Ae. Macias et al., Parenteral infusions bacterial contamination in a multi-institutional survey in Mexico: Considerations for nosocomial mortality, AM J INFECT, 27(3), 1999, pp. 285-290
Background: Parenteral infusions can be contaminated during administration
(extrinsic contamination). A previous survey found that extrinsic contamina
tion was not uncommon in a hospital in Mexico with lapses in aseptic techni
ques. To determine whether this problem exists in other similar institution
s, we undertook a multi-institutional study.
Methods: We surveyed 6 hospitals (A to F) lacking an infection control comm
ittee to determine the level of extrinsic contamination. We visited each ho
spital and obtained samples of all the parenteral infusions in use, drawing
0.5-1 mL from the tubing injection port. Quantitative and qualitative bact
erial cultures were performed. Chlorine levels of the tap water were measur
ed. Visits were repeated until the survey was completed.
Results: A total of 751 infusions were cultured, of which 16 (2.13%) were c
ontaminated. Hospital contamination rates varied from zero to 5.56%. Klebsi
ella pneumoniae was the most common isolate (10 cases). During the first sa
mpling day in hospital C, the 7 infusions from the pediatric ward were foun
d to be contaminated with a similar K pneumoniae strain. In-service educati
on was started in this hospital. Infusion contamination was eliminated foll
owed by a reduction in mortality rate. Overall, a higher risk for infusion
contamination was noted for pediatric patients (P =.01, odds ratio = 3.28,
95% CI, 1.10-9.91) and in wards with inadequate water chlorine levels (P =.
02, odds ratio = 3.64, 95% CI, 1.08-13.51).
Conclusions: If the hospitals surveyed are representative of others in deve
loping countries, an endemic level of parenteral infusion contamination cou
ld exist in many hospitals throughout the world.