ObjectiveTo construct a prediction model for the postoperative recurrence risk of granulomatous lobular mastitis (GLM) based on multiple systemic inflammatory indicators and clinicopathologic characteristics, with the aim of guiding clinical treatment. MethodsThe GLM patients who underwent lesion resection at Sichuan Provincial Maternal and Child Health Hospital from January 2017 to March 2024 were retrospectively collected. The univariate and multivariate logistic regression analyses were used to screen the risk factors for recurrence after GLM lesion resection, and a nomogram prediction model was constructed based on the risk factors. The test level was set at α=0.05. ResultsA total of 533 patients with GLM were included in this study, and 118 cases (22.1%) developed postoperative recurrence. The results of multivariate analysis showed that the no taking oral bromocriptine, having microabscess formation in postoperative pathological examination results, systemic immune inflammation index (SII) >789.0×109/L, and immunoglobulin E (IgE) >64.4 U/mL were the independent risk factors for recurrence after GML. Based on this, the nomogram predicting recurrence risk was constructed. The area under the receiver operating characteristic curve (95%CI) was 0.913 (0.895, 0.932), and its sensitivity and specificity were 90.5% and 88.9%, respectively. The calibration curve showed that the probability of recurrence after GML predicted by using the nomogram was highly consistent with the actual occurrence probability. The decision curve analysis showed that the nomogram had a good clinical net benefit. ConclusionsThe results of this study suggest that no taking bromocriptine orally, having microabscess formation in postoperative pathological examination results, SII, and IgE are risk factors for recurrence in GLM patients undergoing lesion resection. The nomogram model for predicting postoperative recurrence of GLM constructed based on these risks has good efficacy and provides a reference for the early treatment and management of GLM.