The implementation of portable air-cleaning technologies in healthcare settings – a scoping review

Abstract

The COVID-19 pandemic revealed opportunities to improve prevention practices in healthcare settings, mainly related to the spread of airborne microbes (also known as bioaerosols). This scoping review aimed to map methodologies used to assess the implementation of portable air cleaners in healthcare settings, identify gaps, and propose recommendations for future research. The protocol was registered in the Open Science Framework and reported following the checklist provided by the Preferred Reporting Items for Systematic Reviews and Meta-Analysis – an extension for Scoping Reviews (PRISMA-ScR) statement. The search strategy was performed in five databases and one grey literature source. At the last selection phase, 24 articles that fulfilled our inclusion criteria were summarized and disseminated. Of these, 17 studies were conducted between 2020 and 2022; one of them was a protocol of a multicentre randomized controlled trial. The outcomes measured among the studies include airborne microbe counts, airborne particle concentrations, and rate of infections/interventions. The leading healthcare settings assessed were dental clinics (28%), patient's wards (16%), operating rooms (16%), and intensive care units (12%). Most of the devices demonstrated a significant potential to mitigate the impact of bioaerosols. Although some indoor air quality parameters can influence the mechanics of aerosols, only a few studies controlled these parameters in their analyses. Future clinical research should assess the rate of infections through randomized controlled trials with long-term follow-up and large sample sizes to determine the clinical importance of the findings.

Introduction

The COVID-19 pandemic raised awareness of the high risk that represents respiratory pathogens spread by aerosols – particularly in enclosed spaces with poor ventilation [1]. Although it is not a new concern in health facilities since it has long been a top priority for the Centers for Disease Control and Prevention (CDC), the pandemic outbreaks did evidence the need for evolution and innovation to mitigate the impact of aerosols [2].

Aerosols are liquid or solid particles suspended in the air by natural or artificial sources. Depending on their weight, these particles can remain suspended in the air for hours and travel long distances through the airborne route [3]. When aerosols transport micro-organisms, such as bacteria, fungi, spores, and viruses, they are also known as bioaerosols [4]. In cases when it is impossible to reduce the sources of aerosols or the dilution ventilation is insufficient, the implementation of portable air cleaners has been proposed as a coadjutant measure in residential, commercial buildings, and healthcare settings [5].

The portable air cleaners use different technologies such as fibrous media air filters, generally rated as high-efficiency particulate air filters (HEPA) or ultra-low particulate air filters (ULPA), ultraviolet air filtration, and electronic air cleaners, including electrostatic precipitators and ionizers, alone or in combination [6]. Fibrous media air filters remove particles by capturing them on fibrous filter materials. Electrostatic precipitators and ionizers remove particles by an active electrostatic charging process. Ultraviolet air filtration reduces viable airborne micro-organisms by killing or deactivating them [6]. Gas-phase air-cleaning technologies include adsorbent air filters such as activated carbon, chemisorbed media air filters, photocatalytic oxidation, plasma, and intentional ozone generators, designed to remove gaseous air pollutants or convert them to harmless [6].

The effectiveness of different portable air cleaners was reported to range from 12 to 99% depending on the technology used, setting, and outcome assessments across the studies [7]. In theory, one can assume that lowering the airborne particle concentrations and airborne microbial counts in the indoor air would result in lower rates of infection. This scoping review aimed to map and summarize overall research (published and grey literature) assessing the implementation of portable air-cleaning technologies in healthcare settings; additionally, to report the outcomes measured across the studies, the characteristics and range of the used methodologies, challenges, and limitations, and to propose recommendations for future research.

Methods

Protocol and registration

This scoping review was registered in the OSF database (doi: https://osf.io/8g9ap), conducted following the guidelines for conducting systematic scoping reviews of the JBI Briggs Reviewers Manual, and reported following the checklist provided by the PRISMA-ScR statement (Supplementary Table S1) [8,9].

Eligibility criteria and search strategy

The inclusion criteria were guided by the review question: What outcomes have been measured in existing research to assess the implementation of portable air cleaners in healthcare settings? Studies that aimed to assess the implementation of portable air-cleaning devices in healthcare settings (medical and dental clinics and offices, urgent care centres, large hospitals) in real or quasi-real-life scenarios compared to no implementation were considered. No limitations of language or publication date were established.

