Effect of early supplemental parenteral nutrition in the paediatric ICU: a preplanned observational study of post-randomisation treatments in the PEPaNIC trial

Background Large randomised controlled trials have shown that early supplemental parenteral nutrition in patients admitted to adult and paediatric intensive care units (PICUs) is harmful. Overdosing of energy with too little protein was suggested as a potential reason for this. This study analysed which macronutrient was associated with harm caused by early supplemental parenteral nutrition in the Paediatric Early versus Late Parenteral Nutrition In Critical Illness (PEPaNIC) randomised trial. Methods Patients in the initial randomised controlled trial were randomly assigned to receive suppplemental parenteral nutrition (PN) within 24 h of PICU admission (early PN) or to receive such PN after 1 week (late PN) when enteral nutrition was insufficient. In this post-randomisation, observational study, doses of glucose, lipids, and aminoacids administered during the first 7 days of PICU stay were expressed as % of reference doses from published clinical guidelines for age and weight. Independent associations between average macronutrient doses up to each of the first 7 days and likelihood of acquiring an infection in the PICU, of earlier live weaning from mechanical ventilation, and of earlier live PICU discharge were investigated using multivariable Cox proportional hazard analyses. The three macronutrients were included in the analysis simultaneously and baseline risk factors were adjusted for. Findings From June 18, 2012, to July 27, 2015, 7519 children aged between newborn and 17 years were assessed for eligibility. 6079 patients were excluded, and 1440 children were randomly assigned to receive either early PN (n=723) or late PN (n=717). With increasing doses of aminoacids, the likelihood of acquiring a new infection was higher (adjusted hazard ratios [HRs] per 10% increase between 1.043-1.134 for days 1-5, p=0.029), while the likelihood of earlier live weaning from mechanical ventilation was lower (HRs 0.950-0.975 days 3-7, p=0.045), and the likelihood of earlier live PICU discharge was lower (HRs 0.943-0.972 days 1-7, p=0.030). By contrast, more glucose during the first 3 days of PICU stay was independently associated with fewer infections (HRs 0.870-0.913, p=0.036), whereas more lipids was independently associated with earlier PICU discharge (HRs 1.027-1.050, p=0.043 days 4-7). Risk of harm with aminoacids was also shown for low doses. Interpretation These associations suggest that early administration of aminoacids, but not glucose or lipids, could explain harm caused by early supplemental parenteral nutrition in critically ill children.