P0397 / #958: CONCORDANCE STUDY OF BERLIN AND PALICC DEFINITIONS FOR ACUTE RESPIRATORY DISTRESS SYNDROME WITH AUTOPSY FINDINGS IN PEDIATRIC PATIENTS.

Abstract

Aims & Objectives: Acute respiratory distress syndrome (ARDS) is an important cause of mortality. In pediatrics, its diagnosis is based on the Berlin consensus definition (BERLIN) and more recently on the Pediatric Acute Lung Injury Consensus Conference (PALICC). We measure the concordance of ARDS by BERLIN and PALICC definitions with autopsy findings in deceased patients from a pediatric intensive care unit (PICU). Methods: A cross-sectional study of clinical and autopsy records of patients who died in a PICU from a Hospital in Bogotá between 2000 and 2018. Concordance was measured with a Cohen Kappa coefficient test and a comparative analysis between definition components and diffuse alveolar damage (DAD) was performed and expressed as median with interquartile percentiles or relative frequencies with a Chi2 o a K-Wallis test were appropriate. Results: In a total of 222 autopsies, 53 (23.9%) met BERLIN criteria, 56 (25.2%) PALICC and 45 (20.2%) had DAD. Agreement was 0.48 for BERLIN and PALICC in relation to DAD. DAD cases were only significantly associated with bilateral opacities on chest X ray (77.8% vs. 56.0%; p = 0.025) and the presence of severe values in oxygenation (p = 0.04) or saturation (p = 0.04) indices. Conclusions: The agreement between ARDS definitions and DAD is moderate and has not changed with the most recent adjustment in the definition. Bilateral opacities and high oxygenation or saturation index are related to DAD. Models seeking therapeutic strategies should limit cases to these two findings to improve their specificity.