A scoping review on bio-aerosols in healthcare and the dental environment

Background Bio-aerosols originate from different sources and their potentially pathogenic nature may form a hazard to healthcare workers and patients. So far no extensive review on existing evidence regarding bio-aerosols is available. Objectives This study aimed to review evidence on bio-aerosols in healthcare and the dental setting. The objectives were 1) What are the sources that generate bio-aerosols?; 2) What is the microbial load and composition of bio-aerosols and how were they measured?; and 3) What is the hazard posed by pathogenic micro-organisms transported via the aerosol route of transmission? Methods Systematic scoping review design. Searched in PubMed and EMBASE from inception to 09-03-2016. References were screened and selected based on abstract and full text according to eligibility criteria. Full text articles were assessed for inclusion and summarized. The results are presented in three separate objectives and summarized for an overview of evidence. Results The search yielded 5,823 studies, of which 62 were included. Dental hand pieces were found to generate aerosols in the dental settings. Another 30 sources from human activities, interventions and daily cleaning performances in the hospital also generate aerosols. Fifty-five bacterial species, 45 fungi genera and ten viruses were identified in a hospital setting and 16 bacterial and 23 fungal species in the dental environment. Patients with certain risk factors had a higher chance to acquire Legionella in hospitals. Such infections can lead to irreversible septic shock and death. Only a few studies found that bio-aerosol generating procedures resulted in transmission of infectious diseases or allergic reactions. Conclusion Bio-aerosols are generated via multiple sources such as different interventions, instruments and human activity. Bio-aerosols compositions reported are heterogeneous in their microbiological composition dependent on the setting and methodology. Legionella species were found to be a bio-aerosol dependent hazard to elderly and patients with respiratory complaints. But all aerosols can be can be hazardous to both patients and healthcare workers.

a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 present body of evidence regarding bio-aerosols. This results in an up-to-date summary of the literature, allowing us to make recommendations for future research by identifying gaps in current knowledge, and to underline the risks for HCW and immunocompromised. Since this is a scoping review, our objectives are broad and cover three areas concerning bio-aerosols in hospital and dental settings [18,19]: • What are the sources that generate bio-aerosols?
• What is the microbial load and composition of bio-aerosols and how were they measured?
• What is the hazard posed by pathogenic micro-organisms transported via the aerosol route of transmission?

Design and search strategy
A scoping review was performed systematically according to the PRISMA statement for transparent reporting of systematic reviews and meta-analysis [20] and JBI Briggs Reviewers Manual [21] (see S1 PRISMA checklist

Screening process and inclusion criteria
References yielded from the search strategy were imported in Covidence, an online web application for screening systematic reviews, and duplicates were removed. C.Z. and A.L. screened and scored the relevance of the hits independently, based on their title and abstract. The full text manuscripts were retrieved via Endnote, Google, Research Gate or by addressing the corresponding and/or first author. Subsequently, the studies were assessed on their eligibility for inclusion based on the full text. A study was included for final data extraction and summary when it met one of the following criteria: bio-aerosol composition; pathogenicity; sources; conducted in healthcare or the dental setting; published in English, German, French, Spanish or Dutch. Discussion papers, letters to the editor, animal studies, protocols, prevention of bioaerosols, technical studies, reviews without pooled data, narrative reviews, development of drug therapy, or studies conducted in other settings besides healthcare were excluded. Additionally, a reference check and search through grey literature was conducted and included in the flowchart termed 'snowballing'.

Data extraction and summary
Data on the origin of bio-aerosols was categorized based on sources. Studies on the microbial composition of the bio-aerosols were summarized based on the colony forming units (CFU). References that reported sampling time were recalculated for a sampling time of 10 minutes and finally Log-transformed to make comparison possible between studies. These studies are presented in figures. References not reporting sampling time were not summarized and are presented in the study of characteristics table. The micro-organisms reported in individual studies were summarized per type of organism and setting. Potential hazard for patients and HCWs were summarized narratively.

Results
A total of 5823 studies were retrieved, of which 678 duplicates and 4797 irrelevant studies were removed. After reading 311 abstracts, 201 full text studies were assessed for eligibility. This eventually resulted in 62 studies including references from snowballing (see Fig 1. PRISMA flowchart).

