Construction of a nomogram model based on multiple factors to differentiate cryptococcal meningitis from tuberculous meningitis in HIV-Infected patients

Abstract

To develop and validate a nomogram model for differentiating cryptococcal meningitis (CM) from tuberculous meningitis (TBM) in HIV-infected patients, given the diagnostic challenges due to shared clinical manifestations and limitations of existing methods. A retrospective analysis extracted 207 HIV cases (112 CM, 95 TBM). Candidate predictor variables covering general information, blood biochemical, and cerebrospinal fluid(CSF) examination indicators were collected. Least absolute shrinkage and selection operator (LASSO) regression and ten-fold cross-validation identified key predictors, which were used to construct and validate the nomogram model. Model performance was evaluated by the area under the receiver operating characteristic (ROC) curve (AUC), calibration curves, and decision curve analysis (DCA). SHapley Additive exPlanations (SHAP) values were used to interpret the characteristics of the model's predictor variables. Five predictors (extracranial tuberculosis, extracranial fungi, erythrocyte sedimentation rate, albumin, and CSF pressure) were included in the final nomogram. The model achieved AUC of 0.830 (95% CI: 0.758-0.902) in the training set and 0.811 (95% CI: 0.719-0.904) in the testing set, with good calibration and clinical validity shown by calibration curves and DCA. The developed nomogram model effectively distinguishes CM from TBM in HIV-infected patients. It aids clinicians in diagnosis decisions.

Keywords: Cryptococcal meningitis; HIV; LASSO regression; Nomogram; Precision medicine; Tuberculous meningitis.