Percutaneously inserted central venous catheters (CVCs) are frequently used for parenteral nutrition (PN) in neonates. Catheter-related sepsis (CRS) is the most commonly reported complication. The aim of this study was to compare sepsis rates between neonates receiving PN by CVC and by peripheral cannula. Neonates were randomized to receive PN either by CVC or by peripheral cannula. Primary outcomes were sepsis rates and efficacy of PN delivery. Interim analysis (49 neonates) revealed no difference in sepsis rates (CVC group 46%, cannula group 40%, p = 0.57). There was a significant discrepancy in PN delivered (median deficit 3.2% in the CVC group, 10.3% in the cannula group, p=0.0014). After consideration of these findings the study was terminated.
Conclusion: Because the trial was stopped early, small but clinically important differences in the rates of sepsis may have been missed. Neonates in the cannula group accrued significant deficits of PN owing to a lack of venous access. This may contribute to undernutrition in neonates. CVC can be advocated for PN administration, in that sepsis appears to be no higher than when using cannulae and delivery of nutrition is significantly better.