Generation of bio-aerosols
One study reported solely on the generation of bio-aerosols [22]. Therefore, we extracted data on the generation of bio-aerosols from papers selected for the other objectives . The sources of bio-aerosols in dental clinics were: ultrasonic scalers, high speed hand pieces, air turbines, three in one syringes, and air water syringes. Studies conducted in hospitals reported 30 different bio-aerosol generating sources. Humans produced aerosols by coughing, and sneezing. Patients with cystic fibrosis positive for Burkoderia cepacia were also capable of producing pathogenic aerosols. Interventions conducted by HCWs that produced aerosols were: colonoscopy, tracheal intubation, suction before and after intubation, manipulation oxygen mask, bronchoscopy, non-invasive ventilation, insertion of nasogastric tube, defibrillation, chest physiotherapy, and washing the patient. Bed making, ward rounds, tea trolley round, activity at bed, floor mopping, moving furniture, lunch time, drugs round, evening meal, vacuum cleaner, toilet use, cold-mist humidifier, shower, cleaning patients room and the nebulizer were found to be other activities in a hospital to produce aerosols [22].

Hospital environment
Thirty-one studies analyzed the microbial composition of bio-aerosols in the hospital environment [11, 30, 35-37, 39, 41-65]. The studies combined identified a total of 111 organisms by using culture techniques (see Table 1 for overview of micro-organisms identified and Table 2 study characteristics hospital setting). Fifty-six bacterial species (23 Gram-negative and 32 Gram-positive; 1 mycobacteria), 45 fungal genera and ten viral species were identified [11, 30, 35-37, 39, 41-52, 54, 56-66] Most bacteria originated from human skin or the human gut, the environment or water. The identified viruses originated from the human respiratory tract. The methods for collecting air samples from the bio-aerosols and the methods for culturing microorganisms were heterogeneous. The method most frequently used to actively collect microorganisms was the Andersen air sampler (N = 9). Four studies used passive collection of micro-organisms by placing Petri dishes with agar. In all studies, 21 different culture methods were used, wherefrom Tryptic soy agar (N = 7) was most frequently used. Fourteen studies analyzed the bacterial load of the bio-aerosols [11, 39, 41, 42, 45-47, 50, 55, 58, 60, 61, 64, 65]. The mean Log-10 of CFU/m 3 ranged from 0.8 to 3.8 (see Fig 2). Additionally, five studies analyzed the bio-aerosol contamination before and/or after treatment, intervention or of a room when a patient with an infectious disease was present. The measured bacterial or fungal load ranged from Log 0.6-4.2 at baseline to Log 1.2-4.3 after the second measurement (see Fig 3) [30,35,43,56,57]. Seven studies reported on the fungal load in bioaerosols during the day when patients were present in a hospital room. Fungal loads ranged from Log 0.8-3.5 CFU/m 3 in various hospital wards [45,47,50,56,59,61,66]. Multiple studies quantified the air in patient specific areas or via specific methods.
Two studies identified multiple viruses in bio-aerosols after patients with symptoms of a cold coughed, however both studies did not report on the viral load [51,62]. Viral loads in the bio-aerosol ranged between Log 2.2 plaque forming units /m 3 in the air of an infant nursery positive for RSV and Log 5.5 PFU/m 3 in the air contaminated by patients positive for Influenza A virus [52,54]. Another study reported the RNA copy/L and found Log 3.3-5.2 in aerosols produced by patients positive for Influenza A virus [33].

Dental environment
Seventeen studies analyzed the microbial composition of dental clinics [23][24][25][26][27][28][29][67][68][69][70][71][72][73][74][75][76]. The studies cumulatively identified 38 types of micro-organisms by using culture techniques (see Table 3 for complete overview of micro-organisms identified and Table 4 for study characteristics in dental setting). Wherefrom nineteen bacteria (7 Gram-negative and 12 Gram-positive) and 23 fungal genera were detected. The bacteria originated from water, human skin and the oral cavity. None of the included studies looked for viruses or parasites. Similar to the hospital setting, the active Andersen air sampler (N = 4) and the passive culturing method by placing Petri dishes with agar (N = 6) were the most frequent used air sampling techniques. Thirteen different culture methods were used to identify the collected micro-organisms, of which Tryptic soy agar (N = 3) and blood agar (N = 3) were used most often.
The mean bacterial load in the bio-aerosols ranged from Log 1-3.9 CFU/m 3 (see Fig 4). Furthermore, six studies analyzed the bio-aerosol contamination before and after treatment. The bacterial or fungal load ranging from Log -0.7-2.4 CFU/m 3 at baseline and from Log 1-3.1 CFU/m 3 after treatment (see Fig 5) [25,68,71,72]. Only one study reported on the relation between the distance from the bio-aerosol generating source and the bacterial load. They found a higher bacterial load in the bio-aerosols at 1.5 meter from the oral cavity of the patient than in the bio-aerosols within 1 meter from the patient [26]. One study screened for B. cepacia and one screened for M. tuberculosis, however both studies could not retrieve these organisms after regular patient treatment [24,28]. Sampling method: Cough bio-aerosol sampling system method used to collect bio-aerosols in TB positive patients. Andersen six-stage sampler with 7H11 agar collection for 5 minutes. (Continued)

