Our goal was to substantiate the recently hypothesized transmission pathway of potential pathogens (1,6) from a heater–cooler unit to the surgical field by an airborne route in a cardiac surgical operating room equipped with an ultraclean air ventilation system. Smoke dispersal experiments demonstrated that air originating from the heater–cooler unit was propelled by the ventilator of the unit to merge with the ultraclean airflow near the ceiling and ultimately reached the surgical field. This finding was corroborated by particle measurements in the same setup and showed increased particle counts in the surgical field when the heater–cooler unit airflow was oriented toward the operating table than at baseline and when the opposite heater–cooler unit orientation was used. Furthermore, we demonstrated that viable M. chimaera can be recovered on sedimentation plates placed in the air outflow path at a distance of up to 5 m from a contaminated heater–cooler unit. These results corroborate and extend previous observations (6). Previously, bacteria-positive air cultures had been demonstrated only when an active air sampling device was used.

These findings have major and novel implications in 3 areas. First, they substantiate the airborne transmission pathway for outbreaks of M. chimaera cardiac implant–related infections (1,6). Second, they provide prototypical evidence for the failure of ultraclean airflow ventilation systems in preventing surgical site infection. Third, they exemplify the potential infectious risks related to add-on airflow producing devices in the operating room.

A recent outbreak investigation of 6 case-patients with postoperative M. chimaera infections showed that M. chimaera contaminate heater–cooler unit water and are dispersed into the air (6). In that study, when heater–cooler units contaminated with M. chimaera were turned on, air cultures associated with the unit became contaminated with the same mycobacteria, which was proven in 1 instance to display the same random amplified polymorphic DNA PCR pattern. Conversely, air cultures remained bacteria negative when heater–cooler units were switched off or were running but not contaminated. Given the bacteria-positive air cultures and the absence of another common source of infection among the 6 patients, contamination of cardiac implants or the surgical field during surgical intervention had to be assumed. Similar to results of that report, contaminated heater–cooler units and associated infections have been reported in other hospitals in Europe and the United States (1,6,15,16).

Smoke dispersal experiments visualize the airflow pathway by which bacterial aerosols can access the surgical field, and particle count experiments showed a major increase in contamination of air reaching the surgical field. Furthermore, M. chimaera grew not only in samples of a considerable volume of air (6) but also on sedimentation plates up to 5 m from a heater–cooler unit in a nonventilated environment. Thus, the air concentration of mycobacteria was high enough to contaminate implant devices or the surgical field. The sedimentation culture experiments had to be conducted in a test setting because using a contaminated heater–cooler unit in the operating room was no longer advisable. In several other outbreaks of cardiac surgical site infections with nontuberculous mycobacteria, a transmission pathway by air was considered but not proven (3,4,17). Other members of the M. avium complex have been shown to aerosolize from water into air (18–20).

The association of ultraclean air ventilation systems with increased rates of surgical site infections in an epidemiologic evaluation (9) has sparked a debate on the benefit of such systems to prevent surgical site infection in a range of settings for which it has been shown to be effective (8,21–23). Thus, current guidelines offer no recommendation (unresolved issue) for performing orthopedic implant operations in rooms supplied with laminar airflow (24). The operating room smoke and particle count experiments in this study and recently described cases of M. chimaera infections in open-chest heart surgery patients (1,6) demonstrate a possible mechanism of how ultraclean air ventilation systems could fail to prevent, or even contribute to, surgical site infections. The design quality of ultraclean air ventilation systems has been reported as a major factor in their protective function (25). This design quality might also play a pivotal role in the transmission pathway of M. chimaera.

Our evidence also highlights the risks associated with devices that generate airflows in the operating room and that were not part of the original operating room air management design. Wood et al. (10) attempted to establish links between airflow-generating devices in operating rooms, such as forced-air warmers to maintain physiologic patient body temperature, and an increased rate of surgical site infections. However, this attempt was unsuccessful because results of microbial sampling could not be incriminated as a real risk for nosocomial infection (11) or because, although forced-air warmer blowers emitted airborne particles within the size range of free-floating bacteria and fungi, an association with surgical site infections was not investigated (12). A recent review reported that forced-air warmers contaminate ultraclean air ventilation systems (10). However, there appears to be no definite link to an increased risk for surgical site infections on the basis of current research (10). Although our experiments were not designed to answer the question of the overall benefit or harm of ultraclean air ventilation systems, the results exemplify that the combination of a contaminated horizontal airflow and laminar airflow can be responsible for surgical site infections.

Since 2014, heater–cooler units at the University Hospital of Zurich have been uniformly positioned away from the surgical field, and their exhaust air is now captured by a secondary housing and reliably channeled to the operating room exhaust. Although M. chimaera has regrown in some of the monthly water cultures from most factory-new heater–cooler units delivered in January 2014 after 1 year of operation on a daily water-change with filtered tap water scheme, air cultures in the exhaust air have consistently remained bacteria negative.

Our investigations have some limitations. First, the number of experiments was relatively small. However, results were unequivocal and consistent with previous findings (1,6). Second, the specific design of operating rooms, ventilation systems, and the heater–cooler unit model and the interplay of these elements affect the results and limit generalizations. However, specific design characteristics of the operating room ventilation system might potentially explain why infected patients were associated with only some of the many cardiac surgery centers in which heater–cooler units were found to be contaminated by M. chimaera. Third, our experiments were not designed to determine how nontuberculous mycobacteria left the heater–cooler unit water tank and reached the airflow and how the surplus of particles in the heater–cooler unit airflow was generated. These questions need further investigation, including other brands of heater–cooler units. Fourth, our experiments targeted only a nontuberculous mycobacterium that was found to have a major role in the etiology of surgical site infections. Thus, further testing of other pathogens that might be transmitted by the same route should be conducted. A recent study by Knibbs et al. showed that Pseudomonas aeruginosa aerosols emitted by coughing cystic fibrosis patients can become airborne, remain viable for 45 min, and travel up to 4 m (26).

In conclusion, results of our experiments substantiate the recently discovered airborne pathway of a pathogen from a contaminated heater–cooler unit to a surgical field. The specific role of ultraclean air ventilation systems in this pathway needs to be further investigated. Until more detailed evidence is available regarding this issue, all heater–cooler units should be reliably separated from air that can gain access to sterile areas and instruments, and devices that generate drafts should be banned from the operating room.