Contaminated Ventilator Air Flow Sensor Linked to Bacillus cereus Colonization of Newborns

The Missouri Department of Health and Senior Services conducted this investigation in response to the hospital's identification of an increased number of tracheal aspirates that were positive for B. cereus collected from newborns who were on ventilators during March–May, 2011. All tracheal aspirate culture results obtained in the Neonatal Intensive Care Unit (NICU) during January 2010–June 2011 were reviewed. NICU data was also searched for positive B. cereus culture from other specimens, such as blood, body fluids, or tissues. Investigators thoroughly evaluated respiratory management practices in the unit by direct observation, respiratory records review, and an interview with the respiratory therapist.
Several environmental cultures were obtained from the flow sensors of the unit's ventilators over the 1-month period. B. cereus isolates were forwarded to the Centers for Disease Control and Prevention to be molecularly characterized by using multilocus sequence typing (MLST). DNA was prepared from bacterial cultures as described. The DNA was used as a template in PCRs with the primers described on the Bacillus cereus MLST Web site for the 7 loci which define the MLST scheme. The sequences for the loci glpF, gmk, ilvD, pta, pur, pycA, and tpi were then assigned allele designations. The combination of the 7 alleles determines a given sequence type. A greater number of alleles that match between strains indicates a higher level of relatedness. Prevalence of B. cereus–positive specimens was compared by using the Mann-Whitney U test.
Retrospective analysis of tracheal aspirate culture results showed significant increase (p = 0.039) in B. cereus isolation between March and May, 2011. No Bacillus spp. were isolated from blood, other body fluids, or tissues during the study period. The chart review of the case-patients comprising the cluster of B. cereus colonization revealed that none received a diagnosis of clinical B. cereus infection. All patients were treated with vancomycin or tobramycin, or both, for indications not related to B. cereus in tracheal aspirate. One case-patient died 108 days later without evidence that B. cereus contributed to the outcome. All other case-patients recovered and were discharged.

Investigation of the ventilation procedures in the NICU revealed that most equipment used for respiratory care was disposable, designated for single-patient use. The Draeger Evita v500 ventilator was used for mechanical ventilation of infants who were intubated to treat severe respiratory compromise. The Draeger Evita V500 is a microprocessor controlled ventilator offering both mandatory and spontaneous ventilation modes for adult, pediatric, and neonatal patients. Heated and humidified gas flows from the ventilator unit, through the inspiratory circuit and NeoFlow air flow sensor to the patient through an endotracheal tube. Upon exhalation, gas flows back through the air flow sensor into the expiratory circuit and returns to the ventilator through the expiratory flow sensor and exhalation valve. In addition to the ventilator, reusable respiratory equipment comprised a proximal air flow sensor, expiratory flow sensor, exhalation valve, and circuit temperature probe. The sensor closest to the newborn's mouth was an air flow sensor located inside the disposable ventilation circuit. From 9 environmental cultures obtained from 9 air flow sensors, 1 was positive for Bacillus spp., and was later confirmed as B. cereus by the State Public Health Laboratory.

MLST was performed for 8 B. cereus isolates from case-patients and for 1 environmental isolate from the air flow sensor. We were able to fully characterize 4 of the 9 isolates. One locus for the remaining 5 strains did not yield an amplicon for sequencing after repeated attempts and, thus, could not be assigned a sequence type. The isolates that included sequence type (ST) 73 and ST94 were closely related to each other because they differed by merely 1 locus, gmk. The strains that were not fully typed because of the inability to obtain sequences for locus pta were also closely related to ST73 or ST94 because the other loci matched. There was 1 match between strains isolated from 1 case-patient and the air flow sensor, which was ST73. The contaminated air flow sensor was then sterilized by using a steam autoclave. A repeat culture of this sensor after sterilization was negative.
We found that air flow sensors were routinely disinfected by placing them in a container with 70% alcohol solution for 60 minutes. After discovery of the air flow sensor contaminated with B. cereus, the disinfection policy was changed. All air flow sensors were first soaked in Enzol enzymatic detergent solution and then sent for steam autoclave sterilization at 134°C (273.2°F). After implementation of new disinfection and sterilization procedures, no new cases of B. cereus tracheal colonization were identified in the nursery. In this cluster, contaminated proximal air flow sensors were the likely source of tracheal colonization with B. cereus in newborn infants, supported by a genetic match by MLST between a strain isolated from 1 case-patient and the contaminated air flow sensor.

ISweek（http://www.isweek.com/）- Industry sourcing & Wholesale industrialproducts