Oral

INTRODUCTION: Refeeding syndrome (RFS) occurs after overaggressive nutritional rehabilitation in malnourished patients. It may cause electrolyte imbalance that may potentially lead to significant morbidity and mortality. The American Society for Parenteral & Enteral Nutrition (ASPEN) 2020 committee emphasized that studies are needed to identify risk factors for RFS in children who received parenteral nutrition (PN). To date, little is known about the risk factors apart from what was extrapolated from the experiences with enteral nutrition (EN), which remains uncertain of its applicability.

OBJECTIVES: Our study aimed to prevent RFS in children receiving PN by identifying risk factors and formulating a nomogram for risk stratification.

METHODS: This was a retrospective case-control study done at University Malaya Medical Centre (UMMC) that included 100 patients aged 1-month to 18-year-old who received PN for at least 24 hours from 2019 to 2021. Study variables were compared between patients who developed RFS and those who did not, using Chi-square or Mann-Whitney U test. Variables that showed a difference of p-value < 0.25 were analysed with logistic regression.

RESULTS: From the 100 patients included, 38 had RFS. Children who developed RFS had lower weight, lower height, higher weight loss and abnormal baseline electrolytes. Multivariate logistic regression revealed, low height z-score (OR= 0.53, 95% CI: 0.31 – 0.88), higher percentage of weight loss (OR= 1.26, 95% CI: 1.09 – 1.46), and abnormal baseline serum electrolytes (OR= 7.21, 95% CI: 2.04 – 25.50) were associated with RFS. A nomogram was derived from the logistic regression model to predict the probability of RFS (area under curve= 0.88, sensitivity= 0.76, specificity= 0.61).

CONCLUSIONS: Risk factors of RFS in children who received PN were low height, weight loss and abnormal baseline electrolytes. Our study helped to consolidate the knowledge that the risk factors of RFS were similar when comparing children who received PN and EN. In addition, an easy-to-use nomogram was developed to estimate the probability of RFS, which may help in the early identification of children at risk for RFS before starting PN.