Lack of effect of changing needles on contamination of blood cultures

To determine whether changing needles during the collection of blood cultures reduces contamination, we randomly assigned 303 children undergoing blood cultures to 1 of 3 groups: no needle changes (blood instilled directly into culture media through the needle used for venipuncture); 1 needle change (before inoculation into the first of 2 culture bottles); and 2 needle changes (before inoculation into each of 2 culture bottles). Each patient's skin was cleansed with povidone-iodine for 60 seconds before venipuncture. We found similar rates of contamination among the 3 groups: no change, 2 of 92 (2.2%); 1 change, 0 of 106 (0.0%); 2 changes, 2 of 105 (1.9%). The combined contamination rate of all 3 groups (1.3%) was significantly lower than the prestudy rate of contamination (4.4%), based on 315 blood cultures (P = 0.04). These data suggest that careful skin preparation is a more important factor than changing needles in reducing contamination during blood culture collection.