Indoor Environment Control Strategies: Building Ventilation

Overview

In the mid-19^th century Florence Nightingale advocated for outdoor air flow into hospitals, schools, and other indoor environments as a means of infection control. She even suggested air flow rates that are similar to those recommended by contemporary experts [1]. Ventilation in buildings is defined as the rate at which outdoor air flows into an indoor environment and replaces the existing air. Mechanical ventilation uses fans to draw air into an indoor environment, while natural ventilation includes air flow from infiltration through the building envelope and from the opening of windows or doors. Because it is understood that infected humans can generate infectious aerosols that can float through indoor environments leading to inhalation exposure by others sharing the indoor air, ventilation offers to reduce the concentration of infectious aerosols — through dilution and removal — that might otherwise accumulate.

The importance of ventilation for controlling aerosols assumed to mix evenly throughout an indoor environment was described in the foundational work of Wells, leading to the Wells-Riley equation [2] given by,

with P probability of infection, I number of infectors shedding pathogens into aerosols, q quantum generation rate, p pulmonary ventilation rate, t time sharing the air of an indoor space, and Q ventilation rate. The quantum generation rate is the generation rate of inhalation doses leading to 63% rate of infection, infectious dose 63 (ID₆₃). Given human exhaled breath as the predominant source of CO₂ in an indoor environment, Rudnick and Milton built on the Wells-Riley equation proposing the rebreathed-air equation using CO₂ as a marker of how much air being inhaled by an individual is made of the exhaled breath of people in a room [3]. The rebreathed-air equation is,

with n people sharing the air, and rebreathed-fraction equal to the CO₂ concentration in the room minus the CO₂ concentration outdoors divided by the CO₂ concentration of exhaled breath, which is a physiologic constant of 38,000 ppm integrated over the exposure time. Aside from providing direct estimates of exposure to exhaled breath in a well-mixed room, the rebreathed-air equation could be used with CO₂ sensors and in the absence of the challenges involved in assessing and estimating room ventilation, required in the Wells-Riley formulation. Given a certain number of people occupying an indoor environment of a specified volume, CO₂ is a function of the rate of ventilation as outdoor air with lower CO₂ (generally 400-450 ppm) replaces indoor air containing built up exhaled breath.

Elevated CO₂ levels have become synonymous with increasing risk transmission for SARS-CoV-2 and other respiratory infections in indoor environments. Epidemiologic evidence from observational, experimental, and modelling studies supports the relationship between infection risk and ventilation or CO₂, occupancy, exposure time, and airborne infection generation rate [2-8].

Epidemiologic studies and models of infection control by building ventilation

There are numerous studies linking measured or estimated levels of ventilation with detected respiratory infections in the context of outbreaks or community incidence. Systematic reviews of this literature have concluded that indoor ventilation levels are an important factor in airborne infection control [9-11]. However, the current body of existing literature does not contain sufficient information to adequately support specific, minimum levels of ventilation for what might be considered acceptable infection control in various settings. Studies to generate such knowledge would have to adequately characterize the magnitude, and duration of the exposure between infected and susceptible individuals, the ventilation, and chains of infection transmission. Such studies must be well designed and executed and require substantial investment of time and resources.

Human challenge-transmission trials with controlled exposure aimed at investigating transmission modes have been conducted [12]. Yet these experimental studies face implementation obstacles and generally rely on artificial inoculation which may not recapitulate the most infectious cases required for small-scale studies to achieve enough transmission and reach sufficient statistical power [4, 13]. Improvements to this experimental design – such as the use of naturally infected cases, and controlling exposure through changes in ventilation levels – may yield findings toward updated standards. Alternatively, modulating ventilation in high exposure settings and prospectively monitoring infection status may provide direct evidence of ventilation control effectiveness.

The table below gives an overview of the main findings from a selection of studies that linked ventilation with respiratory infection risk. The studies included were those where there was available measurement or estimation of ventilation during exposure between primary and secondary cases, and some assessment of infection risk given the total number of people exposed. These studies consistently show a protective effect of ventilation on infection risk of approximately 25 – 300% or more, often reaching statistical significance.

