Introduction

During the spring and summer of 2020, COVID-19 case rates in prison were more than five times the rate of the non-incarcerated U.S. population [1]. In other congregate living facilities, the importance of indoor air quality and ventilation has been widely demonstrated [2, 3]. To our knowledge, no studies have described the means by which ventilation conditions in carceral settings may impact COVID-19 transmission. Addressing this knowledge gap is particularly important given the aging infrastructure and overcrowding in many prisons, as well as the increased vulnerability to complications from infectious diseases or respiratory health issues among incarcerated populations and carceral employees [4, 5].

Respiratory droplets and aerosols containing the SARS-CoV-2 virus exhaled by infected individuals may remain suspended in air for significant periods of time [6].Ventilation works to prevent transmission of the virus indoors by removing air containing viral particles and replacing it with clean air. Poorly ventilated indoor spaces pose a particularly high risk for individuals living or working in high occupancy settings where many individuals may be exposed to air containing infectious viral particles [7].

Carceral facilities in the U.S. tend to be of older construction and there is minimal data on the existence and performance of their ventilation systems. In California, the average age of the 35 California Department of Corrections and Rehabilitation’s (CDCR’s) state correctional facilities exceeds 45 years with approximately 35 percent of the facilities exceeding 50 years of age including two facilities with buildings dating back to the 1800s [8]. Thus, the lack of data on ventilation systems within carceral settings hinders the development of infection control strategies. This knowledge gap also prevents ventilation guidelines from being a part of pandemic preparedness plans and overall ventilation for maintining health in the context of other disasters (e.g. wildfires) in carceral settings.

In this paper we focus on characterizing the ventilation conditions within carceral settings by measuring ventilation rates and direction of airflow in housing units within CDCR. Our findings—particularly when compared to WHO standards for ventilation in healthcare and quarantine facilities—can inform recommendations about pandemic preparedness in the prison context.

Materials and methods

Results

Ventilation rate

The median measured (or estimated) ventilation rates in cells and dormitories ranged from 18.1 CFM/person to 69.9 CFM/person (Table 1). All of the cells measured had ventilation rates less than the WHO recommended minimum of 60 L/S, or 127 CFM/person for healthcare and quarantine facilities [11].

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Ventilation characteristics of six CDCR facilities, 2020–2021.

https://doi.org/10.1371/journal.pone.0293533.t001

Discussion

Our study describes ventilation conditions that were measured at CDCR facilities during the COVID-19 pandemic. Notable findings include low per-person ventilation rates compared to WHO standards, worsened ventilation rates in winter months, and positive static pressure in cells (potentially moving infectious air from cells to common areas), all of which may heighten the risk of COVID-19 transmission among residents and staff. None of the housing units we measured met the minimum 127 CFM/person ventilation rate guideline set by WHO for indoor ventilation in health care and quarantine facilities, and which has been estimated to decrease airborne disease transmission by up to 38% in other studies [11, 12]. The median ventilation rates measured exceeded minimum ventilation rate recommendations for correctional facilities in non-pandemic scenarios. These standards were initially designed to control body odor and we argue that carceral facilities need to increase ventilation rates in the event of a pandemic such that the ventilation function is geared for infection control [13, 14].

The lower air exchange rates observed in the same housing units in the winter versus the summer suggest the role of heating and recirculation systems, as well as the tendency to keep building doors closed in the winter, in reducing overall ventilation rates. During our initial winter visits in December 2020, as much as 90% of the air within housing units was recirculated heated air and the doors were closed to all housing units that we visited. Consequentially, transmission risk within the housing units we sampled could have been significantly higher during the winter months. Following the release of our results, CDCR policy subsequently dramatically reduced recirculation as the risk became apparent. To avoid increased transmission during winter months, facilities should avoid recirculating heated air to the extent possible and instead rely on as much outdoor air as possible. In settings where air conditioning (as opposed to swamp coolers) is used in summer months, the same would apply. Where it is not feasible to use outdoor air, effective filtration is essential [15].

