Development and validation of a nomogram for predicting hemoptysis risk in patients with bronchiectasis: A study based on Random Forest algorithm

Hemoptysis is a severe and potentially life-threatening complication of bronchiectasis. There is currently a lack of reliable tools for the individualized prediction of hemoptysis risk in these patients. This study aimed to utilize a machine learning algorithm to develop and validate a nomogram for predicting the risk of hemoptysis in patients with bronchiectasis. This retrospective study enrolled 131 patients with bronchiectasis, who were randomly divided into a training set and an internal validation set in a 7:3 ratio. The Random Forest algorithm was employed, using mean decrease accuracy as the metric, to screen for important predictors from clinical variables. Subsequently, the selected variables were incorporated into a multivariate logistic regression model to construct the predictive nomogram. The model’s discrimination, calibration, and clinical utility were comprehensively evaluated using the receiver operating characteristic curve, calibration curve, decision curve analysis, and clinical impact curve. The top 10 important predictors identified by the Random Forest algorithm included body mass index, C-reactive protein, creatinine, age, urea, low-density lipoprotein, platelet count, total cholesterol, diabetes, and fasting plasma glucose. Multivariate logistic analysis confirmed that body mass index, C-reactive protein, creatinine, age, and urea were independent predictors of hemoptysis. After correcting the mismatched logistic regression parameters, the constructed nomogram still demonstrated excellent predictive performance in both the training and validation sets, with an area under the curve of 0.944 (95% confidence interval: 0.873–1.000) in the validation set. The calibration curve showed a high degree of consistency between predicted probabilities and observed outcomes. Decision curve analysis and the clinical impact curve further confirmed that the model provided significant net clinical benefit across a wide range of threshold probabilities. This study successfully developed and validated a nomogram that incorporates Random Forest-based variable screening. The model integrates 5 readily available clinical parameters to accurately and individually predict the risk of hemoptysis in patients with bronchiectasis, demonstrating good calibration and clinical utility. It serves as a valuable tool to assist clinicians in the early identification and management of high-risk patients.