The impact of heating, ventilation, and air conditioning design features on the transmission of viruses, including the 2019 novel coronavirus: A systematic review of filtration

Historically, viruses have demonstrated airborne transmission. Emerging evidence suggests the novel coronavirus (SARS-CoV-2) that causes COVID-19 also spreads by airborne transmission. This is more likely in indoor environments, particularly with poor ventilation. In the context of airborne transmission, a vital mitigation strategy for the built environment is heating, ventilation, and air conditioning (HVAC) systems. HVAC features could modify virus transmission potential. A systematic review was conducted to identify and synthesize research examining the effectiveness of filters within HVAC systems in reducing virus transmission. A comprehensive search of OVID MEDLINE, Compendex, and Web of Science Core was conducted to January 2021. Two authors were involved in study selection, data extraction, and risk of bias assessments. Study characteristics and results were displayed in evidence tables and findings were synthesized narratively. Twenty-three relevant studies showed that: filtration was associated with decreased transmission; filters removed viruses from the air; increasing filter efficiency (efficiency of particle removal) was associated with decreased transmission, decreased infection risk, and increased viral filtration efficiency (efficiency of virus removal); increasing filter efficiency above MERV 13 was associated with limited benefit in further reduction of virus concentration and infection risk; and filters with the same efficiency rating from different companies showed variable performance. Adapting HVAC systems to mitigate virus transmission requires a multi-factorial approach and filtration is one factor offering demonstrated potential for decreased transmission. For filtration to be effective, proper installation is required. Of note, similarly rated filters from different companies may offer different virus reduction results. While increasing filtration efficiency (i.e., increasing MERV rating or moving from MERV to HEPA) is associated with virus mitigation, there are diminishing returns for filters rated MERV 13 or higher. Although costs increase with filtration efficiency, they are lower than the cost of ventilation options with the equivalent reduction in transmission. Systematic review registration: PROSPERO 2020 CRD42020193968.


Introduction
In March 2020, the World Health Organization (WHO) declared a global pandemic due to Coronavirus Disease 2019  which is attributed to Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) [1].Worldwide, public health authorities have sought evidence regarding routes of virus transmission and corresponding public health measures to mitigate virus spread.Certain viruses can be transmitted via an aerosol route [2].Aerosol transmission is facilitated by virus-laden aerosols expelled by humans, remaining airborne for extended periods of time.Recent evidence suggests that SARS-CoV-2 can spread via airborne transmission, particularly in indoor environments with poor ventilation [3,4].In April 2021, the American Society of Heating, Refrigerating, and Air-Conditioning Engineers (ASHRAE) released a statement declaring that "[a]irborne transmission of SARS-CoV-2 is significant and should be controlled.Changes to building operations, including the operation of heating, ventilating, and air-conditioning systems, can reduce airborne exposures."[5] Determining appropriate measures to protect occupants of indoor spaces based on informed, interdisciplinary research is critical to managing the spread of infectious disease [6].
Heating, ventilation, and air conditioning (HVAC) systems can mitigate airborne transmission of viruses by removing or diluting contaminated air inside the building envelope where humans breathe [6][7][8][9].HVAC design features, such as, filtration, ventilation, ultraviolet radiation, and humidity can influence transmission.The first official air filter testing standard in the HVAC industry was published by ASHRAE in 1968 (ASHRAE 52-68 [10]) with an update in 1976 (ASHRAE .In Europe, ASHRAE 52-76 was adapted as Eurovent 4/5 which classified filters from EU1 to EU9 [10].In 1999, ASHRAE 52.2 introduced the classification Minimum Efficiency Reporting Value (MERV).ASHRAE 52.2 was modified in 2007, 2012, and 2017.By 2016 in Europe, ISO 16890 replaced EN779 which classified filters as G1-G4; M5-M6; F7-F9 [10].EN779 had replaced Eurovent 4/9 which superseded Eurovent 4/5 [10].In ASHRAE Standard 52.2-2017, filter efficiency is based on particle size removal efficiency (PSE) [11].In other words, filter efficiency is the fraction of particles removed from air passing through the filter [12].Particles fall into three size ranges: E 1 0.30-1.0μm; E 2 1.0-3.0μm; and E 3 3.0-10.0μm.The PSEs in these size ranges are used to determine the MERV rating, which ranges from MERV 1 to MERV 16.Another common filter is a High Efficiency Particulate Air (HEPA) filter; the filter efficiency of a HEPA filter is better than MERV 16 [12].Based on these dates, studies and recommendations may report a variety of filter efficiencies from EU, F, MERV, and HEPA.
In January 2021, ASHRAE made recommendations for reducing airborne infectious aerosol exposure, which included the use of MERV 13 or higher filters for air recirculated by HVAC systems [5].When accessed in April 2021, the ASHRAE website indicated that MERV 13 is recommended but MERV 14 or better is preferred [13].In March 2021, the WHO released a roadmap concerning indoor ventilation during COVID-19 [14].Based on a scoping review that identified six studies specific to SARS-CoV-2 and technical guidance by leading international HVAC organizations, the roadmap presents standards for healthcare, non-residential and residential environments based on type of ventilation (natural vs. mechanical).In specific situations (e.g., depending on ventilation rate and airflow patterns, system designs with air recirculation modes or heat recovery), HEPA or MERV 14 filters are recommended [14].This systematic review considered whether virus transmission is affected by HVAC design features, particularly, filtration.In this review, an extensive and comprehensive search of the literature was conducted to identify and synthesize published research evaluating the effectiveness of filtration in reducing virus transmission.The insight drawn from this review could help answer questions of the utility of filtration to mitigate the transmission of SARS-CoV-2 in mechanically ventilated indoor environments.Further, understanding the association between filter efficiency and infection risk could inform control measures.Finally, a comprehensive synthesis of the existing scientific literature can identify gaps and guide priorities for future research.