Table. Epidemiologic studies that evaluate the effect of ventilation on infection risk

Study	Exposure scenario	Infection assessment	Ventilation assessment	Ventilation comparison	Risk of infection^a
[14] (Brundage et al., 1988)	Army trainees living in close quarters in barracks over 47 months.	Retrospective records assessment for febrile (≥38^oC) respiratory illness.	Estimated based on barrack design and occupancy.	0.9 vs 6.8 L/s/p; ‘Modern’, tighter sealed versus older, ‘leaky’ barracks.	Adjusted risk ratio 1.51 (95% CI 1.46-1.56).
[15] (Hoge et al., 1994)	Inmates living in different areas of a jail over 4 weeks during an S. pneumoniae outbreak.	Prospective culture of respiratory swabs & serology for S. pneumoniae.	Cross-sectional measurement of CO₂ and evaluation of air flow from the ventilation system.	>3.4 vs 2.0 L/s/p; Cell block style with more vs less outside air.	Adjusted odds ratio 2.02 (95% CI 1.07-3.82).
[5] (Menzies et al., 2000)	Health care workers exposed to TB cases over 3 years.	Cross-sectional analysis of tuberculin skin test (TST) conversion.	Cross-sectional CO₂ decay and smoke release experiments.	<2 vs ≥2 ACH; workers in lower versus higher ventilated spaces.	Adjusted hazard ratio 3.4 (95% CI 2.1-5.8) for TST conversion negative to positive.
[16] (Menzies, et al., 2003)	Laboratorians working in hospitals with TB cases over 3 years.	Cross-sectional analysis of tuberculin skin test conversion.	Cross-sectional CO₂ decay and smoke release experiments.	16.7 vs 32.5 ACH (averages); lower versus higher ventilated spaces.	Unadjusted infection risk greater (p<0.001; unpaired t test).
[17] (Sun et al., 2011)	University dormitory population in 17 buildings.	Questionnaire of common cold incidence over the previous year.	Cross-sectional measurement of CO₂ decay experiments (peak vs outdoor).	5 vs 1 L/s/p; Average dorm room ventilation rate of (CO₂ decay calculation).	Incident risk ratio 7 (35% versus 5% study population) for ≥6 common colds.
115] (Du et al., 2020)	Retrospective cohort of household and university campus contacts (at least 30 hours of shared air) of infectious TB cases from an outbreak. Ventilation interventions applied.	Sputum tests and chest x-ray to detect active TB cases, with sequencing to confirm probable transmission clusters.	Measured CO₂ concentrations before and after intervention.	>1,000 vs <1,000 ppm CO₂ (pre- vs post-intervention classroom) with exposure to cases; CO₂ ppm (estimated clean air flow L/s/p) were 3,200 (1.7) & 600 (23.6-25.1).	Adjusted hazard ratio 32.8 (95% CI 2.0-540.3) for acquiring active TB infection.
[18] (Zhu et al., 2020)	University dormitory population in ‘low’ (LVB) and ‘high ventilated (HVB) dormitory buildings.	Prospective symptom monitoring with qRT-PCR for acute respiratory infections.	Continuous measurement of CO₂ >5 mo. Measurement of building envelope pressure and local weather data.	2.0 vs 5.9 L/s/p; LVB vs HVB participant room means. 1.9 vs 2.1L/s/p; leeward vs windward room means in LVB.	Unadjusted incident rate ratio 4.04 (95% CI 0.69-163.02) LVB vs HVB; 1.3 (0.7-2.61) leeward vs windward.
[12] (proof-of-concept [POC], Killingley et al., 2012); [13] (main study, Nguyen-Van-Tam et al., 2020)	Compared infection rate between 2 human challenge trials with different ventilation levels.	Daily, upper respiratory swab sample following inoculation, and pre- vs post-infection serology.	POC estimated given sealed suite with bathroom exhaust [4]. Main estimated by CO₂ decay & tracer gas experiments [13].	0.8 L/s/p vs 4 L/s/p; POC vs main study.	Unadjusted risk ratio of transmission 6.4 (8.3 vs 1.3%).
^aExposure is compared at lower vs higher ventilation level.

