Identifying Prognostic Factors of Central Serous Chorioretinopathy Patients Treated With Micro-Pulse Laser Therapy and Development of a Predictive Model.

Aims/Background The prognosis of central serous chorioretinopathy (CSC) patients treated with micro-pulse is affected by various factors. Developing a predictive model can be helpful in assessing treatment efficacy and patient outcomes. This study investigated the key factors influencing CSC prognosis and developed a predictive model. Methods This retrospective study included 230 CSC patients who received micro-pulse laser therapy in the Second Affiliated Hospital of Zunyi Medical University, China, between January 2020 and May 2024. Using a random number table approach, patients were divided into a modeling group (n = 138) and a validation group (n = 92) in a 3:2 ratio. The modeling group was subdivided into a good prognosis group (n = 86) and a poor prognosis group (n = 52). Univariate and binary logistic regression analyses were performed to identify factors influencing poor prognosis in these patients, and a prediction model was constructed. Furthermore, the R programming language was employed to create receiver operating characteristic (ROC) curves. Moreover, calibration curves were built, and decision curve analysis (DCA) was used to assess the clinical utility of the model. Results There were no statistically significant differences in terms of general data among patients between the modeling group and validation group (p > 0.05). In the modeling group, the differences in age, body mass index (BMI), gender, residence, monthly family income, education level, number of leaks, and onset type were statistically insignificant (p > 0.05), while disease duration, leakage area, leakage type, and distance to leakage point showed statistically significant differences (p < 0.05). Binary logistic regression analysis revealed that disease duration, degree of retinal pigment epithelial detachment (PED), leakage type, and distance of leaks were significant factors influencing the prognosis of CSC patients undergoing micro-pulse laser therapy (p < 0.05). The model equation was Logit (P) = -2.217 + (1.059X1) + (1.344X2) + (1.258X3) + (1.494X4). The model demonstrated strong agreement between predicted and actual risks in both the modeling and validation groups, evidenced by calibration curves with slopes close to 1. In the modeling group, the model had an area under the curve of 0.942 (95% confidence interval (CI): 0.896-0.969, p < 0.001), with a standard error of 0.021 and a Youden index of 0.75, indicating a sensitivity of 93.6% and a specificity of 81.7%. In the validation group, the model's area under the curve was 0.860 (95% CI: 0.826-0.912, p < 0.001), with a standard error of 0.036 and a Youden index of 0.63, showing a sensitivity of 82.9% and a specificity of 79.8%. DCA revealed that the model had a significant positive net benefit, indicating good clinical utility. Conclusion Disease duration, degree of retinal PED, leakage type, and distance of leaks were significant prognostic factors in CSC patients undergoing micro-pulse laser therapy. A predictive model was successfully established and validated, enabling physicians to perform more accurate outcome assessments and facilitating tailored management approaches.