Our paper on Implementing a negative-pressure isolation ward for a surge in airborne infectious patients was a fruitful collaboration between hospital engineers, Veterans Health Infection Control professionals, and University of Colorado Boulder air quality researchers.  This work has recently garnered attention because of a critical need for isolation capacity during the highly transmissible COVID19 pandemic.

In this demonstration study we worked with a modern hospital in California to figure out and test how to turn a 30-bed hospital ward into a negative pressure isolation ward in a few hours and maintain the capacity for 24 hours.  The demonstration was very successful, especially due to the creativity of the lead scientist Dr. Nick Clements and the expertise of the lead hospital engineer Steven Elliott.

One important implication of this work is that while the negative pressure capacity can be held between the ward and the rest of the hospital, on the ward, the rooms have varying pressure differentials and therefore the hospital staff need to wear personal protective equipment while on the ward.

Additional excellent questions and points about the work were recently brought up in email exchanges between hospital engineers at other facilities – and I am sharing these ideas here FYI.

Point 1: It is absolutely critical that any exhaust air out of the ward be dedicated with its own fan.  The standard and most common HVAC system comprises of Supply and Return Fans with a mixing air damper.  Under the scenario provided in the article, the mixing damper would be 100% closed with outside air damper and exhaust air damper open.

Response: We operated in the study with the mixing air damper closed and with one supply (60% capacity) and one return fan (100% capacity). Only the isolation rooms had their own dedicated exhaust fans. Requiring a dedicated exhaust fan for the whole ward in case of surge would be great, but that’s not the reality in most hospitals, and during a surge these facilities would just have to do the best they can. There will be potential for recirculation, though without testing it is impossible to know how much of a risk this poses. Our guess would be small, but it ultimately depends on the craftsmanship of the sheet metal contractor who did the duct install.

Point 2: Negative air condition will occur if the fans are modulated appropriately.  What this scenario does not address is leakage between the return ducting (downstream of return fan) and the supply duct (upstream of the supply fan).   Most if not all mixing dampers can not achieve 100% shut-off with 0% chance of leak-by.  Furthermore, electrical penetrations, sheet metal joints and other pathways exist between supply/return ductwork which allow additional cross contamination.  As such, a percentage of the contaminated air from the isolation ward will be recirculated back into the space.  This places the hospital staff at great risk because they are working in a space with contaminated air.

Response: This worked for the California hospital where we did the study due to their AHU servicing only the containment ward floor, and any recirculated material will just end up back in the isolation ward.  But it would pose a significant risk if the AHU serviced non-isolation ward areas. There may be some isolation ward duct leakage that causes virus particles to end up in wall interstitial spaces, that then could get picked up by other ducts and sent to other parts of the hospital, which could happen. One would need tracer testing to verify for sure.

Point 3: It cannot be guaranteed that all patients in the ward share only one infectious disease.  Example:  20 out of 20 patients in an isolation ward have COVID-19.  1 out of 20 patients in the wards also has TB.  Using the existing ventilation fans in the manner described in the article will now expose the 19 other patients to TB due to leakage of the exhausted return air into the supply ducts.

Response: We also want to stress that the point of the ward is to house only patients with the surge disease, and that hospital staff should wear full PPE on the ward. Patients with other diseases (including airborne) should not be on the ward if possible, since there is no control over pressures between rooms. This is a not a design where the ward hallways are safe and rooms are converted to AIIR’s, everywhere on this ward is potentially contaminated. This is one reason we installed UV lights in the stairwells for example.

Lastly, our research suggests this is type of surge response is possible, but without testing and planning BEFORE a pandemic, we don’t necessarily recommending adopting the negative isolation ward as a surge response, as there may be too much risk involved in not thoroughly understanding all these uncertainties. (Although in a crisis such as where we are now it may be an important option.) Our recommendation is to have this as a preparedness measure for future outbreaks.  There are some hospitals that do have this in their flu pandemic plans, but we imagine many more who do not.

Shelly Miller, PhD
Nick Clements, PhD