Derivation and Validation of the Fungal Pneumonia Assessment and Likelihood Predictor Score

Overview

The FUNGAL-P study is a single-center observational study designed to derive and validate a clinical prediction score for the early identification of fungal pneumonia in adult patients presenting with pneumonia. The study includes a retrospective derivation cohort and a prospective validation cohort of patients undergoing microbiological evaluation of lower respiratory tract samples. Clinical, laboratory, radiological, microbiological, and treatment-related variables associated with fungal pneumonia will be analyzed to identify independent predictors of fungal infection. These predictors will be combined to develop the FUNGAL-P score. The derived score will subsequently be evaluated in a prospective validation cohort to assess its diagnostic performance, calibration, and clinical utility. The ultimate goal is to facilitate earlier recognition of fungal pneumonia and support timely diagnostic testing and antifungal treatment in patients presenting to the Emergency Department or hospital with pneumonia.

Fungal pneumonia is an increasingly recognized cause of severe respiratory infection and is associated with substantial morbidity and mortality. Early diagnosis remains challenging because clinical manifestations are often nonspecific and conventional diagnostic criteria may not be readily applicable in emergency care settings. Delayed recognition may result in delayed diagnostic investigations and initiation of targeted antifungal therapy. The FUNGAL-P study is a no-profit, single-center observational study conducted at Careggi University Hospital (Florence, Italy). The study was designed to derive and validate a clinical prediction rule for the early identification of fungal pneumonia or fungal-bacterial coinfection among adult patients presenting with pneumonia. The study consists of two sequential phases. Retrospective derivation phase The derivation cohort includes adult patients with microbiologically confirmed pneumonia diagnosed between January 1, 2022 and December 31, 2023. Eligible patients underwent bronchoalveolar lavage (BAL), bronchial aspirate, or endotracheal aspirate collection within 48 hours of hospital presentation. Microbiological investigations included fungal and bacterial cultures, galactomannan testing, molecular assays for Aspergillus species and Pneumocystis jirovecii, and multiplex molecular respiratory pathogen testing. Patients with microbiologically confirmed fungal pneumonia constituted the study group, whereas patients with bacterial or viral pneumonia served as controls. Cases of fungal-bacterial coinfection were classified as fungal pneumonia. Patients without microbiological identification of a pathogen were excluded. Prospective validation phase The validation cohort includes consecutive adult patients with pneumonia enrolled between January 1, 2024 and December 31, 2024. Patients underwent routine microbiological investigations according to standard clinical practice. The performance of the derived FUNGAL-P score was evaluated prospectively without influencing clinical decision-making. Outcome definition The primary study outcome is fungal pneumonia, including infections caused by Aspergillus species, Pneumocystis jirovecii, and other fungal pathogens, as well as fungal-bacterial coinfections. Diagnosis is based on an integrated assessment of clinical presentation, radiological findings, microbiological results, fungal biomarkers, and molecular testing results. Final case adjudication includes review of clinical follow-up data, imaging studies, microbiological findings, and treatment decisions. Data collection Demographic characteristics, medical history, comorbidities, fungal infection risk factors, vital signs, laboratory findings, radiological data, microbiological results, administered treatments, and clinical outcomes are collected for all participants. Statistical analysis Independent predictors of fungal pneumonia are identified using multivariable logistic regression. Candidate variables are selected based on prior evidence and univariable analyses. Predictors retained in the final model are incorporated into the FUNGAL-P score. Diagnostic performance is evaluated using sensitivity, specificity, predictive values, likelihood ratios, receiver operating characteristic (ROC) curves, and area under the ROC curve (AUROC). Calibration is assessed using calibration plots and the Hosmer-Lemeshow test. Internal validation is performed using bootstrap resampling, and clinical utility is evaluated through decision curve analysis. Primary Objective To derive a clinical prediction score for fungal pneumonia or fungal-bacterial coinfection in adult patients presenting with pneumonia. Secondary Objectives To prospectively validate the FUNGAL-P score. To assess score calibration and discrimination. To determine sensitivity, specificity, positive predictive value, and negative predictive value. To evaluate the clinical utility of the score for identifying patients at increased risk of fungal pneumonia.