P-639. Predictive Model to Identify Adult Patients with RSV at Risk of Developing Severe Acute Respiratory Infection

TL;DR

This study develops a clinical scoring system to predict which adult RSV patients are at risk of severe respiratory infection, based on factors like age and symptoms.

Contribution

A novel clinical prediction score was developed and validated to identify RSV patients at risk of severe acute respiratory infection.

Findings

Six independent risk factors for severe acute respiratory infection (SARI) in RSV patients were identified.

The clinical scoring system demonstrated excellent discriminative capacity with an AUC of 0.82.

The model was internally validated and stratified patients into four risk categories.

Abstract

Respiratory Syncytial Virus (RSV) infection in adults can evolve into severe conditions. Identifying risk factors for Severe Acute Respiratory Infection (SARI) would optimize clinical management. To develop and validate a predictive model to identify adult patients with RSV at risk of developing SARI.Independent Risk FactorsIndependent Risk Factors for SARI in Adult RSV PatientsClinical Prediction ScoreClinical Prediction Score for SARI Risk in Adult RSV Patients

Independent Risk Factors

Independent Risk Factors for SARI in Adult RSV Patients

Clinical Prediction Score

Clinical Prediction Score for SARI Risk in Adult RSV Patients

Analysis of 187 adults with PCR-confirmed RSV (Mexico Hospital, 2022-2024). Using multivariate logistic regression, independent factors associated with SARI were identified. A clinical scoring system was developed, and its diagnostic performance was evaluated through ROC analysis and internal validation by bootstrap.ROC CurveROC Curve for SARI Risk Prediction ModelCalibration PlotCalibration Plot of Predicted vs Observed SARI Risk

ROC Curve

ROC Curve for SARI Risk Prediction Model

Calibration Plot

Calibration Plot of Predicted vs Observed SARI Risk

Six independent risk factors for SARI were identified: age ≥80 years (OR 3.86, 95%CI 1.47-10.13), preexisting lung disease (OR 3.52, 95%CI 1.68-7.37), presence of dyspnea (OR 4.25, 95%CI 1.98-9.13), fever (OR 2.87, 95%CI 1.31-6.30), ≥3 comorbidities (OR 2.34, 95%CI 1.08-5.05), and temperature >36.5°C (OR 2.15, 95%CI 1.01-4.59). The scoring system assigned 2 points to age ≥80 years, lung disease, and dyspnea; 1 point to fever, ≥3 comorbidities, and temperature >36.5°C. The model showed excellent discriminative capacity (AUC 0.82, 95%CI 0.74-0.89) with sensitivity 76.9%, specificity 75.6%, and negative predictive value 89.5%. Internal validation confirmed the model's robustness (optimism-corrected AUC 0.79).

The developed scoring system effectively stratifies SARI risk in adult patients with RSV, with four categories: low risk (0-1 points, < 10%), moderate (2-3 points, 10-30%), high (4-5 points, 30-60%), and very high (6-9 points, >60%). This tool could optimize clinical decisions, facilitating identification of patients requiring closer monitoring or early intervention. External validation is required before widespread implementation.

All Authors: No reported disclosures

Figures