Hazard of a bio-aerosol
Seven studies reported on the hazard of micro-organisms to HCWs and/or patients, see Table 5 study characteristics hazard in healthcare and the dental setting [12,31,34,45,[77][78][79] Three studies looked into the risks for patients when exposed to Legionella pneumophila containing sources that may produce bio-aerosols [34,38,78]. They reported that cooling towers, air conditioning or mechanical ventilation systems could be a source of L. pneumophila. Kool et al. concluded that patients had an increased risk to acquire L. pneumophila in hospitals when they used corticosteroids (OR = 13; 95CI% 1.6-102) and when intubated (OR = 10; 95% CI 1.3-73) [34]. Another study identified smoking, drinking alcohol, having chronic lung disease and cancer as risk factors for getting an infection with L. pneumophila [78]. For the dental clinic there is one case study reported that reported of irreversible septic shock and died after two days in a patient that was infected with L. pneumophila [79].
One systematic review reported on the pooled odds ratio (OR) for the transmission and exposure to Severe Acute Respiratory Syndrome (SARS) in HCWs during bio-aerosol generating procedures. Tracheal intubation (OR = 6.6; 95%CI 2.3-18.9) and noninvasive ventilation   A scoping review on bio-aerosols in healthcare and the dental environment (OR = 3.1; 95%CI 1.4-6.8) were risk factors for acquiring SARS. Other bio-aerosol generating interventions such as manipulation of an oxygen mask were not significant risk factors [31]. Another study calculated that the risk ratio for acquiring clinical respiratory infections was 2.5 (95%CI 1.3-6.5) for HCWs performing a high risk procedure [12]. Augustowska et al. studied the effect of bacteria and fungi on asthmatic patients. They reported a decrease in maximum breathing capacity due to the increase of bacterial or fungal load in the air [45]. A case-study in a dental clinic described the risk of acquiring Herpes Simplex Virus (HSV)-1 for the dentist and the dental hygienists when they treated a patient with active HSV-1. One member of the treatment team became infected with HSV-1, probably by the bio-aerosol containing HSV-1, induced by ultrasonic scaling or by rubbing her eyes while working. The infected HCWs manifested recurrent HSV-1 infections [77].

Discussion
By conducting a scoping review we were able to summarize existing evidence on the generation, composition, load and hazards of bio-aerosols in hospital and dental environment. We found that bio-aerosols are generated via multiple sources such as machines, different types of interventions; instruments; and human activity. The composition of bio-aerosols depended on the method of sampling (active versus passive), microbiological techniques (culture based versus DNA-based, different culture plates used) and the setting of the study (specific clinics versus general dental clinics). Bio-aerosols can be hazardous to both patients and HCWs. Multiple studies described the threat of Legionella species to elderly and patients with respiratory complaints.
The composition of bio-aerosols was extensively studied in hospital environments (N = 31) compared to dental environments (N = 16). Regarding the micro-organism composition of bio-aerosols, we conclude that bio-aerosols contain a high variety of bacterial and fungal  (75) 1. Closed dental operatory using ultrasonic scaler. After: 44 (14) 2. Operative treatments in closed dental operatory.  After: 20 (11.5)

Pasquarella, 2012
Setting: Dental clinic Total mean bacteria CFU/m 3 before and after treatment: Sampling method: Passive air sample placing Tryptone Soya Agar for 1 hour, 1 m above the floor. Active air sampling using SAS sampler, 180L/min volume 500 L.

Szymanska, 2008
Setting: Dental clinic Total bacteria % before disinfection: Sampling method: Passive air sampling by placing Petri dishes during ultrasonic scaling. Baseline measure for 10min exposed dishes in middle of operatory. 2 plates on tray table over patient's chest 40cm away from mouth. 2 plates on cart 150cm from patient's mouth next to wall, behind patient and dentist at 100cm height exposed for 20min. BHI agar.
Before strains from different sources such as the human skin and intestine; and the environment such as soil and water. Based on the sampling and culturing techniques, fungi and Gram-positive bacteria were found most often. Pathogens such as Legionella and Pseudomonas species were found in bio-aerosols that were distributed by instruments using tap water. Few studies looked  Cross-sectional study with microbiological examination of the air and the lung function in asthmatic patients.
-Decrease of forced expiratory volume in 1 second.

