External Validation of the "Quick" Pediatric Logistic Organ Dysfunction-2 Score Using a Large North American Cohort of Critically Ill Children With Suspected Infection

Abstract

Objectives: A quick Pediatric Logistic Organ Dysfunction 2 score on day 1, consisting of tachycardia, hypotension, and altered mentation, was shown to predict mortality with an area under the receiver operating characteristic curve of 82% (95% CI, 76-87%) in children admitted to a PICU with suspected infection. We performed an external validation of the quick Pediatric Logistic Organ Dysfunction 2, including its performance in predicting mortality in specific age groups.

Design: Analysis of retrospective data obtained from the Virtual Pediatric Systems PICU registry.

Setting: Prospectively collected clinical records from 130 participating PICUs in North America.

Patients: Children admitted between January 2009 and December 2014, with a diagnosis of infection at discharge, for whom all required data were available.

Interventions: None.

Measurements and main results: Systolic blood pressures, heart rates, and Glasgow Coma Scale scores were used to evaluate the quick Pediatric Logistic Organ Dysfunction 2 using area under the receiver operating characteristic curve analysis. Performance was compared with Pediatric Risk of Mortality 3 and Pediatric Index of Mortality 2 risk scores. Data from 42,196 children with complete data were analyzed, with median age 2.7 years (interquartile range, 0.7-8.8 yr; range 0-18 yr) and a 4.27% mortality rate. Mortality was 13.4% for quick Pediatric Logistic Organ Dysfunction 2 greater than or equal to 2 and 2.5% for quick Pediatric Logistic Organ Dysfunction 2 less than 2, representing a false-negative rate of 49.5%. Also 311 children (17%) who died had a quick Pediatric Logistic Organ Dysfunction 2 score of 0. The area under the receiver operating characteristic curve was 72.6% (95% CI, 71.4-73.8%) for quick Pediatric Logistic Organ Dysfunction 2, compared with 85.0% (95% CI, 84.0-86.0%) for Pediatric Risk of Mortality 3 and 81.5% (95% CI, 80.5-82.5%) for Pediatric Index of Mortality 2. Performance of quick Pediatric Logistic Organ Dysfunction 2 was worst in the greater than 12 years age group (area under the receiver operating characteristic curve, 67.8%; 95% CI, 65-70.5) and best in the less than 1 month age group (area under the receiver operating characteristic curve, 78.9%; 95% CI, 75.3-82.4).

Conclusions: Quick Pediatric Logistic Organ Dysfunction 2 performed markedly worse in our cohort, compared with the original study, and the high rate of false negatives limits its clinical utility in our population. Further work is needed to develop a robust quick pediatric sepsis diagnostic tool for both research and clinical care.