Concomitant acromioplasty is associated with enhanced structural healing, long-term durability, and functional outcomes in arthroscopic rotator cuff repair: a retrospective propensity score and causal machine-learning analysis

Background The clinical value of acromioplasty (AC) combined with arthroscopic rotator cuff repair (RCR) remains controversial. This study evaluates the associations and heterogeneity of treatment response for concomitant AC during arthroscopic RCR. Methods We retrospectively analysed 734 patients (220 AC, 514 RCR-only). Propensity scores were estimated via a SuperLearner ensemble, followed by inverse probability of treatment weighting to balance covariates. Structural healing, functional outcomes (American Society of Shoulder and Elbow Surgeons and Constant–Murley Shoulder scores), and re-tear events were assessed at 12 months. Doubly robust machine-learning approaches (DR-learner, causal forest, Bayesian additive regression trees) were employed to estimate the average and individual treatment effects. Shapley Additive Explanations (SHAP) were used to improve model interpretability. Results Acromioplasty was associated with improved structural healing (risk difference: −7.8%; 95% confidence interval [CI], −11.9% to −3.6%; number needed to treat [NNT] ≈ 13). Weighted Kaplan-Meier analysis showed significantly lower cumulative re-tear incidence at 60 months (AC: 45.9% vs. RCR-only: 80.0%; hazard ratio: 0.41; 95% CI: 0.33 to 0.51; p < 0.001). Functional scores were significantly higher with AC (ΔASES = 7.9; ΔConstant = 7.5; both p < 0.001), although these differences did not exceed established minimal clinically important difference thresholds. Considerable treatment heterogeneity was identified: patients with a high critical shoulder angle (CSA) (>35°) and older age (>62 years) showed the strongest associations with potential benefits. The SHAP analysis identified age, body mass index, tear size, and CSA as key factors influencing outcomes. Conclusion Concomitant AC is associated with enhanced structural healing and long-term durability in arthroscopic RCR. Although functional scores were statistically higher in the AC group, the differences did not exceed established MCID thresholds, indicating limited clinical significance for functional outcomes. Significant treatment heterogeneity highlights the importance of individualised surgical strategies, particularly favouring patients with high CSAs and of older age.