An initial limited search was performed in PubMed to analyse text words contained in the title and abstracts across the articles that fulfilled our eligibility criteria. The following MeSH terms and keywords were combined: Hospital∗ OR ‘Health Facilities’ OR ‘Dental Clinic∗’ AND ‘Air Filters∗’ OR ‘Air Purifier∗’ OR ‘Portable Air Cleaner∗’ OR ‘Air Circulation’ OR ‘Air Filtration’ OR ‘High-Efficiency Particulate Air Filter∗’ OR ‘Ultraviolet Air Filtration’ OR ‘Plasma Air Filtration’. After selecting keywords and index terms, a second search was performed by two independent reviewers. Five databases (PubMed, Embase, Scopus, Cochrane Library, and Web of Science) were used to identify all the published articles on the topic, and one grey literature source (Grey Matters) was used to identify unpublished articles (Supplementary Table S2).

The reference list of included articles was also assessed to search for additional studies, and search alerts were activated in each database. Sources were last accessed in June 2022. All citations found were imported into a reference manager (EndNote, version 20.3, Thomson Reuters), and duplicates were removed automatically and manually.

Extraction of data and charting

Two independent authors (M.A. and J.D.) extracted and charted data from the included studies. The following information was tabulated: author, country, year of publication, aims, healthcare setting, description of the device used (commercial name, airflow settled, noise, and type of technology), and outcomes measured. The studies were also summarized and charted according to the outcome assessed, describing how these outcomes were measured (methodology and measurement tool) and reported results.

Results

Study selection

In total, 2023 citations were identified, and 425 duplicates were removed. After title and abstract analyses, 31 articles were selected for full-text reading, of which one report was not retrieved. Six articles were excluded: one extended abstract [10], four experimental studies performed in a test chamber or a simulated room [[11], [12], [13], [14]], and one study in which the air filter was installed in the heating, ventilation, and air-conditioning (HVAC) system [15]. Thus, 24 articles were summarized and disseminated in this scoping review (see PRISMA flow diagram, Figure 1 ).

Figure 1.

PRISMA 2020 flow diagram for identification and screening of studies. HVAC, heating, ventilation, and air-conditioning system.

Characteristics of the included studies

Of the 24 studies that fulfilled our inclusion criteria [[16], [17], [18], [19], [20], [21], [22], [23], [24], [25], [26], [27], [28], [29], [30], [31], [32], [33], [34], [35], [36], [37], [38], [39]], 17 studies were conducted and published between 2020 and 2022 during the COVID-19 outbreaks. However, only one study described the detection of SARS-CoV-2 RNA in air samples collected in addition to other airborne microbes (a range of other bacterial, viral, and fungal pathogens) [23]. The outcomes measured – alone or combined – among the included studies were airborne microbial counts, airborne particle concentrations, and rate of infections or interventions (Figure 2 ). The characteristics of the included studies are described in Table I , including the commercial names and technology of the devices tested, as well as the main technical descriptions when the authors or the manufacturers' websites supplied the information.

Figure 2.

Outcome measurements across the included studies to assess the implementation of portable air cleaners in healthcare settings over time.

Table I.

Characteristics of the included studies, the devices used, and the outcomes measured in the different healthcare settings

