Undernutrition at PICU Admission Is Predictor of 60-Day Mortality and PICU Length of Stay in Critically Ill Children.

Background: There are few studies that assess the role of different nutritional assessment variables at pediatric intensive care unit (PICU) admission in predicting clinical outcomes.

Objective: To identify nutritional variables in the first 4 days of PICU stay that predict 60-day mortality and time to discharge alive from the PICU.

Design: Single-center prospective study in Southern Brazil, conducted between July 2013 and February 2016. At PICU admission, children with z scores <-2 for body mass index (BMI)-for-age, mid-upper arm circumference (MUAC)-for-age, and triceps skinfold thickness (TSF)-for-age were considered as undernourished.

Participants/setting: There were 199 patients, aged <15 years, with PICU stay >48 hours.

Main Outcome Measures: Sixty-day mortality and time to discharge alive from the PICU.

Statistical Analysis Performed: Cox regression model was applied to determine predictors of 60-day mortality and time to discharge alive from the PICU.

Results: Median age was 23.1 months (interquartile range=3.9 to 89.1), and 63% were male, with 18% prevalence of undernutrition at admission by BMI-for-age. Median PICU stay was 7 days (interquartile range=4 to 12), and 60-day mortality was 12%. After adjusting for sex, age, Pediatric Index of Mortality 2, and presence of complex chronic conditions, undernutrition based on BMI-for-age (hazard ratio [HR]=3.75; 95% CI=1.41 to 9.95; P=0.008), MUAC-for-age (HR=7.62; 95% CI=2.42 to 23.97; P=0.001), and TSF-for-age (HR=4.01; 95% CI=1.14 to 14.15; P=0.031) was associated with higher risk of 60-day mortality. Based on MUAC-for-age with the same adjustment model, undernourished children had longer time to discharge alive from the PICU (HR=0.45; 95% CI=0.21 to 0.98; P=0.045).

Conclusions: Undernutrition at PICU admission based on different anthropometric variables was predictive of 60-day mortality and longer time to discharge alive from the PICU.