Neonatal Nosocomial Sepsis from Multiple Resistant Klebsiella pneumoniae Transmitted through Multi-Dose Bottles of an Intravenous Nutritional Supplement.

Introduction

Nosocomial neonatal sepsis from gram-negative bacteria is an important problem in Saudi Arabia, and is compounded by antibiotic resistance. We investigated a problem of sepsis in a neonatal special care unit (NSCU), where a strain of Klebsiella pneumoniae resistant to ampicillin, gentamicin, chloramphenacol, cephalothin and tobramycin (resistant Klebsiella) had been causing sepsis in over 3% of neonates for more than one year.

Methodology

We identified 47 neonates who over a three-month period developed clinical sepsis with the resistant Klebsiella isolated from blood (case-neonates). We chose 102 infants admitted during the same time to the NSCU who did not develop sepsis (control-neonates). We reviewed medical files for birth circumstances, underlying illness and exposures to intravenous therapy, medical procedures and equipment. After stratification by birthweight, exposures during the 48 hours before onset or blood culture for case-neonates were compared with exposures of control-neonates during the same time period.

Results

Case-neonates were more likely to die than control-neonates (risk ratio [RR] = 1.9: 95% confidence interval [CI] 1.4-2.6) and the risk ratio for death was highest in neonates with birthweights over 2000 gms (RR=4.3; 95% CI 1.8-10). All case-neonates had an intravenous line, compared with 92% (94) control-neonates (p=0.056, two-tailed Fisher's exact test). One intravenous additive, Aminosteril, was strongly associated with cases (odds ratio=281; 95% CI 23-3324). One-liter multi-use bottles of Aminosteril were kept in the NSCU refrigerator. Resistant Klebsiella was isolated from two of three partially used Aminosteril bottles and from the stool of three of 32 neonates in the NSCU.

Conclusion

Resistant Klebsiella sepsis was caused by contamination of multi-dose Aminosteril bottles. Since the Aminosteril bottles were consumed every two to three days, the stool of the neonates probably provided the reservoir for the organism. We suspect that poor handwashing before handling the Aminosteril and the intravenous system allowed contamination. This hospital and others with neonatal Klebsiella sepsis will need strict control of multi-use bottles and enforced handwashing before handling intravenous systems. Only one new case has occurred in this hospital over two months since beginning these controls.