Study/country	Healthcare setting (area/volume)	Description of the portable air cleaner tested				Outcomes measured
		Device	CADR (m3/h)	Noise (dB)	Technology	AM	AP	RI
Persson et al.a [16] Sweden	ORs (NI) in seven hospitals	Novaerus Protect 800	260	45	Plasma	No	No	Yes
Capparè et al. [17] Italy	Dental OR (20 m2/60 m3) in a hospital	Professional XXL inn-561	NI	38 and 58	HEPA 14	Yes	Yes	No
Oberst et al. [18] Germany	OR (21 m2/52 m3) in an orthopaedic clinic	AP-40 Air filter	320	35–51	HEPA 13; activated carbon; plasma	No	Yes	No
Arikan et al. [19] Turkey	ICUs (105 m2/315 m3) in a hospital	Novaerus Defend 1050	267	67	HEPA 13; activated carbon	Yes	No	Yes
		Novaerus Protect 800	260	45	Plasma
Corrêa et al. [20] Brazil	Emergency care unit (NI)	Non-commercial device	781	NI	UV-C lamps	Yes	No	No
Maurais et al. [21] Canada	A dental OR and a mobile dental OR (NI)	MedEVAC-A	255	56	HEPA	No	Yes	No
		AF400M HEPA	510	52	HEPA
		XPOWER X-2580 Professional	510	NI	HEPA; activated carbon
Tzoutzas et al. [22] Greece	Dental OR (170 m2/510 m3) in a dental school	Aurabeat AG+ NSP-X1	375	≤58	UV lamp; silver Ion; plasma	No	Yes	No
Morris et al. [23] UK	Repurposed ‘surge’ COVID ward (39.52 m2/NI) and ‘surge’ ICU (72.22 m2/NI) in a hospital	AC1500 HEPA14/UV	400–1000	≤50 and ≤42	HEPA 14; UV-C lamp	Yes	No	No
		Medi 10 HEPA13/UV	700–1300	35–60	HEPA 13; ozone-free UV-C; activated carbon
Buising et al. [24] Australia	Patient’s ward (12.8 m2/37 m3) in a hospital	AX5500K Air Purifier	467	21–50	HEPA 13; activated carbon	No	Yes	No
Lee et al. [25] Australia	A single-bed patient room (2 m2/37 m3) in a hospital	Industrial air cleaner Model A	200	NI	HEPA	No	Yes	No
		Industrial air cleaner Model B	400	NI	HEPA
		AX60RR5080WD	467	NI	HEPA; activated carbon
Razavi et al. [26] Canada	A dental OR (9 m2/36 m3) in a dental clinic	JADE, SCA5000C	260 and 530	NI	HEPA; UV-C lamps; activated carbon, PCO	No	Yes	No
Ren et al. [27] USA	10 dental OR (NI) in a dental clinic	Honeywell 50250	425	NI	HEPA	No	Yes	No
Verbeure et al. [28] Belgium	Room for oesophageal HRM (20 m2/NI) in a university hospital	City M Air Purifier	435	16–53	HEPA/molecular	No	Yes	No
Messina et al. [29] Italy	An ISO-7b OR (NI/90 m3) in a hospital	Illuvia® 500 UV	850	NI	HEPA; UV-C lamps; PCO	No	Yes	No
Pouvaret et al. [30] France	A 12-bed adult haematology unit (NI/66 m3) in a cancer institute	Air handling unit R4000™	8000	NI	ULPA 15; UV-C lamps	Yes	Yes	No
Rao et al. [31] USA	Paediatric wards setting (NI) in a hospital	PECO Air Purifier MH1	NI	NI	PECO	No	No	Yes
Anis et al.a [32] USA	OR for joint arthroplasty in a hospital	T1 C-UVC system	850	NI	C-UVC chamber	Yes	Yes	No
Bischoff et al. [33] USA	Emergency rooms (NI) in a hospital	Illuvia® 500 UV	850	NI	C-UVC; chamber; PCO	Yes	No	No
Ozen et al. [34] Turkey	A haematology ward (NI) in a teaching hospital	Uvion Air Aseptizör	2500	55	HEPA 14; UV-C lamps	No	No	Yes
Le et al. [35] Vietnam	ICU (NI/125 m3) in a hospital	A non-commercial device	250	NI	UV-A lamps; activated carbon; PCO	Yes	No	No
Abdul Salam et al. [36] Singapore	Six wards (NI) in an acute tertiary-care teaching hospital	HealthPro 150	350	NI	HEPA	No	No	Yes
Hallier et al.a [37]	Three separate dental OR (NI) in a teaching dental hospital	FlexVac™	500	NI	HEPA	Yes	No	No
Chotigawin et al. [38] USA	A renal unit (86.4 m2/259 m3) in a hospital	A non-commercial device	NI	NI	HEPA; PCO	Yes	No	No
Pelleu et al. [39] USA	Three dental OR (NI/45 m3, 51 m3, and 92 m3)	NI	1360	NI	HEPA	Yes	No	No

CADR, clean air delivery rate; UV, ultraviolet; AI, average irradiance; AM, airborne microbial count; AP, airborne particle concentration; RI, rates of infections or interventions; HEPA, high-efficiency particulate absorbing filter; OR, operating room; NI, not informed; ICU, intensive care unit; PCO, photocatalytic oxidation; HRM, high-resolution manometry; ISO, International Organization for Standardization; PECO, photo-electrochemical oxidation; C-UVC, crystalline UV-C; ULPA, ultra-low particulate air filter.

