Construction and Validation of a Nomogram to Identify the Risk of Cavitation in Pulmonary Tuberculosis.

Abstract

The present study aimed to construct and validate a nomogram based on clinical metrics to identify CPTB. The present study retrospectively recruited pulmonary tuberculosis (PTB) patients admitted to Jiashan County First People's Hospital in China from November 2018 to September 2023. PTB patients were classified into the CPTB group and the non-CPTB group based on chest computed tomography findings, and were randomly allocated to the training set (70%) and the validation cohort (30%). The training set and validation set were used to establish and validate nomogram, respectively. Multivariate logistic regression analysis (MLSA) was used to identify the independent risk factors for CPTB in patients with PTB. Statistically significant variables in the MLSA were then used to construct a nomogram predicting CPTB in patients with PTB. The receiver operating characteristic (ROC) curve, calibration curve analysis (CCA), and decision curve analysis (DCA) were used for the evaluation of the nomogram. A total of 293 PTB patients, including 208 in the training set (85 CPTB) and 85 in the validation set (33 CPTB\), were included in this study. Stepwise MLSA showed that sputum smear (≥2+), smoking(yes), glycosylated hemoglobin A1c(HbA1c), hemoglobin (HB), and systemic inflammatory response index (SIRI) were independent risk factors for the development of cavitation in patients with PTB. The nomogram identifying the high-risk CPTB patients was successfully established and showed a strong predictive capacity, with area under the curves (AUCs) of 0.875 (95% CI:0.806-0.909) and 0.848 (95% CI:0.751-0.946) in the training set and validation set respectively. In addition, the CCA and DCA corroborated the nomogram's high level of accuracy and clinical applicability within both the training and validation sets. The constructed nomogram, consisting of sputum smear positivity, smoking, HbA1C, HB, and SIRI, serves as a practical and effective tool for early identification and personalized management of CPTB.