Methods
This systematic review addressed the research question: is virus transmission affected by HVAC design features, particularly, filtration?The current review was part of a larger research program to review literature on HVAC design features (ventilation, filtration, ultraviolet radiation, and humidity) and virus transmission.Results for other HVAC design features of interest are reported separately due to the volume and heterogeneity of the research.The systematic review is registered [15] and an a priori protocol was developed and is publicly available [16].Standards for the conduct of systematic reviews defined by the international Cochrane organization [17] were followed with modifications for questions of etiology [18].The review is reported according to accepted reporting standards [19].

Search strategy
From inception to June 2020, three electronic databases (Ovid MEDLINE, Compendex, Web of Science Core) were searched by a research librarian (GMT) using concepts related to virus, transmission, and HVAC.The Ovid MEDLINE search strategy is detailed in Table 1.Prior to implementing the searches, the strategies were peer-reviewed by two librarians (TL, AH).In January 2021, the search was updated.Reference lists of all relevant papers were screened, in addition to relevant review articles.Conference abstracts were identified through Compendex and Web of Science and were not included, but literature was searched to see whether any potentially relevant abstracts had been published as complete papers.Limits by year or language of publication were not placed on the search; however, only English-language studies were included due to the volume of available literature and resource constraints.EndNote was used to manage references with duplicate records removed prior to screening.

Study selection
Study selection had two stages.First, two reviewers independently screened titles and abstracts of references identified by electronic database searches.Relevance of each record was classified as Yes, Maybe, or No. Any conflicts between Yes/Maybe and No were resolved by one reviewer.To ensure consistency among the review team, pilot testing was conducted with three sets of studies (n = 199 each).The review team met to discuss discrepancies and develop decision rules after each set of pilot screenings.Second, two reviewers independently reviewed full text articles and applied inclusion and exclusion criteria.Reviewers classified studies as Include or Exclude.Any conflicts between Include and Exclude were resolved by consensus of the review team.One reviewer resolved conflicts between different exclusion reasons.The second stage of screening was pilot tested with three sets of studies (n = 30 each).The review team met to resolve discrepancies after each pilot round.Screening was conducted using Covidence software.

Inclusion and exclusion criteria
Inclusion and exclusion criteria are listed in Table 2.This systematic review was part of a larger effort to examine different HVAC design features and virus transmission.While all design features were included in the search and screening process at once, only studies evaluating filtration were synthesized here.A variety of agents were included in the search; however, studies of viruses or agents that simulated viruses were prioritized.Other agents (e.g., bacteria, fungi) would be included only if studies were not available that were specific to viruses.Studies using bacteriophages, which are viruses that infect bacterial cells [20], were included.This review focused on the indoor built environment (e.g., office, public, residential buildings) which had mechanical ventilation.Primary research that provided quantitative results of the correlation

# Searches
or association between filtration and virus transmission was included.No restrictions were placed on year of publication and only English-language, peer-reviewed publications were included.

