An outbreak of Enterobacter cloacae associated with contamination of a blood gas machine.

Over a 3-month period, five cases of Enterobacter cloacae bacteraemia occurred on our neonatal unit. In at least three of these, isolation of the organism coincided with clinical deterioration and evidence of sepsis. In one case, the same strain was isolated from an abscess on the neonate's forearm. The isolates had identical sensitivity patterns being resistant to all beta-lactams tested except imipenem. The extended time course of the infections made cross-infection an unlikely explanation. Moreover, close questioning of the staff and observation of their practices with regard to blood culture collection, failed to reveal any likely mechanism for pseudobacteraemia. On extensive investigation of the environment to try to identify a potential source of the organism, a strain of Enterobacter cloacae, was isolated from the probe of the blood gas machine and the probe cover. No other environmental samples were found to harbour the organism. Subsequent typing procedures showed the blood gas isolate to be indistinguishable from the clinical isolates. Five neonates were successfully treated with imipenem and gentamicin. The exact mechanism whereby these bacteraemias occurred remains obscure. In one case, the baby had positive blood cultures within 2 h of being on the unit and contamination of the blood culture bottle by the doctor taking the culture was suspected. Most of the episodes, however, appeared clinically to be genuine septicaemias. When vigorous infection control procedures were instituted to prevent staff acquisition of the organism from the machine, cases on the unit ceased.