• Francis Leclerc, Alain Duhamel, Valérie Deken, Bruno Grandbastien, Stéphane Leteurtre, and Groupe Francophone de Réanimation et Urgences Pédiatriques (GFRUP).
• 1CHU Lille, Service de réanimation pédiatrique, Lille, France. 2Univ. Lille, EA 2694 Santé publique, épidémiologie et qualité des soins, Lille, France. 3CHU Lille, Département de biostatistiques, Lille, France. 4CHU Lille, Service de gestion du risque infectieux, des vigilances et d'infectiologie (SGRIVI), Lille, France.
• Pediatr Crit Care Me. 2017 Aug 1; 18 (8): 758-763.

ObjectiveA recent task force has proposed the use of Sequential Organ Failure Assessment in clinical criteria for sepsis in adults. We sought to evaluate the predictive validity for PICU mortality of the Pediatric Logistic Organ Dysfunction-2 and of the "quick" Pediatric Logistic Organ Dysfunction-2 scores on day 1 in children with suspected infection.DesignSecondary analysis of the database used for the development and validation of the Pediatric Logistic Organ Dysfunction-2.SettingsNine university-affiliated PICUs in Europe.PatientsOnly children with hypotension-low systolic blood pressure or low mean blood pressure using age-adapted cutoffs-and lactatemia greater than 2 mmol/L were considered in shock.Measurements And Main ResultsWe developed the quick Pediatric Logistic Organ Dysfunction-2 score on day 1 including tachycardia, hypotension, and altered mentation (Glasgow < 11): one point for each variable (range, 0-3). Outcome was mortality at PICU discharge. Discrimination (Area under receiver operating characteristic curve-95% CI) and calibration (goodness of fit test) of the scores were studied. This study included 862 children with suspected infection (median age: 12.3 mo; mortality: n = 60 [7.0%]). Area under the curve of the Pediatric Logistic Organ Dysfunction-2 score on day 1 was 0.91 (0.86-0.96) in children with suspected infection, 0.88 (0.79-0.96) in those with low systolic blood pressure and hyperlactatemia, and 0.91 (0.85-0.97) in those with low mean blood pressure and hyperlactatemia; calibration p value was 0.03, 0.36, and 0.49, respectively. A Pediatric Logistic Organ Dysfunction-2 score on day 1 greater than or equal to 8 reflected an overall risk of mortality greater than or equal to 9.3% in children with suspected infection. Area under the curve of the quick Pediatric Logistic Organ Dysfunction-2 score on day 1 was 0.82 (0.76-0.87) with systolic blood pressure or mean blood pressure; calibration p value was 0.89 and 0.72, respectively. A score greater than or equal to 2 reflected a mortality risk greater than or equal to 19.8% with systolic blood pressure and greater than or equal to 15.9% with mean blood pressure.ConclusionAmong children admitted to PICU with suspected infection, Pediatric Logistic Organ Dysfunction-2 score on day 1 was highly predictive of PICU mortality suggesting its use to standardize definitions and diagnostic criteria of pediatric sepsis. Further studies are needed to determine the usefulness of the quick Pediatric Logistic Organ Dysfunction-2 score on day 1 outside of the PICU.

26 keywords...

hide…