Risk of bias assessment
For experimental studies, risk of bias was assessed based on three key domains: selection bias, information bias and confounding [21,22].Reviewers assessed each domain as low, unclear, or high risk of bias using signalling questions [17] from guidance documents for the different study types that were included; e.g., animal studies, laboratory experiments, epidemiological studies [21][22][23].For modelling studies, risk of bias was assessed using the following three key domains: definition, assumption, and validation [23,24].The domain definition considers model complexity and data sources, assumption considers the description and explanation of model assumptions, and validation considered model validation and sensitivity analysis [24].Reviewers assessed each domain as low, unclear, or high risk of bias based on signalling questions [23][24][25].Risk of bias items were pilot tested among three review authors.Then, two reviewers (GMT, BAF) applied the criteria independently to each relevant study and met to resolve discrepancies.

Data extraction
Reviewers extracted general information about the study (authors, year of publication, country of corresponding author, study design) and methods (setting, population [as applicable], agent studied, intervention set-up).Details on filtration parameters (where available) were extracted, including any filter efficiency ratings or classifications (such as MERV and EU ratings or HEPA) or filter efficiencies expressed as percentages, where applicable.More recent filter efficiencies are based on ASHRAE Standard 52.2-2017 and represent particle size removal efficiencies (PSE) with particles falling into three size ranges: E 1 0.30-1.0μm; E 2 1.0-3.0μm; and E 3 3.0-10.0μm [11].Efficiency of particle removal is usually represented by conventional filter efficiency ratings, such as MERV, EU, F, and HEPA.Quantitative data were extracted, as well as results of any tests of statistical significance related to filtration features.The primary outcome of interest was quantitative measures of the association between filtration and virus transmission.As such, data on actual transmission were extracted where available (i.e., infections), as well as information regarding viruses, filters, setting and experimental test set-up, and outcomes.The distinction between the efficiency of particle removal described above and the efficiency of virus removal is important.The efficiency of virus removal is calculated as the difference between the virus concentration upstream of the filter and the virus concentration downstream of the filter divided by the virus concentration upstream of the filter expressed as a percentage.Only one aerosolized virus study used terminology that referred explicitly to virus, viral filtration efficiency [26], while the remaining studies on virus removal used a variety of terms typical for particle removal: removal efficiency [27], filtrating efficiency [28], filter efficiency [29], and filter reduction efficiency [30,31].European ratings were presented with MERV rating conversions [32].Using a data extraction form spreadsheet to ensure comprehensive and consistent capture of data, one reviewer extracted data and a second reviewer verified data for accuracy and completeness.Any discrepancies were discussed by the review team.

Data synthesis
As anticipated, meta-analysis was not possible due to heterogeneity across studies in terms of study design, filtration features examined, outcomes assessed, and reporting of results.Evidence tables were developed to describe the studies and their results, with a narrative synthesis of the results.To allow for meaningful synthesis and comparison across studies, the studies are presented and discussed in three groups: animal studies, aerosolized virus studies, and modelling studies.

Results
12,177 unique citations were identified following searches executed using the search protocol; 2,428 were identified as potentially relevant following title/abstract screening and 568 met the inclusion criteria following full-text review (Fig 1).Fifty-eight of the 568 were relevant to filtration; 23 studies met the inclusion criteria for filtration and viruses following full-text review.
Five of the seven studies found that filtration was associated with decreased virus transmission [33,[35][36][37]39].Pitkin et al. also found that filtration was associated with decreased daily risk of infections for PRRSV [35].One study found the use of filtration was associated with decreased presence of PRRSV downstream of the filter [34] and another [38] found the use of filtration was associated with decreased relative risk of PRRSV introductions.
Dee et al. [33] and Hopkins and Drury [39] found that increased filter efficiency was associated with decreasing transmission.Hopkins and Drury documented that improper installation of a high efficiency filter allowed transmission to occur [39].Additionally, Dee et al. found no difference in the transmission reduction between MERV 14 and MERV 16 with no transmission events in both situations [36].Two studies found that filter performance varied by manufacturer when testing similarly rated filters from different companies [34,39].

