Abstract / Summary
As compared with withholding parenteral nutrition (PN) for one week (late-PN), early supplementation of insufficient enteral nutrition with PN (early-PN) increased infection risk and delayed recovery of critically ill children in the PEPaNIC-RCT, with harm attributed to higher amino acid doses. We hypothesized that this harm could be explained by increased ureagenesis, revealing metabolic intolerance to enhanced feeding. This is a secondary analysis of the PEPaNIC-RCT (ClinicalTrials.gov-NCT01536275), in which patients (newborn-17 years) admitted to pediatric intensive care units (PICUs, Leuven-Rotterdam-Edmonton) were randomly allocated to early-PN or late-PN. For patients not receiving renal replacement therapy (RRT), we documented plasma/serum urea and urea/creatinine ratio (UCR) profiles throughout two PICU weeks and investigated whether an increased maximum value in the randomized intervention window (first week) statistically explained harm by early-PN, via multivariable logistic regression and Cox proportional hazards analyses. We studied risk of new infections and duration of PICU dependency as primary outcomes, 90-day mortality as safety outcome, and duration of hospital dependency as secondary outcome. Of 1440 PEPaNIC-patients recruited, 665 early-PN and 663 late-PN non-RRT patients had UCR data. Urea concentrations and UCRs were higher in early-PN than in late-PN patients throughout the first 11 days. Higher first week maximum UCR independently associated with (OR/HR expressed per 30 age-normalized units) an increased risk of new infection [adjusted OR (95%CI) 1.746 (1.353-2.265), p < 0.0001], lower likelihood of earlier live PICU [adjusted HR 0.745 (0.671-0.824), p < 0.0001] and hospital discharge [adjusted-HR 0.778 (0.703-0.860), p < 0.0001], and increased 90-day mortality [adjusted OR 1.432 (1.029-1.993), p = 0.033], and statistically explained (part of) any effect of the randomized intervention. Similar results were obtained for urea concentrations. Increased plasma/serum urea or UCR with early-PN partially explained the harm by early-PN, suggesting these may identify critically ill children metabolically intolerant to enhanced nutrition. This opens perspectives for incorporating urea or UCR into a dynamic monitoring tool to guide the appropriate timing and dosing of nutritional support. ClinicalTrials.gov NCT01536275, registered February 2012.