Influence of Prolonged Use of Intravenous Administration Sets in Paediatric Cancer Patients on CVAD-related Bloodstream Infection Rates and Hospital Resources

Abstract

Background:

To assess the effects of extending the routine intravenous administration set (IVAS) change-interval from 72 h (group 1) to 7 days (group 2) on the incidence density for central venous access device (CVAD)-related bloodstream infections (BSIs) and on resource expenditures in a singlecentre pilot study.

Procedure:

Prospective pre-/post-intervention comparison of two consecutive 12-month surveillance periods (2001–2003) in a 17-bed paediatric oncology tertiary care unit. IVAS changes and nosocomial infections (NIs) were prospectively analysed using a standardized unit-based surveillance system (Oncopaed NI).

Results:

All 175 eligible patients were enrolled, 96 in group 1 and 79 in group 2. Both groups had similar distributions of primary diagnoses and risk factors. The proportion of IVAS changes performed after 3 days increased from 5.6% to 22.5%, but only 8% of IVASs in group 2 were changed after 7 days. Most IVAS changes (64.8% in group 1 and 92.9% in group 2) were done because of therapeutic interventions (blood products, parenteral nutrition [TNP]) before the scheduled endpoint. Overall, the rates and incidence densities of NIs were significantly lower during the second period. The corresponding results for CVAD-related BSIs did not show significant differences. No death attributable to a NI occurred. The ‘7-day’ strategy resulted in cost savings for devices (3,300$/year) and of nursing time (23 working days/year).

Conclusions:

Extending the routine IVAS change-interval from 3 days to 7 days appears to be safe and cost-effective in a paediatric oncology unit with high infection control standards and continuous surveillance for NIs. These results do not prove that 7-day intervals prevent infections, but they do suggest that this policy probably is not harmful and that a prospectively randomized study with sufficient power is needed.