Aerosolized virus studies (n = 7, Table 4)
Five of seven aerosolized virus studies found that filtration was associated with virus removal using PRRSV [30], equine arteritis virus (EAV) [30], bovine enterovirus (BEV) [30], human adenovirus [29], and the bacteriophages T7 [27,28] and MS2 [31].One study found that increasing MERV rating was associated with decreasing virus concentration for bacteriophage T4 at different locations within an apartment unit [40], and another found that increasing MERV rating was associated with increasing viral filtration efficiency using bacteriophage MS2 in a test duct set-up [26].
Like the animal studies by Dee et al. [33] and Hopkins and Drury [39], two aerosolized studies found increasing MERV ratings to be associated with increased virus removal efficiencies [26,40].Zhang et al. noted that the filter efficiency associated with E 1 (i.e., 0.30-1.0μm; the smallest particle size range for MERV classification) was a good approximation of the viral filtration efficiency [26].In addition, prior to the use of MERV ratings as developed in 1999, Washam et al. identified cost differences between filters of similar efficiency [27].
From the nine modelling studies, five studies found that increasing filter efficiency associated with decreased virus concentration [41,46] and decreased risk, including absolute risk and relative risk of infection [42], transmission risk [44], and infection risk [48].One study found that the use of filtration was associated with decreased attack rate [43].Two studies found HEPA filtration was associated with decreased risk [45] and decreased probability [49] of SARS-CoV-2 infection.Finally, Janni et al. found that a combination of MERV 8 prefilter and MERV 16 was associated with a lower concentration of PRRSV compared to a combination of MERV 8 prefilter and MERV 14 [47].
Of the nine modelling studies, four evaluated the use of a single filter [43,45,48,49]; all agreed that installing a filter is more beneficial than not installing a filter.Like the experimental studies of Dee et al. [36] and Spronk et al. [37], three modelling studies comparing single filters [41,42,46] agreed that there is a threshold beyond which increasing filter efficiency may not affect the probability of infection.One study using a MERV 8 prefilter was able to model a difference in virus concentration comparing MERV 14 and MERV 16 [47].Mazumdar and Chen compared HEPA filters of different efficiencies and determined that improvement was not evident when increasing filter efficiencies past 99.90%[41].Azimi and Stephens [42] and Brown et al. [46] specifically found that there were diminishing returns for MERV 13 or higher filters when using influenza, while Dee et al. [36] and Spronk et al. [37] found no difference between MERV 14 and MERV 16 filters for PRRSV.Azimi and Stephens [42], using influenza, compared varying MERV filters and a HEPA filter and found that, for filters MERV 13 or greater, the risk reduction was not limited by the filter efficiency; instead by the amount of recirculating air through the HVAC system.Similar to Washam et al. [27], Azimi and Stephens considered the cost of filtration [42].Compared with the cost of an equivalent amount of outdoor air ventilation, they found filtration was the less expensive option for all MERV ratings except MERV 4 [42].

Risk of bias
Twelve of the 14 experimental studies had low risk of bias for all three domains: selection bias, information bias, confounding.Two experimental studies had low risk of bias for selection bias and confounding but had unclear risk of bias for information bias because of a lack of  clarity in the methodological details [28,29].Seven of the nine modelling studies had low risk of bias for all three domains: definition, assumption, validation.Two modelling studies had low risk of bias for assumption and validation but had unclear risk of bias for definition because of a lack of clarity about the HEPA filter efficiency [45] and contribution of fresh air [49].

Discussion
While there is substantial literature on the use of filters to remove particles, this review was designed to consider the use of filters to remove viruses.The 23 articles that satisfied our strict criteria revealed several important findings.First, filtration was associated with decreased virus transmission.Second, filters removed viruses from the air.Third, increasing filter efficiency (efficiency of particle removal) was associated with decreased transmission, decreased infection risk, and increased viral filtration efficiency (efficiency of virus removal).Fourth, increasing filter efficiency above MERV 13 was associated with limited benefit in further reduction of virus concentration and infection risk.Fifth, filters with the same filter efficiency rating from different companies had varied performance regarding transmission in animal studies.