^aA study protocol or a pilot study.

^bISO 1 indicates the cleanest and ISO 9 the dirtiest air.

The healthcare settings assessed were dental clinics (28%), patients' wards (16%), operating rooms (16%), intensive care units (12%), single-bed patient rooms (8%), emergency units (8%), renal units (4%), and haematology units (4%), and rooms for high-resolution oesophageal manometry (4%), including teaching hospitals and clinics in different countries (Australia 9%, Belgium 4%, Brazil 4%, Canada 9%, Germany 4%, France 4%, UK 9%, Greece 4%, Italy 8%, Sweden 4%, Singapore, Turkey 8%, USA 25%).

Assessment of airborne microbial counts using portable air cleaners

Airborne microbial counts were assessed in 11 studies (three of them were assessed in addition to airborne particles), one randomized clinical trial, and 10 in-situ experiments in real-life scenarios by different sampling methods and measurement tools [17,19,20,23,30,32,33,35,[37], [38], [39]] as shown in Table II . Arikan et al. and Pouvaret et al. assessed surface microbial counts in addition to airborne microbial counts by a surface swab test [19,30].

Table II.

Summary of the methodologies, sampling methods, and outcomes assessing airborne microbes

Study	Study design	Source of aerosols	Bioaerosols measured	Sampling/calibration/analysis	Indoor air parameters			Outcomes reported using PACs
					ACH	T (°C)	RH (%)
Capparè et al. [17]	Randomized clinical trial	Dental AGPs	Airborne bacteria, yeast, and fungi	Active sampling by impaction with Petri dishes containing TSA/250 L per point/direct counting on plates (cfu/m3)	–	–	–	Significant reduction in airborne bacterial and fungal counts
Arikan et al. [19]	In-situ experiment in a real-life scenario	Patients and procedures	Airborne bacteria	Active sampling by impaction with Petri dishes containing SBA/100 L per point/direct counting on plates (cfu/500 L)	–	20–25	30–60	Significant reduction in airborne bacterial counts
Corrêa et al. [20]	In-situ experiment in a real-life scenario	Patients and procedures	Airborne bacteria and fungi	Passive sampling by sedimentation technique with Petri dishes containing BHI and SDA/not applicable/direct counting on plates (cfu/m3)	–	–	–	Significant reduction in airborne bacterial and fungal counts
Morris et al. [23]	In-situ experiment in a real-life scenario	COVID patients and procedures	Airborne SARS-CoV-2 and a range of other bacterial, viral, and fungal pathogens	Active sampling by filtration and qPCR assays/not informed/nucleic acids were extracted from each sampler component	–	–	–	Significant reduction of airborne SARS-CoV-2 and other airborne pathogens detected
Pouvaret et al. [30]	In-situ experiment in a real-life scenario	Patients and procedures	Airborne bacteria and fungi	Active sampling by impaction with SCA and standard agar plates/100 L/min for 5 min per point/direct counting on plates (cfu/m3)	2	35	–	Significant reduction in airborne bacterial and fungal counts
Anis et al. [32]	In-situ experiment in a real-life scenario	Patients and procedures	Airborne bacteria	Active sampling by impaction with blood agar plates/30 L/min for 10 min per point/direct counting on plates (cfu/m3)	–	–	–	Non-significant reduction in airborne bacterial counts
Bischoff et al. [33]	In-situ experiment in a real-life scenario	Patients and procedures	Airborne bacteria	Active sampling by impaction with blood agar plates (TSA II with SBA)/not informed/direct counting on plates (cfu/m3)	–	–	–	Significant reduction in airborne bacterial counts
Le et al. [35]	In-situ experiment in a real-life scenario	Patients and procedures	Bacteria and fungi	Passive sampling by sedimentation technique/not applicable/direct counting on plates (cfu/mL)	–	–	–	Significant reduction in airborne bacterial and fungal counts
Hallier et al. [37]	In-situ experiment in a real-life scenario	Dental AGPs	Bacteria	Active sampling by impaction with blood agar plates/100 L/min for 5 min per point/direct counting on plates (cfu/m3)	–	21–24	–	Significant reduction in airborne bacterial counts
Chotigawin et al. [38]	In-situ experiment in a real-life scenario	Patients and procedures	Bacteria and fungi	Active sampling by impaction with TSA and SDA plates/28.3 L/min for 3 min per point/direct counting on plates (cfu/m3)	–	–	74–76	Significant reduction in airborne bacterial counts but a non-significant reduction in airborne fungal counts
Pelleu et al. [39]	In-situ experiment in a real-life scenario	Dental AGPs	Bacteria and fungi	Active sampling by impaction with TSA plates/not informed/direct counting on plates (cfu/m3)	–	–	–	Significant reduction in airborne bacterial counts

