Construction and Validation of a Regression Model for Predicting Hemothorax in Patients with Multiple Rib Fractures

Objective To analyze risk factors for hemothorax in patients with multiple rib fractures and establish a nomogram prediction model. Methods A modeling cohort of 280 patients with multiple rib fractures treated at our hospital from June 2022 to June 2024 was selected. Patients were divided into a hemothorax group ( n =70) and a non-hemothorax group ( n =210). Additionally, 120 patients treated at our hospital during the same period were collected as the validation group, also divided into a hemothorax group ( n =33) and a non-hemothorax group ( n =87) based on hemothorax occurrence. General clinical data were collected. Independent risk factors were identified using a multivariate logistic regression model, and a nomogram prediction model was constructed using the “rms” package in R software. Results In the modeling cohort, patients in the hemothorax group exhibited statistically significant differences compared to the non-hemothorax group in terms of history of hypertension, number of fractures, bilateral fractures, chest injury severity score, concomitant pulmonary contusion, flail chest, white blood cell count, hemoglobin, and D-dimer ( P < 0.05). In the validation cohort, patients with multiple rib fractures and complicated hemothorax showed statistically significant differences compared to the non-complicated group in terms of history of hypertension, number of fractures, bilateral fractures, chest injury severity score, concomitant pulmonary contusion, flail chest, white blood cell count, hemoglobin, and D-dimer ( P < 0.05). Bilateral fractures, flail chest, white blood cell count, hemoglobin, and D-dimer were independent risk factors for hemothorax in patients with multiple rib fractures (P < 0.05). A nomogram prediction model was constructed based on these factors. The Hosmer-Lemeshow goodness-of-fit test revealed χ² =13.303, P =0.102 for the modeling group and χ² =8.526, P =0.384 for the validation group, indicating good consistency and model fit. The decision curve analysis revealed that the predictive model provided additional clinical net benefit when the risk threshold was set between 0.08 and 0.96 in the modeling group and between 0.08 and 0.95 in the validation group. The receiver operating characteristic curve revealed an area under the curve (AUC) of 0.854 (95% CI : 0.794–0.914) for the modeling group and 0.835 (95% CI : 0.744–0.926) for the validation group, indicating good predictive performance. Conclusion The nomogram model constructed based on bilateral fractures, flail chest, white blood cell count, hemoglobin, and D-dimer provides important strategic guidance for the predictive assessment and clinical care intervention of hemothorax in patients with multiple rib fractures.