Relationship between blood culture collection method and proportion of contaminated cultures in neonates

Abstract

This study aims to document methods of blood culture collection used in our neonatal unit and to determine whether or not the proportion of contaminated cultures varied according to method of collection. Two specific comparisons of interest were the proportion of contaminated cultures in samples collected via (i) preferred versus non-preferred methods, and (ii) peripheral cannulae at the time of insertion versus separate arterial or venous puncture. Data were collected on every blood culture taken in the Grantley Stable Neonatal Unit of the Royal Brisbane and Women's Hospital over a 12-month period. The method used to collect blood for culture was recorded, and the proportion of contaminated cultures taken by each method was calculated. Blood was collected via peripheral cannulae at the time of insertion in 63.1% of cultures, umbilical catheters at the time of insertion in 18.1%, peripheral arterial or venous puncture in 7.7%, and indwelling lines in 2.1%, and the method was unknown in 9.0%. The proportion of contaminated blood cultures was 11.5% with collection via peripheral vascular puncture, 11.1% via indwelling lines, 2.1% via peripheral cannulae at the time of insertion and 1.1% via umbilical catheters at the time of insertion. There was no significant difference in the proportion of contaminated cultures taken via preferred and non-preferred methods. The proportion of contaminated cultures taken via peripheral cannulae at the time of insertion was significantly less than when samples were taken via a separate arterial or venous puncture. In our unit, blood cultures are frequently collected via peripheral cannulae at the time of insertion. Blood cultures taken in this manner are less likely to be contaminated than cultures collected via separate vascular puncture.