Objective: To investigate an outbreak of Burkholderia (formerly (Pseudomonas) cepacia respiratory tract colonization and infection in mechanically ventilated patients.
Design: A retrospective case–control and bacteriologic study.
Setting: Veterans Affairs medical center.
Patients: 42 mechanically ventilated patients who developed respiratory tract colonization or infection with B. cepacia and 135 ventilator-dependent controls who were not colonized and did not develop infections.
Measurements: Clinical and demographic data; benzalkonium chloride concentrations and pH levels in albuterol sulfate solutions; repetitive-element polymerase chain reaction (PCR)-mediated molecular fingerprinting on eight patient isolates and three environmental B. cepacia isolates that were available for study.
Results: 42 patients had B. cepacia respiratory tract colonization or infection. Observation of intensive care unit and respiratory care personnel showed faulty infection control procedures (for example, the same multiple-dose bottle of albuterol was used for many mechanically ventilated patients). More case patients (39 [92.9%]) than controls (95 [70.4%]; P = 0.006) received nebulized albuterol, and case patients (67.5 treatments) received more treatments than controls (18 treatments; P < 0.001). In-use albuterol solutions had pH values that were unstable, and benzalkonium chloride concentrations declined over time to levels capable of supporting bacterial growth. Medication nebulizers and in-use bottles of albuterol harbored B. cepacia. Molecular fingerprints of patient isolates and environmental B. cepacia isolates were identical using repetitive-element PCR. No further isolates of B. cepacia were identified after institution of appropriate infection control procedures.
Conclusions: Multiple-dose medications and reliance on benzalkonium chloride as a medication preservative provide a mechanism for nosocomial spread of microorganisms, particularly if infection control procedures are not carefully followed. Repetitive-element PCR is a useful fingerprinting technique for molecular epidemiologic studies of B. cepacia.