Nomogram-based prediction of postoperative proliferative vitreoretinopathy following scleral buckling surgery for rhegmatogenous retinal detachment

To identify independent risk factors for postoperative proliferative vitreoretinopathy (PVR) in patients with primary rhegmatogenous retinal detachment (RRD) treated with scleral buckling surgery and to develop a nomogram for predicting postoperative PVR.

Patients who underwent scleral buckling surgery for primary RRD were retrospectively enrolled. Patients were randomly assigned to a training cohort (n=515) and a validation cohort (n=55). Candidate variables included demographic characteristics, systemic comorbidities, preoperative ocular status, and retinal break features. Independent predictors of postoperative PVR were identified using univariate and multivariate logistic regression analyses. A nomogram was constructed to predict the risk of PVR at 1, 3, and 6mo after surgery. Model performance was evaluated using the concordance index (C-index), receiver operating characteristic (ROC) curves, calibration plots, and decision curve analysis (DCA).

A total of 570 eyes were included, with mean age of 52.65±15.70y, and 51.4% of patients were male. Postoperative PVR occurred in 28.8% of eyes overall, including 28.5% in the training cohort and 30.9% in the validation cohort. Postoperative PVR developed in a subset of patients during follow-up. Preoperative PVR, a history of diabetes mellitus, and more than four retinal breaks were identified as independent risk factors for postoperative PVR. In the training cohort, the C-indices of the nomogram at 1, 3, and 6mo were 0.888, 0.931, and 0.948, respectively, and 0.885, 0.885, and 0.909 in the validation cohort. ROC and calibration analyses demonstrated good discrimination and agreement, while DCA showed favorable net clinical benefit across a wide range of threshold probabilities.

Preoperative PVR, diabetes mellitus, and more than four retinal breaks are independent predictors of postoperative PVR after scleral buckling for RRD. The proposed nomogram provides accurate individualized risk prediction at 1, 3, and 6mo postoperatively, and may assist clinicians in postoperative surveillance and decision-making.