AV, aspirated volume; T, temperature; RH, relative humidity; CO₂, carbon dioxide; TSA, tryptic soy agar; cfu, colony-forming unit; PAC, portable air cleaner; SBA, sheep blood agar; BHI, brain–heart infusion; SDA, Sabouraud–dextrose agar; qPCR, quantitative polymerase chain reaction; CDC, Centers for Disease Control and Prevention; SCA, Sabouraud–chloramphenicol agar; AGP, aerosol-generating procedure; HVA, high-volume aspiration.

In all the studies, the sources of aerosols were patients and procedures performed during the sampling period. Table II also shows the calibration of the measurement tools (when reported), airborne microbes assessed, analysis method, and outcomes reported using portable air cleaners in each study assessing microbiological contamination. Regarding indoor air parameters that can influence microbiological results, only four of these studies assessed temperature or relative humidity [19,30,37,38], and only one assessed the air changes per hour (ACH) in the room [30].

Assessment of airborne particle concentrations using portable air cleaners

The concentration of airborne particle or particle matter of different aerodynamic diameters was measured in 12 studies through in-situ experiments in real [17,18,21,22,26,[28], [29], [30],32] and quasi-real-life [[24], [25], [26], [27]] scenarios, alone or combined. Table III shows details of the measuring tools used (when reported), airborne particle sizes measured, and outcomes reported using portable air cleaners in each study assessing airborne particulate concentration.

Table III.

Summary of the methodologies, measurement tools, and outcomes assessing airborne particles

Study	Study design	Source of aerosols	Particle size (μm)/measuring tool	Indoor air parameters			Outcomes reported using portable air cleaners
				ACH	T (°C)	RH (%)
Capparè et al. [17]	In-situ experiment in a real-life scenario	Dental AGPs	0.3, 0.5, 1.0 and 5.0/particle counter system (Lasair III; Particle Measuring Systems, Boulder, CO, USA)	NI	–	–	Significant reduction in total airborne particle values
Oberst et al. [18]	In-situ experiment in a real-life scenario	Surgical procedures	2.5, 1.0 and 10/particle measuring device (IGERESS, Yuzhi Technology, Shenzen, China)	–	–	–	Significant reduction in airborne particle values for all sizes measured
Maurais et al. [21]	In-situ experiment in a real-life scenario	Dental AGPs and dental non-AGPs	≤10/laser photometers sensor (DustTrak DRX; Norrscope, Chelmsford, UK)	6 and 13	–	–	Significant reduction in airborne particle values in all conditions assessed
Tzoutzas et al. [22]	In-situ experiment in a real-life scenario	AGPDs	10 and 2.5/laser photometers sensor	7	23–26	30–60	Significant reduction in airborne particle values for the duration of the experimental period with a few exceptions
Buising et al. [24]	In-situ experiment in a quasi-real-life scenario	Glycerine-based aerosol (<1 μm)	≤2.5/laser photometers sensors (TSI DustTrak DRX 8533 and II8530)	–	–	–	Significant reduction in airborne particle values
Lee et al. [25]	A numerical experiment in a quasi-real-life scenario	Aqueous glycol solution (1.0 μm)	1.0 (predicted by the size of solution used) ACH calculating for predicting the clearance time	13.9	–	–	Airborne particle clearance time was significantly improved (3 times faster in <10 min)
Razavi et al. [26]	Numerical and in-situ experiment in a real and quasi-real-life scenario	Dental AGPs and simulated dental AGPs	0.3, 0.5, 1.0, 2.0, 5.0, and 10/aerodynamic particle sizer spectrometer and two optical particle counters for predicting the clearance time	7.23 and 14.73	22–25	49–60	Airborne particles clearance time was significantly improved (≥6.3 times faster)
Ren et al. [27]	In-situ experiment in a quasi-real-life scenario	Burning three sticks of incenses	0.3, 0.5 and 1.0/aerosol particle counter (Lasair III 310 C, USA)	3–45	22–23	34–52	Significant reduction in airborne particle accumulation and accelerated removal. Especially prominent in rooms with poor ventilation.
Verbeure et al. [28]	In-situ experiment in a real-life scenario	Patients undergoing an oesophageal HRM	0.3, 0.5, 1.0, 3.0, 5.0, and 10/particle counter (Lasair II, USA)	–	–	–	A non-significant reduction in airborne particle values
Messina et al. [29]	In-situ experiment in a real-life scenario	Surgical procedures	≥0.3, ≥0.5, ≥1.0, ≥3.0, ≥5.0 and >10/particle counter (Climet Ci-550; Climet Instruments Co., Redlands, CA, USA)	15	–	–	Significant reduction in airborne particle values of all sizes
Pouvaret et al. [30]	In-situ experiment in a real-life scenario	Patients and procedures	0.3, 0.5, 1 and 5/optical particle counter (AeroTrak; TSI, Minneapolis, MN, USA) following ISO 21501-4 standard	2	35	–	Significant reduction in airborne particle values (above the ISO 6a class compared with ISO 7a and ISO 8a classes reached with no PAC)
Anis et al. [32]	In-situ experiment in a real-life scenario	Surgical procedures	0.5–10.0/particle counter (BioTrak; TSI, Minneapolis, MN, USA)	–	–	–	Significant reduction in total airborne particle values