The positive pressure observed in most cells sampled indicates the airflow from potentially infected cells to common areas. Despite isolating individuals from each other in single-person cells within isolation and quarantine units, transmission ensued. In a prison pandemic scenario, isolation and quarantine units in particular should be under negative, rather than positive, pressure. A cell under negative pressure would have a lower air pressure inside the cell compared to the outside common space and ensure that fresh air flows into the room while any harmful virus laden air located inside the room is removed with exhaust systems. Negative pressure rooms are commonly used within healthcare settings to avoid the spread of disease and a core part of the WHO ventilation recommendation in the context of COVID-19. Alternatively, ultraviolet irradiation could be used in common spaces to disinfect air that is on the return path from individual cells [16].

The heterogeneity in the pressure and airflow direction from cells within the same building in facility 1 can be attributed to the poor performance of supply and return vents in individual cells or the practice of residents covering vents with cardboard and cloths to block the entrance of cold air. These factors may also explain why ventilation rates were higher in the solid cells sampled in facility 3, but lower in facility 4. Regular rebalancing exercises can ensure that vents within the system are performing to specification, avoiding inadvertent airflow from areas holding infected individuals to areas with uninfected individuals.

The combination of low ventilation rates and cells under positive pressure suggest the dangers of housing people with unknown infection status in the same building. Despite isolation that is established through individual cells with solid doors, positive pressure from cells to common areas can promote the movement of virus-laden air to areas with uninfected individuals. In cell blocks with open-grate cells, no pressure gradient was observed suggesting a bidirectional flow of air between cells and common areas and the potential for infectious agents generated in any cell to diffuse freely throughout the building.

Our findings on low ventilation conditions within prison housing units have implications for other health outcomes in incarcerated populations and prison staff. Improving ventilation rates and filtration in housing units can reduce the transmission of other airborne diseases such as tuberculosis that are highly prevalent in incarcerated populations [17, 18]. Poor ventilation not only leads to the transmission of infectious diseases, it can also exacerbate the exposure of incarcerated populations to extreme heat [19]. In California, prisons are built in areas that are more prone to extreme heat and wildfire events that are growing in their frequency and intensity. Ventilation improvements are a basic step in mitigating the health risks that these growing environmental threats pose on populations. The methods that we present in this paper can be used to characterize ventilation systems and prioritize improvements.

It its important to note that some practices aimed at controlling COVID-19 may work to exacerbate exposure to wildfire smoke. For example, minimizing air recirculation and introducing as much fresh outdoor air as possible is a strategy for decreasing COVID-19 transmission risk. In contrast, to minimize smoke exposure during a fire event, the use of recirculation mode on the air conditioning unit should be maximized to prevent outdoor smoke from entering the indoor space [20]. Further work is needed to develop specific guidance for facilities that may be balancing the need to simultaneously control for wildfire smoke exposure and COVID-19 transmission.

Conclusions

Despite the toll that COVID-19 has had on people who are incarcerated, there has been little understanding of successful mechanisms for reducing transmission in these facilities. We conclude that ventilation improvements should be a central component of emergency respiratory pandemic preparedness and response plans. Given the location of prisons in wildfire prone areas, properly functioning ventilation and air filtration systems are a minimum requirement in reducing disproportionate exposures to smoke for incarcerated populations. Ventilation assessments such as the ones that were conducted here can be used on an ongoing basis to identify high-risk areas and monitor the impact of mitigation measures designed to improve ventilation. The methods that we use here are relevant for incarcerated populations in California as well as other states where overcrowding, poorly maintained ventilation systems, and susceptible residents are subjected to disproportionate exposure to health risks and disease.

Supporting information

Acknowledgments

We offer our sincere gratitude to those who helped us conduct this research, particularly the residents and staff of California prisons, as well as those who have been working to respond to COVID-19 within prisons and jails across the country.