Deciphering Prognostic Indicators in Non-HIV Cryptococcal Meningitis: Constructing and Validating a Predictive Nomogram Model

Abstract Cryptococcal meningitis (CM) is a well-recognized fungal infection, with substantial mortality in individuals infected with the human immunodeficiency virus (HIV). However, the incidence, risk factors, and outcomes in non-HIV adults remain poorly understood. This study aims to investigate the characteristics and prognostic indicators of CM in non-HIV adult patients, integrating a novel predictive model to guide clinical decision-making. A retrospective cohort of 64 non-HIV adult CM patients, including 51 patients from previous studies and 13 from the First Hospital of Shanxi Medical University, was analyzed. We assessed demographic features, underlying diseases, intracranial pressure, cerebrospinal fluid characteristics, and brain imaging. Using the least absolute shrinkage and selection operator (LASSO) method, and multivariate logistic regression, we identified significant variables and constructed a Nomogram prediction model. The model's calibration, discrimination, and clinical value were evaluated using the Bootstrap method, calibration curve, C index, goodness-of-fit test, receiver operating characteristic (ROC) analysis, and decision curve analysis. Age, brain imaging showing parenchymal involvement, meningeal and ventricular involvement, and previous use of immunosuppressive agents were identified as significant variables. The Nomogram prediction model displayed satisfactory performance with an akaike information criterion (AIC) value of 72.326, C index of 0.723 (0.592–0.854), and area under the curve (AUC) of 0.723, goodness-of-fit test P = 0.995. This study summarizes the clinical and imaging features of adult non-HIV CM and introduces a tailored Nomogram prediction model to aid in patient management. The identification of predictive factors and the development of the nomogram enhance our understanding and capacity to treat this patient population. The insights derived have potential clinical implications, contributing to personalized care and improved patient outcomes.