A nomogram for assessing the risk of deep vein thrombosis following surgery for pilon fractures

PURPOSE: Deep vein thrombosis (DVT) can occur following fractures of the lower extremities. This study investigated DVT incidence following pilon fractures and analyzed the related risk factors. A nomogram was developed and validated for predicting the likelihood of postoperative DVT. METHODS: From January 2017 to January 2023, data on 1689 patients with pilon fractures at Hebei Medical University's Third Hospital was obtained for this study. The DVT group consisted of DVT cases, while members of the non-DVT group did not have DVT. In order to collect information on demographics, comorbidities, injuries, and laboratory biomarkers, the inpatient medical record system was consulted. The multivariate logistic regression analysis incorporated factors found to be significant (p < 0.05) in the univariate analysis. Independent factors predictive of DVT were identified with a backward stepwise regression approach. Predictors were analyzed in R with the construction of a nomogram. RESULTS: The DVT rate (166 of 1238) in the current study was 13.4%. Using univariate analysis, several DVT predictors were identified, including smoking history (p < 0.001), open fracture (p = 0.001), fracture blister (p < 0.001), body mass index (BMI) (p < 0.001), and Elixhauser comorbidity index (ECI) (p < 0.001). Furthermore, globulin levels were considerably greater in the DVT group relative to the non-DVT group (p = 0.001), while platelets counts were lower (p < 0.001). Higher BMI (p < 0.001, odds ratio (OR) = 0.837, 95% confidence interval (CI) (0.763 - 0.919)), open fracture (p < 0.01, odd ratio (OR) = 0.431, 95% CI (0.286 - 0.649)), fracture blister (p = 0.018, OR = 0.589, 95% CI (0.380 - 0.912)), smoking history (p = 0.011, OR = 0.561, 95% CI (0.360 - 0.874)), and a higher ECI (p < 0.001, OR = 0.776, 95% CI (0.681 - 0.884)) were all strongly associated with DVT, as per the stepwise logistic regression analysis. Five variables independently linked to DVT were the basis for developing the nomogram. CONCLUSIONS: The present research indicates that pilon fractures are more susceptible to DVT than typical ankle fractures, attributed to their high-energy injury characteristics. It was found that BMI, the existence of an open fracture, the lack of blisters, a smoking history, and ECI are independently predictive of DVT following surgery in pilon fracture cases. Consequently, the current study developed a nomogram based on these discovered indicators to facilitate early targeted interventions.