Using adenosine triphosphate bioluminescence level monitoring to identify bacterial reservoirs during two consecutive Enterococcus faecium and Staphylococcus capitis nosocomial infection outbreaks at a neonatal intensive care unit

Introduction: This study aimed to assess the role of adenosine triphosphate (ATP) bioluminescence level monitoring for identifying reservoirs of the outbreak pathogen during two consecutive outbreaks caused by Enterococcus faecium and Staphylococcus capitis at a neonatal intensive care unit (NICU). The secondary aim was to evaluate the long-term sustainability of the infection control measures employed one year after the final intervention measures. Methods: Two outbreaks occurred during a 53-day period in two disconnected subunits, A and B, that share the same attending physicians. ATP bioluminescence level monitoring, environmental cultures, and hand cultures from healthcare workers (HCW) in the NICU were performed. Pulsed-field gel electrophoresis (PFGE) typing was carried out to investigate the phylogenetic relatedness of the isolated strains. Results: Four cases of E. faecium sepsis (patients A-8, A-7, A-9, B-8) and three cases of S. capitis sepsis (patients A-16, A-2, B-8) were diagnosed in six preterm infants over a span of 53 days. ATP levels remained high on keyboard 1 of the main station (2076 relative light unit [RLU]/100 cm²) and the keyboard of bed A-9 (4886 RLU/100 cm²). By guidance with the ATP results, environmental cultures showed that E. faecium isolated from the patients and from the main station’s keyboard 1 were genotypically indistinguishable. Two different S. capitis strains caused sepsis in three patients. A total 77.8% (n = 7/9) of S. capitis cultured from HCW's hands were genotypically indistinguishable to the strains isolated from A-2 and A-16. The remaining 22.2% (n = 2/9) were genotypically indistinguishable to patient B-8. Three interventions to decrease the risk of bacterial transmission were applied, with the final intervention including a switch of all keyboards and mice in NICU-A and B to disinfectable ones. Post-intervention prospective monitoring up to one year showed a decrease in blood culture positivity (P = 0.0019) and catheter-related blood stream infection rate (P = 0.016) before and after intervention. Conclusion: ATP monitoring is an effective tool in identifying difficult to disinfect areas in NICUs. Non-medical devices may serve as reservoirs of pathogens causing nosocomial outbreaks, and HCWs' hands contribute to bacterial transmission in NICUs.