The effect of low-dose heparin on maintaining peripherally inserted percutaneous central venous catheters in neonates

Abstract

Objective: To evaluate the effect of low-dose heparin on duration of peripherally inserted percutaneous central venous catheter (PCVC) patency and catheter occlusion.

Study design: Our research was a prospective, randomized, controlled, double-blind clinical study including 246 eligible neonates with PCVCs. It was conducted at the neonatal intensive care unit (NICU) of Diyarbakir Children's Hospital between 1 February 2007 and 31 October 2008. In the heparin group (n=118), the total parenteral nutrition (TPN) solutions contained heparin, which was infused at a rate of 0.5 IU kg⁻¹ h⁻¹ . The no-heparin group (n=121) received TPN fluids without heparin.

Result: In the heparin group, the duration of catheter patency was longer than that of the no-heparin group reaching a statistically significant difference (12.4 ± 4.5 vs 9.7 ± 4.0, P<0.0001). In the heparin group, the proportion of patients completing the TPN therapy successfully was higher than in the no-heparin group (P=0.0001; relative risk (RR): 3.32 (95% confidence interval (CI): 1.92 to 5.73)) also demonstrating a statistically significant difference. The rate of catheter occlusion among infants in the heparin group was smaller (P=0.0001; RR: 3.44 (95% CI: 1.92 to 6.44)). There were no statistically significant differences between the two groups in the incidence of other catheter-related problems. The length of the occluded catheter outside the skin, TPN infusion rate and catheter duration time were found to be the factors associated with catheter occlusion. Heparin prevented catheter occlusion with a mechanism unrelated to these factors.

Conclusion: In neonates with PCVCs, low-dose continuous infusion of heparin (0.5 IU kg ⁻¹h⁻¹ within TPN fluids is an effective measure in terms of reducing catheter occlusion, allowing successful completion of the therapy, without increasing adverse effects. Furthermore, this effect of heparin may occur independently via occlusion-related factors that we evaluated in the study.