T, temperature; RH, relative humidity; CO₂, carbon dioxide; PAC, portable air cleaner; AGPs, aerosol generator dental procedure; NI, not informed; ACH, air changes per hour; HVAC, heating, ventilation, and air conditioning; HRM, high-resolution manometry; ISO, International Organization for Standardization.

^aISO 1 indicates the cleanest and ISO 9 the dirtiest air.

In one study, in addition to particle concentration levels, they also assessed the aerosol size distribution with an aerodynamic particle size spectrometer [26]. Only in two studies was the clearance time of the portable air cleaner calculated [25]. Seven studies calculated the pre-existing ACH in the rooms (Table III). Capparé et al. mention that the pre-existing ACH was calculated, but the data is missing in the results [17]. Only four studies assessed temperature and relative humidity [26,27,30,40].

Rate of infections or interventions using portable air cleaners

Five studies measured the correlation or association between the implementation of portable air cleaners and decreased rates of infections or intervention [16,19,34,36]. Three were non-randomized prospective studies [19], one was retrospective [36], and the other was the protocol of a multicentre randomized, double-blind, placebo-controlled trial [16]. This protocol is part of the EPoS trial1 that was conducted at seven hospitals from 2017 to 2022 to assess the implementation of air-cleaning devices and the incidence of surgical site infections in orthopaedic surgery. Description of the impacts measured, study design, follow-up, and outcomes reported by the studies are summarized in Table IV .

Table IV.

General characteristics of the included studies in which rates of infections or interventions were assessed

Study	The potential impact assessed	Study design	Follow-up	Outcomes reported using PACs
Persson et al. [16]	Decreased rate of surgical site infections in ORs of seven hospitals	Study protocol of a multicentre randomized, double-blind, placebo-controlled trial. They need ∼45,000 patients to attain a power of 80%	60 months	Results are not yet published
Arikan et al. [19]	Decreased rate of hospital-acquired infections in ICUs in a hospital	Non-randomized prospective study	8 months	Significant positive correlation with the decreased rate of hospital-acquired infections
Rao et al. [31]	Improvement of health outcomes for patients admitted with respiratory distress in the paediatric hospital setting	Non-randomized prospective study of 562 patients	3 months	Non-significant association with the decreased overall length of stay in the hospital and ICU, intubation, nebulizer, and non-invasive ventilation use. However, the authors reported that these reductions were clinically meaningful with a significant impact on the healthcare system.
Ozen et al. [34]	Decreased rate of infections in patients being treated for haematologic malignancies during construction near the hospital	Non-randomized prospective study	12 months	Significant association with decreased overall rates of infections. The preventive effect was more pronounced in patients with acute lymphocytic leukaemia, patients undergoing consolidation therapy, and patients with moderate neutropenia.
Abdulsalam et al. [36]	Decreased incidence rate of invasive aspergillosis infection in a hospital	Non-randomized retrospective study of 134 cases	31 months	Significant association with the decreased incidence rate of invasive aspergillosis infection

PAC, portable air cleaner; OR, operating room; EPoS, European Polyp Surveillance Trial; ICU, intensive care unit.

