between-group differences remaining below the minimal clinically important difference of 10.4 points. Conclusion Using latent class analysis on real-world, patient-reported data, PM after ARCR could be structured into two medication, two immobilization, and four rehabilitation pathways. Despite differences in baseline severity and management approach, outcomes converged by twelve months, remaining below the minimal clinically important difference. These pathways reflected care driven by clinical presentation rather than standardized protocols and may serve as prognostic strata in future work, though causal conclusions cannot be drawn from this observational study.