Abstract: Objective To analyze the risk factors for recurrence of primary spontaneous pneumothorax and to establish a prediction model. Methods The clinical data of 803 patients clearly diagnosed with primary spontaneous pneumothorax in the First Affiliated Hospital of Xinjiang Medical University from January 2010 to January 2021 were retrospectively analyzed,and 70% of the patients were randomly included in the modeling group (562 patients) and 30% in the validation group (241 patients).Risk factors for recurrence were analyzed by univariate and multivariate Cox regression using R 4.2.1 software,and a Nomogram prediction model was developed.Receiver operating characteristic curves were plotted,and the area under the curve (AUC) was calculated to assess model discrimination,and calibration curves were plotted to assess model calibration. Results The overall recurrence rate was 22.67% (182/803).Multivariate Cox regression analysis showed that age,smoking index,dystrophic severity score and treatment regimens were independent risk factors for recurrence of primary spontaneous pneumothorax,and the AUC of the Nomogram prediction model was 71.7% (95% CI 64.1-79.2),with high predictive efficiency. Conclusion This recurrence prediction model of primary spontaneous pneumothorax can assist clinicians to accurately assess the risk of recurrence in individual patients.

Key words: primary spontaneous pneumothorax, recurrence, risk factors, prediction model