Implications for practice
Some findings are straightforward to address in practice.For example, proper installation is required for a filter to be effective in removing viruses and preventing transmission [39].
Other findings are challenging to address in practice.Two studies documented variations in the performance of similarly rated filters that were purchased from different manufacturers [36,39].Variations in performance can occur in filters with the same MERV rating from the same manufacturers, not just different manufacturers.This alone implies that within the broad criteria of MERV ratings, there is enough variation in performance to point to a need for more filter sub-ranges or more stringent testing methods of MERV value.A practical benchmark emerging from this review was, from the findings of Zhang et al. [26], that a conservative estimate of the viral filtration efficiency could be approximated by the E 1 efficiency used to determine MERV rating in ASHRAE Standard 52.2-2017.Not surprisingly, increasing filter efficiency, whether based on increasing filter efficiency before MERV ratings were established [39], increasing MERV rating [42], or increasing from MERV to HEPA [34], was associated with mitigation of virus transmission.An important practical consideration is that the improvement is limited above the filter efficiency MERV 13.This concept of diminishing returns was documented in both animal [36,37] and modelling studies [41,42,46].Spronk et al. [37] and Dee et al. [36] found that both MERV 14 and MERV 16 were associated with no transmission, while Azimi and Stephens [42] and Brown et al. [46] showed that filters rated MERV 13 or better were associated with limited improvement.One model demonstrated decreased virus concentration with MERV 14 versus MERV 16 but both had MERV 8 prefilters [47].While this could suggest an important role of prefilters at higher MERV ratings, the results are difficult to generalize across MERV ratings because only two were modelled.Although one study investigated prefilters, no studies examined filter replacement frequency within the context of viral filtration efficiency.
Diminishing returns are not limited to MERV ratings as Washam et al., prior to the use of MERV ratings, found that similar efficiency can be achieved using a lower cost option [27].In addition, the cost increases with increasing MERV rating and HEPA filtration relative to the cost for MERV 13; however, the cost of filtration for these filters remains less than the cost of outdoor air ventilation with the equivalent transmission reduction [42].This finding is consistent with current ASHRAE recommendations for reducing airborne infectious aerosol exposure [12].Interestingly, ASHRAE indicates that MERV 13 is recommended but MERV 14 is preferred [13].The use of MERV 14 is supported by WHO in the context of COVID-19 [14].
Installing or upgrading an air filtration system can sometimes be an expensive intervention.However, illness costs money as well and there is considerable research available showing simply how improved indoor air quality (IAQ) infrastructure investment is profitable when building owners and users interests are aligned [50].Installing air filtration on large sow farms is estimated to cost $150-200 per sow [51].Alonso et al. postulates that air filtration like this is quite common now due to the major economic impact of PRRSV on swine production [38].When considering poultry, Hopkins and Drury agreed that the additional cost of high efficiency filters would be justified by the investment return of the flock [39].A single influenza case results in an approximate economic loss of $375 in the United States [52].Azimi and Stephens found that upgrading a MERV 7 filter to MERV 13 would only be $17 in labour cost annually in a typical office environment while reducing the number of infections by 1 [42].Not only does air filtration reduce the number of viral infections, it improves the overall air quality by reducing the amount of allergens and asthma triggers in the air.An extra $50 per year for high efficiency air filters in a family residence is seen as a small additional cost against the enormous financial and health burden of asthma [46].Some studies have shown that there can be a plateau in effectiveness for filters past MERV 13 in typical situations [42,46].Mazumdar and Chen found that a 94% efficiency filter would not be able to protect passengers sitting far away from the infectious source in an airliner cabin and that a 99.9% filter was needed for their application [41].They note that further increase in filter efficiency might be difficult and its effectiveness not clear.All filters with the same rating are not equal either.Differences in the filter media or frame design across companies and inconsistency in the filter rating systems across countries can contribute to discrepancies in filter efficiency [34].The source of the raw material of the filter media can also impact filter quality [35].Zhang et al. found that the viral filtration efficiency (VFE) is generally correlated with the MERV rating but they are not the same [26].They found that the VFE is always higher than E1, but lower than E2 or E3 efficiencies.This is all to say that choosing the right filter for the right application is important.The location, use case, existing system, budget, and acceptable level of risk should be taken into account when considering air filtration systems [33,36,37].
When considering scientific testing of filters, the size of the aerosol challenge should be considered very carefully.Many studies have stated that the level of contamination in their experiments was probably much larger than what would be expected in their respective typical situations [27,34,39,40].This could lead to certain interventions being wrongfully deemed inadequate.
Filtration should not be seen as the only approach to infection mitigation.Filtration should be paired with other infection mitigation strategies such as vaccination [44], distancing shared space users, and implementing a wait time before different users access the space [49], and other strict biosecurity measures [30].Installation of a high efficiency filter increases the pressure drop which can lead to greater energy use if the system has to run more often to service the same volume of air.If installing or upgrading the air filtration system is not possible, Bandaly et al. suggests switching to 100% outdoor air because some viral particles that pass through an inferior filter will still be infectious [29].
From our review of the literature, it is clear that upgrading or installing high efficiency air filtration is cost effective in the long run and reduces virus transmission.It is recognized that not all spaces are the same and it might not be feasible to implement high efficiency air filtration.If this is the case, then other infection mitigation measures such as UVGI, increasing ventilation rate, or controlling the relative humidity should be considered.

