Validation of the PIDS/IDSA Severity Criteria in Children with Community-Acquired Pneumonia

Abstract

Background

The Pediatric Infectious Diseases Society (PIDS)/Infectious Diseases Society of America (IDSA) guideline for community-acquired pneumonia (CAP) recommends intensive care unit (ICU) admission or continuous monitoring for children meeting severity criteria. Our objective was to validate these criteria.

Methods

This was a retrospective cohort study of children age 3 months-18 years diagnosed with CAP in a pediatric Emergency Department (ED) from 9/2014-8/2015. Children with complex chronic conditions and recent ED visits were excluded. The primary predictor was the PIDS/IDSA severity criteria. Outcomes included disposition (i.e., admission vs. discharge) and interventions/diagnoses necessitating hospitalization (i.e., need for hospitalization [NFH]).

Results

Of 518 children, 56.6% were discharged. 54.3% of discharged patients and 80.8% of those hospitalized <24 hours were classified as severe. Of those admitted, 10.7% did not meet severity criteria. Overall, 69.5% (n=360) met PIDS/IDSA severity criteria. Of these children, 73.1% did not demonstrate NFH. The areas under the receiver operator characteristic curves (AUC) for PIDS/IDSA major criteria were 0.63 and 0.51 for predicting disposition and NFH, respectively. For PIDS/IDSA minor criteria, the AUC was 0.81 and 0.56 for predicting disposition and NFH, respectively. The sensitivity, specificity, LR+ and LR- of the PIDS/IDSA criteria were 89%, 46%, 1.65, and 0.23 for disposition and 95%, 16%, 1.13, and 0.31 for NFH.

Conclusions

More than half of children classified as severe by PIDS/IDSA criteria were not hospitalized. The PIDS/IDSA CAP severity criteria have only fair ability to predict need for hospitalization. New predictive tools specifically for children are required to improve clinical decision making.