A multidimensional preoperative model combining systemic inflammation and anatomical MRI predictors for shoulder function after rotator cuff surgery

Accurate preoperative identification of patients at risk for poor functional recovery after arthroscopic rotator cuff repair (ARCR) remains challenging. This study aimed to investigate the combined predictive value of the preoperative systemic immune-inflammation index (SII) and shoulder magnetic resonance imaging (MRI) features for functional recovery after ARCR and to develop a nomogram for individualized risk assessment. This single-center retrospective study included patients with full-thickness rotator cuff tears who underwent ARCR between January 2022 and January 2025. Preoperative MRI parameters (tear size, Patte grade, supraspinatus fatty infiltration grade, acromiohumeral distance [AHD], and tangent sign) were assessed, and SII was calculated from peripheral blood counts and analyzed per 100-unit increase. At 12 months postoperatively, patients were categorized by Constant-Murley score into ideal (≥ 80) and non-ideal (≤ 79) recovery groups. Independent predictors were identified using multivariate logistic regression, and a nomogram was constructed. Model performance was evaluated by receiver operating characteristic curves, the Hosmer–Lemeshow test, and bootstrap internal validation. A total of 406 patients were included, with 328 (80.8%) in the ideal recovery group and 78 (19.2%) in the non-ideal recovery group. Multivariate analysis identified preoperative SII (odds ratio [OR] = 1.001, 95% confidence interval [CI]:1.000–1.002, P = 0.001), supraspinatus fatty infiltration grade (OR = 1.673, 95%CI:1.15–2.49, P = 0.009), tear size (OR = 1.574, 95%CI:1.17–2.13, P = 0.003), Patte grade (OR = 2.098, 95%CI:1.26–3.48, P = 0.004), positive tangent sign (OR = 2.623, 95%CI:1.31–5.18, P = 0.006), and decreased AHD (OR = 0.798, 95%CI:0.69–0.92, P = 0.003) as independent predictors of poor recovery. Diabetes mellitus showed borderline significance ( P = 0.071). The nomogram achieved an area under the curve of 0.761 (95%CI:0.706–0.817) with good calibration (Hosmer–Lemeshow P = 0.074). Adding SII to the conventional MRI model increased the AUC from 0.738 to 0.761, but this improvement did not reach statistical significance ( P = 0.102). Preoperative SII, supraspinatus fatty infiltration, tear size, tendon retraction, supraspinatus muscle atrophy, and decreased AHD independently predict poor functional recovery after ARCR. The nomogram integrating these predictors provides a preliminary tool for individualized risk assessment, though external validation is needed. The incremental value of SII over traditional MRI parameters requires further evaluation in larger multicenter cohorts.