Implications for research
The modelling studies spanned from 2009 to 2020 and the aerosolized virus studies spanned from 1966 to 2021; some of the more recent studies were motivated by the COVID-19 pandemic.Many viruses were examined; however, it is particularly interesting that there were only two coronavirus studies (both SARS-CoV-2) for filtration, both modelling, and only one aerosolized study that used MS2 as a potential surrogate for SARS-CoV-2 based on size [26].Additionally, one of the studies discussed the implications of filtration in relation to SARS--CoV-2 but used influenza to set relevant infectivity parameters [49].The limited number of coronavirus (and specifically SARS-CoV-2) studies suggests a potential research gap to be addressed in the future.There is a very rich body of literature of engineering or laboratory studies which focuses on collecting physical metrics expected to affect transmission; these were not considered in this review as our focus was on connecting transmission itself with installed equipment.The trade-off between isolating specific parameters related to filtration physics or analysing more combined links in the chain of transmission is an important consideration and we stress that the literature gaps tend to be in studies which consider systems holistically.This points to a need for interdisciplinary studies that combine knowledge of well controlled filter experiments with less well controlled "real world" situations which inherently capture combined synergies in the transmission chain as well as possible "unknown unknowns" in that chain.
Interestingly, the animal studies only spanned from 1971 to 2013.Furthermore, no field studies examining potential mitigation of virus transmission in humans using filtration were found.This suggests a research opportunity to investigate the use of filtration on mitigating virus transmission in "real-world" built environments occupied by humans, as well as epidemiological studies during disease outbreaks.Taken together, these findings identify an opportunity for more research that investigates HVAC filtration mitigation of human viruses and, specifically, human coronaviruses.
More recent aerosolized virus studies use the ASHRAE Standard 52.2 test duct [26,31].This establishes an important level of consistency for future research.The aerosolized virus studies that state an explicit interest in SARS-CoV-2 used the bacteriophage MS2, which has a smaller size than SARS-CoV-2.Future research could explore the ramifications, if any, of this size differential on virus transmission given the size of droplet or droplet nuclei in which the virus itself is suspended.
Of note, we included studies that were published up to 2021 and there may be relevant studies that have been published since then.However, in a search of PubMed, Compendex and Web of Science in January 2023, we did not identify any studies that would alter the conclusions of this review (https://www.mcmasterforum.org/docs/default-source/productdocuments/living-evidence-syntheses/covid-19-living-evidence-synthesis-15.1-effectivenessof-ventilation-for-reducing-transmission-of-covid-19.pdf?sfvrsn=b1237b23_5).

Conclusion
Evidence clearly demonstrates that filters are effective in removing viruses from the air and reducing the potential for virus transmission and risk of infection.However, experimental and modelling studies indicate diminishing returns above MERV 13, with associated cost implications.Further, above MERV 13, other HVAC features (e.g., amount of recirculating air) may impact risk reduction beyond filtration efficiency.This systematic review identified no field studies or epidemiological investigations of the "real-world" effectiveness of filters in mitigating virus transmission in humans, representing an important gap and priority for future research.Only two modelling studies of SARS-CoV-2 and two experimental studies of MS2 were directly relevant to the current COVID-19 pandemic and recent coronavirus outbreaks.

Fig 1 .
Fig 1. Flow of studies through the selection process (note: Search was conducted for all HVAC design features but only studies of filtration are included in this manuscript).https://doi.org/10.1371/journal.pgph.0002389.g001