Studies are limited that offer direct comparisons between infection outcomes under varying levels of well-characterized ventilation conditions. Reports of outbreaks and case reports of transmission events in poorly ventilated environments are more abundant. A recent report of 9 secondary SARS-2 infections from a single primary case in a restaurant with recirculating air-conditioning units and an average ventilation rate of 0.9 L/s/p suggested a ventilation rate of 38.6 L/s/p to inhibit transmission [19]. More comprehensive, albeit, older systematic reviews of ventilation on infection control can be found elsewhere [9, 10]. An interdisciplinary panel from the earlier of the two reviews concluded that “there is strong and sufficient evidence” to demonstrate that lower ventilation rates and indoor airflow from infected to uninfected people are associated with increased transmission of infectious diseases “such as measles, tuberculosis, chickenpox, influenza, smallpox and SARS [9]. Based on the literature reviewed, the panel stated “This evidence supports the use of negatively pressurized isolation rooms for patients with these diseases in hospitals….”. However, the panel also concluded that the available data were insufficient to form a basis for specifying the minimum ventilation rates needed to limit infectious disease transmission in various types of buildings. Similar conclusions were drawn by the later literature review that considered the role of mechanical ventilation, in particular, on aerosol transmission [10]. Although prospective studies of large populations over time with precisely characterized exposure and infection outcome data would carry stronger evidence, existing studies and prioritization of aerosol transmission control (e.g., applying the precautionary principle) provide sufficient scientific evidence to enhance ventilation [20].

Numerous studies have documented transmission of SARS-CoV-2 in low ventilated indoor environments and this is contrasted with comprehensive contact tracing that has detected low levels of transmission during outdoor exposures where dilution ventilation is abundant [21, 22]. Outdoor exposure may lead to transmission when there is crowding, close physical contact, prolonged exposure, and poor face mask adherence [21]. It is well regarded that ventilation can physically remove contaminated air and reduce potential airborne exposure and transmission. The critical question centers on how to implement ventilation for infection control in indoor settings where thermal comfort, energy use, and financial cost are also driving considerations. Reported superspreading events appear to contain a common thread of occurring in crowded settings, low ventilated indoor environments, environments where air is recirculated without filtration, places where people are singing, and settings where facemasks adherence is inconsistent. To help promote the widespread avoidance of such exposure conditions, WHO supported a public messaging campaign “avoid the three Cs”: crowded places, close-contact settings, and confined and enclosed spaces.

Recent evidence supports masking as both source control and as PPE against aerosol inhalation exposure [23]. There have been reports of SARS-CoV-2 transmission among healthcare workers exposed to asymptomatic cases with surgical masks and eye protection but without fitted respirators [24, 25]. These findings emphasize the importance of airborne transmission for SARS-CoV-2, and underscore the role of respirators for PPE, and ventilation to dilute indoor contaminants.

Agency guidelines regarding building ventilation

Agencies have provided guidance for ventilation to control respiratory infections, and recently, SARS-CoV-2. There has been widespread acknowledgement of infectious aerosols contributing to the spread of infection epidemic, which has spurred efforts to advise building operators on infection control actions. Given the challenges of developing precise ventilation standards to effectively reduce airborne infection (specific beyond “more ventilation is better”), the collective wisdom of infectious disease experts and building engineers generally emphasizes some level of increased ventilation to mitigate SARS-CoV-2 spread, while maintaining thermal comfort, to reduce airborne transmission risk [26-28]. Upgrading HVAC filters to MERV13 or better, using portable air cleaners (with MERV13 filters or better), and using upper room germicidal UV (GUV) are also suggested to enhance infection control.

The Federation of European Heating, Ventilation, and Air Conditioning Associations (REHVA) recommends airborne infection isolation rooms (AIIR) to have at least 6-12 air changes per hour (ACH), with new builds having ≥12 [28]. REHVA suggests upgrading ventilation in any healthcare wards with infectious disease cases to match AIIR levels, and at least 4 ACH in other zones of healthcare facilities. The American Society of Heating, Refrigeration, and Air-Conditioning Engineers (ASHRAE) goes beyond this, advising at least 12 ACH in healthcare settings, regardless of aerosol generating procedure [29]. For school health clinics, ASHRAE provided guidance ranging from 6-10 ACH [30]. This is generally higher than the non-pandemic standards.

REHVA shows how increasing ventilation above 1L/s/m² can reduce airborne infection risk and that 4 L/s/m² may be advisable as a lower bound in office meeting rooms or classrooms (corresponding to 5 ACH) [28]. They classify maintenance of indoor CO₂ levels below 800 and 1,000ppm (given outdoor CO₂ level of 400ppm) as “good” and “acceptable” ventilation, respectively. WHO suggested 5-6 ACH for public buildings in the context of SARS-CoV-2 pandemic control [31, 32]. This is similar to what has been suggested in the US for holding K-12 school during the pandemic with 6 ACH as ideal, 5-6 as excellent, 4-5 as good, and 3-4 as bare minimum [33].

Guidelines are expected to evolve as understanding of ventilation infection control effectiveness increases. Ventilation is considered a helpful control within the context of other layers of protection including, in the context of the COVID-19 pandemic, use of face masks, physical distancing, de-densifying indoor spaces, upgrading filters in central HVAC systems, using portable air cleaners, deploying GUV, and considering air flow dynamics to reduce exposure to infectious aerosols.

1. Iddon, C., Florence Nightingale: nurse and building engineer, in CIBSE Journal. 2015, CIBSE Available from: https://www.cibsejournal.com/general/florence-nightingale-nurse-and-building-engineer/.

2. Riley, E.C., G. Murphy, and R.L. Riley, Airborne spread of measles in a suburban elementary school. Am J Epidemiol, 1978. 107(5): p. 421-432.

3. Rudnick, S.N. and D.K. Milton, Risk of indoor airborne infection transmission estimated from carbon dioxide concentration. Indoor Air, 2003. 13(3): p. 237-245. https://dx.doi.org/10.1034/j.1600-0668.2003.00189.x.

4. Bueno de Mesquita, P.J., C.J. Noakes, and D.K. Milton, Quantitative aerobiologic analysis of an influenza human challenge-transmission trial. Indoor Air, 2020. 30(6): p. 1189-1198. https://dx.doi.org/https://doi.org/10.1111/ina.12701.

5. Menzies, D., et al., Hospital Ventilation and Risk for Tuberculous Infection in Canadian Health Care Workers. Annals of Internal Medicine, 2000.

6. Gao, X., et al., Evaluation of intervention strategies in schools including ventilation for influenza transmission control. Building Simulation, 2012. 5(1): p. 29-37. https://dx.doi.org/10.1007/s12273-011-0034-7.

7. Gao, X., et al., Potential impact of a ventilation intervention for influenza in the context of a dense indoor contact network in Hong Kong. Science of The Total Environment, 2016. 569–570: p. 373-381. https://dx.doi.org/10.1016/j.scitotenv.2016.06.179.

8. Gao, X., et al., Building Ventilation as an Effective Disease Intervention Strategy in a Dense Indoor Contact Network in an Ideal City. PLOS ONE, 2016. 11(9): p. e0162481. https://dx.doi.org/10.1371/journal.pone.0162481.

9. Li, Y., et al., Role of ventilation in airborne transmission of infectious agents in the built environment - a multidisciplinary systematic review. Indoor Air, 2007. 17(1): p. 2-18.

10. Luongo, J.C., et al., Role of mechanical ventilation in the airborne transmission of infectious agents in buildings. Indoor Air, 2016. 26(5): p. 666-678. https://dx.doi.org/10.1111/ina.12267.

11. Qian, H. and X. Zheng, Ventilation control for airborne transmission of human exhaled bio-aerosols in buildings. J. Thorac. Dis., 2018. 10(Suppl 19): p. S2295-S2304. https://dx.doi.org/10.21037/jtd.2018.01.24.

12. Killingley, B., et al., Use of a human influenza challenge model to assess person-to-person transmission: proof-of-concept study. The Journal of Infectious Diseases, 2012. 205(1): p. 35-43. https://dx.doi.org/10.1093/infdis/jir701.