Evaluation of nomogram prediction model for postoperative deep vein thrombosis in patients with non-small cell lung...

Non-small cell lung cancer (NSCLC) patients undergoing surgery face substantially elevated risks of postoperative deep vein thrombosis (DVT), a complication that significantly impairs recovery and survival. This study aims to develop and validate a nomogram for predicting postoperative DVT in NSCLC patients.

A retrospective analysis of the clinical data was conducted, including 120 NSCLC patients with DVT and 140 without DVT. Univariate and multivariate logistic regressions identified the risk factors for postoperative DVT in NSCLC patients, and these were integrated into a nomogram using R software. External validation was performed in an additional 174 patients.

Univariate screening identified six significant predictors: advanced age (≥70 years), diabetes mellitus, adenocarcinoma histology, stage III/IV disease, presence of neoadjuvant chemotherapy, and thoracotomy approach. Multivariate refinement established four independent determinants: geriatric age [odds ratio (OR) =4.631; 95% confidence interval (CI): 1.213-17.682], diabetes (OR =5.102; 95% CI: 1.167-22.316), adenocarcinoma subtype (OR =7.460; 95% CI: 1.817-30.626), and stage III pathology (OR =5.159; 95% CI: 1.407-18.913). The resulting nomogram demonstrated exceptional predictive accuracy with area under the curve (AUC) values of 0.900 (95% CI: 0.839-0.960) in the derivation cohort and 0.903 (95% CI: 0.854-0.951) upon external validation. Model calibration proved satisfactory through Hosmer-Lemeshow goodness-of-fit testing (P=0.54).

This validated nomogram achieves superior discrimination (AUC >0.90) vs. existing risk tools while using only routine clinical parameters. It enables stratified thromboprophylaxis, optimizing resource-limited settings.