Discussion

Most of the 24 studies included in this review (71%) were conducted after the COVID-19 outbreak from 2020 to 2022. Only one study assessed SARS-CoV-2 RNA in air samples collected in a ward and an intensive care unit adapted for COVID-19 patients [23]. Findings showed a significant reduction in detectable SARS-CoV-2 RNA when the devices were operating [23].

In summary, 20 out of 24 studies demonstrated significant potential to prevent and mitigate the impact of bioaerosols in healthcare environments, regardless of the scenario and methodology used. These results were especially prominent in settings with poor ventilation, such as dental clinics, where economic issues or lack of guidelines may limit the installation of an appropriate ventilation system. According to the US Occupational Information Network (O∗NET), which calculates risk levels for different occupations, dentists and clinical dentistry professionals are at the top of the risk scale when comparing ‘exposure to disease and infection’ versus ‘physical proximity to other people’ [41]. In the remaining four studies (three in-situ studies and one prospective study) the reduction of aerosols was non-significant [28,31,32,38]. Across the nine different healthcare facilities of the included studies in this review, 28% of the studies that fulfilled our inclusion criteria were performed in dental clinics, followed by patients' wards (16%), operating rooms (16%), and intensive care units (12%).

This review has limitations. Variability in aerosol measurement remains a challenge. Active air samplers exhibit high variability, yielding different results in the exact location simultaneously. A calibration following validated standards is strictly necessary, but it was not mentioned in all studies in this review.

Although the ACH sums up all methods of aerosol removal – natural or mechanical (e.g. unknown leakage, settling, opening windows, HVAC system, etc.) – which could significantly impact the measured outcome, only seven studies calculated the pre-existing ACH in the rooms. Also, temperature, relative humidity, and air velocity directly influence aerosols' mechanics, but only seven studies controlled these parameters in their statistical analyses.

Methodologies and outcome measurements were not standardized in current research, compromising the overall quantitative measure of the magnitude of the effect. Although several studies assessed airborne microbial counts, the conclusions of these studies cannot be extrapolated to species that require specific growing conditions or different sampling requirements. Some species cannot be detected with conventional culture counting methods, requiring additional analysis that includes molecular identification methods.

Although the included studies assessed the outcomes using well-known methodologies, most of them lacked the complexity associated with analysing indoor air quality in healthcare settings. Standardization of methods is necessary to obtain a body of evidence with less heterogeneity, which would allow for establishing the size of the effect, making recommendations, direct comparisons, and cost/benefit analyses of the implementation of portable air cleaners in healthcare settings. Given the global economic pressure on clinical settings and the rapid evolution of practices to live with COVID-19, the device manufacturers should focus on efficiency and being affordable for their implementation in low- or medium-income countries.

Future research should assess (a) active airborne microbial sampling (at least overall fungi and bacteria) or quantitative PCR analysis, (b) airborne particle concentrations (<5 μm), and (c) indoor air parameters (mainly ACH, temperature, relative humidity, and air velocity) to be controlled in statistical analyses, including the flow of people and the procedures performed during sampling. Most importantly, studies need to evaluate the influence of portable air cleaners on rates of infections through prospective randomized or non-randomized trials with long-term follow-up and large sample sizes.

We recommend calculating the sample size for microbiological sampling, preferably based on a pilot study for assessing the variability of the setting since physical and biological variables may affect the aerosol mechanics or viability of micro-organisms. A full description of these technologies should be supplied – including their design, ease of use, noise level, maintenance cost, and the energy consumption – enabling us to compare the cost-effectiveness of the different devices tested.

Author contributions

M.A. designed the study and drafted the paper with input from all authors. M.A. and J.D. performed the searches and data extraction. M.A., J.D., and B.L. analysed methodologies. G.M. and A.G. revised the manuscript critically for important intellectual content and final approval of the published version.

Conflict of interest statement

None declared.

Funding statement

This study was supported by the Fundação Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES), Brazil, Finance Code 001; and did not receive any specific grant from other funding agencies in the public, commercial, or not-for-profit sector.

Appendix A. Supplementary data

The following is the Supplementary data to this article: