Prediction of pulmonary infection during chemotherapy in non-small cell lung cancer: development and internal validation of a nomogram.

OBJECTIVE: To develop and internally validate a nomogram for predicting pulmonary infection during chemotherapy in patients with non-small cell lung cancer (NSCLC). METHODS: This retrospective study included 246 NSCLC patients receiving cytotoxic chemotherapy. The primary outcome was pulmonary infection during chemotherapy. Patients were randomly divided into training (n=196) and validation (n=50) cohorts. Predictors were selected using AIC-guided forward stepwise logistic regression to construct a nomogram. Model performance was evaluated by discrimination, calibration, threshold-based classification metrics, and decision curve analysis, with internal validation performed using bootstrap resampling. RESULTS: Pulmonary infection occurred in 55 patients (22.4%). The final model included five predictors: absolute neutrophil count, serum albumin, age, combined immunotherapy, and chronic obstructive pulmonary disease. The apparent AUC was 0.786 (95% CI 0.703-0.869), with an optimism-corrected AUC of 0.735 and a validation AUC of 0.765 (95% CI 0.562-0.968). The nomogram classified patients into low-, intermediate-, and high-risk groups with infection rates of 5.4%, 18.2%, and 37.5%, respectively. Decision curve analysis suggested modest clinical net benefit within clinically relevant threshold probability ranges. CONCLUSION: This internally validated nomogram, based on routinely available pre-chemotherapy variables, may provide a practical approach for preliminary pulmonary infection risk stratification in patients with NSCLC.