Nutrition Risk Assessed by STRONGkids Predicts Longer Hospital Stay in a Pediatric Cohort: A Survival Analysis.

Abstract

We evaluated the impact of Screening Tool for Risk on Nutritional Status and Growth (STRONGkids) classification in time to discharge and verify whether the nutrition risk assessed by this method is an independent predictor of hospital length of stay (LOS) in pediatric inpatients. A cohort study was conducted in a Brazilian hospital from February 2014 to July 2018. The outcome in the survivor analysis was hospital discharge. Kaplan-Meier curves were used to estimate the cumulative survival time according to STRONGkids categories. Multivariable Cox proportional hazard models were fitted, and the adjusted hazard ratio (aHR), with respective 95% CI, was used to measure the strength of association. The discriminatory ability of STRONGkids was verified by a receiver operating characteristic curve RESULTS: A total 641 patients were included in the study: 54.9% males, median age of 2.8 years. The frequencies of low, moderate, and high nutrition risk were 15.6%, 63.7%, and 20.7%, respectively. The mean LOS was 5.9 days. Survival curves differed significantly according to nutrition-risk categories. Patients classified as high risk had a 52% less chance of hospital discharge when compared with low-risk patients (aHR: 0.48; 95% CI, 0.35-0.65). STRONGkids score ≥ 3 showed the best discriminatory power to identify LOS. From this score, there was a significant increase in the days of hospitalization. The nutrition risk assessed by STRONGkids independently predicts LOS in pediatric patients. For this outcome, patients with 3 points (moderate risk) should be treated with the same priority as those with high risk.