Poster Discussion Abstracts
Aims & Objectives:
To compare the classifications of (p-RIFLEand AKIN) for predicting AKI in pediatric DKA.
Setting: Pediatric Critical Care Division of a tertiary care institute. Subjects: Children aged ≤14 years with DKA as per ISPAD-2014 guideline. Intervention: Analyzed the prospectively collected data of 69 patients managed between Oct-2014 to April-2017. AKI presence and severity was assessed for each classification using the change in serum creatinine and estimated creatinine clearance (eCCl) levels calculated by the modified-Schwartz formula (=0.413xLength in cm / SCr in mg/dl). For baseline eCCl, the lowest value of SCr in the preceding 3-month of admission, if not, eCCl assigned of 100 ml/min/1.73m2.
Mean±SD of age was 99 ± 45 months and male:female is 20:49. Mild, moderate and severe DKA was present in 20 (29%),20 (29%) and 29 (42%) respectively. AKI was present in 64 (93%; R48%,I42%,F3%) and 59 (86%; I41%,II26%,III19%) according to p-RIFLE and AKIN criteria respectively. Intraclass correlation between two system for identification and the severity of AKI was 91%(95%CI 86% to 94%, p< 0.001). According to the severity of DKA, identification of AKI was similar between the p-RIFLE and AKIN (mild 80% vs.70%; moderate 100% vs. 95% and severe 97% vs. 90%). No difference in mean±SD time to resolution of DKA in p-RIFLE(25.4 ± 18.6 vs. 19.6 ± 5.2;p=0.099) and AKIN (21.3 ± 6.5 vs. 25.6 ± 19.2; p=0.196) criteria among AKI versus no AKI patients.
p-RIFLE and AKIN criteria show excellent consistency in the identification of AKI in pediatric DKA. p-RIFLE shows early detection of AKI and AKIN shows over diagnosis of the severe type of AKI.