# Novel model to predict risk of invasive fungal infection and fungal prophylaxis timing

## TL;DR

A new model predicts fungal infection risk in immunocompromised patients and shows that targeted antifungal treatment can reduce infections.

## Contribution

A nomogram model using immune indicators predicts IFI risk and guides optimal timing for fungal prophylaxis in immunocompromised patients.

## Key findings

- The model includes age, IgG level, and CD4+ cell count as predictive indicators with an AUC of 0.723.
- Fluconazole prophylaxis in high-risk patients significantly reduced IFI incidence compared to placebo.
- The model's negative predictive value was 90.2%, indicating strong reliability in identifying low-risk patients.

## Abstract

Patients on immunosuppressive drugs or those who are critically ill are at high risk for invasive fungal infections (IFIs). The assessment of IFI risk and the initiation of prophylaxis in these patients remain unclear. A nomogram model was developed to evaluate clinical and immune indicators in relation to IFI risk and validated by immunocompromised patients. High-risk patients, as identified by the model, were selected for a prospective randomized study to assess the efficacy of the model and timing of fungal prophylaxis initiation. Patients deemed high risk received either fluconazole or a placebo until IFI occurrence or risk downgrade for 90 days. We compared the incidence of IFI and mortality between the two groups. The nomogram, created from a training cohort (n = 384), included age, IgG level, and CD4+ cell count as predictive indicators of IFI and was validated in a separate cohort (n = 281) with an area under the curve of 0.723. A total of 265 patients were recruited into the prospective study, with 163 high-risk patients randomly assigned to receive either fluconazole (n = 83) or a placebo (n = 80). The model had a positive predictive value of 48.8% and a negative predictive value of 90.2%. High-risk IFI defined by this model could be reduced to the low-risk cohort level with fluconazole prophylaxis (P < 0.01). This nomogram model reliably predicts the risk of IFI and timing of mycoprophylaxis in immunocompromised patients. Targeted mycoprophylaxis significantly reduces the incidence of IFI, particularly yeast infections, and may help to prevent fungal colonization.

We still lack enough evidence to decide when and how to begin fungal prophylaxis in immunocompromised patients. A model based on immune function indicator is an effective tool for predicting risk of invasive fungal infection (IFI) in immunocompromised patients. Patients were collected to evaluate the performance of the model retrospectively and prospectively.

This study is registered with the Chinese Clinical Trial Registry as ChiCTR2400079810.