Browning 2012
Awareness of the risks involved in treating active herpes labialis in a dental clinic for viruses, and in total only ten different viruses were identified, because no open ended detection or identification methods are available for viruses. Therefore only specific targeted techniques were used. Moreover, none of the studies conducted in dental practice have used methods to identify the presence of viruses in the generated aerosols. Therefore, we must keep in mind that the yielded results were dependent on the methodology of the individual study. The results of the individual studies, and the heterogeneity we found in this review, are dependent on the methods leading to an over-or under estimation of the complete bio-aerosol profile. The same inconsistency is discussed in previous studies in which the researchers compared two main sampling methods [80]. The methodological variety between studies, e.g. differences in method of sampling and culturing or sequencing, differences in sampling time and sampling area; and differences in distance to the bio-aerosol generating source caused difficulties in comparing results. When a study used selective medium or agar it results in an overview of selected micro-organisms. This leaves out other micro-organisms that were present at that moment. The same accounts for duration of sampling or passive versus active sampling. In passive sampling, the researcher waits for a certain amount of time for micro-organisms to fall on a Petri dish, while other micro-organisms were still floating in the air and take more time to fall on surfaces. The spread in a bio-aerosol is heterogeneous, so whatever is 'catched' on that moment may vary from the second, third or even fourth sampling attempt. So, the method chosen (active or passive) should be dependent on the aim of the air quantification [80]. Furthermore, in many studies variables in the experimental setup were not described, like sampling time, distance and sampling location. Also, no standard deviations of the microbiological loads were reported consistently. Besides, the data might be an underestimation of reality since studies looked for specific micro-organisms, in specific settings by selective sampling and culture dependent techniques, thereby missing other micro-organisms present in the bio-aerosols. Also, there was very little data available on the persistence of micro-organisms in the air over time and the spatial distribution. None of the included studies looked for -Suction before intubation: 3.5 (0.5-24.6) NS

Tran 2012
Risk of transmission of acute respiratory infections to HCWs exposed to bio-aerosol generating procedures parasites, although it has been reported that these are present in many tap water dependent bio-aerosol producing systems with plastic tubing [81,82]. We found little evidence to state the presence or absence of direct threats or health risks for patients or HCWs with regards to bio-aerosols. In the hospital setting, two studies reported on the hazard for the staff [12,31], and four on the hazard for patients [34,45,78,79]. The search yielded one study for this objective assessing the hazard of an infectious disease to dental staff [77]. However, it is known that on average, dental practitioners carry elevated levels of Legionella antibodies [83], but the hazard to non-healthy HCWs and patients remains, based on our findings, unknown. An estimation of the hazard of bio-aerosols is usually made based on the microbial content and load of the bio-aerosols. We conclude that bio-aerosols can be hazardous to certain populations that are extensively exposed to bio-aerosol generating procedures or immunocompromised people.

Limitations
The search yielded 40 references that were to be screen based on full text. However, we could not recover these 40 full text manuscripts to assess the their eligibility for inclusion, even though we tried to contact the first and/or corresponding author, by retrieving his/her email via the abstract or Google. We assume that the body of evidence for each objective would have been larger if all 40 studies, or at least a part, would have been available and included. Another limitation was that the outcomes and methods were inconsistent throughout all included studies. Also, there was little data on the hazard of bio-aerosols, thus no strong conclusions could be drawn.

Recommendation for future research
We recommend that future research on bio-aerosols should create an extensive and complete methodology for the quantification of air contamination. Time and frequency of air sampling, distance from sources, location of sampling in both passive and active air sampling techniques should be described thoroughly. Furthermore, the identification of micro-organisms should be done by both selective and non-selective methods and cover organisms that find their origin in water, human, and environment. Also, we believe that infections due to a bio-aerosol should be structurally reported so that the risk for HCWs and patient can be analyzed. Finally, the risks for HCWs, especially dentists, working in an environment in which they are continuously exposed to bio-aerosols, and to their patients remain unclear and therefore need further research. This is needed in order to comprehend the risks of bio-aerosols generated in clinical settings to attention to staff and patients to improve awareness, hygienic standards, risks, and prevention methods.

Conclusion
Bio-aerosols are generated via multiple sources such as different interventions, instruments and human activity. Bio-aerosols have different microbiological profiles depending on the setting and the used methodology. Bio-aerosols can be hazardous to both patients and healthcare workers. Legionella species were found to be a bio-aerosol dependent hazard to elderly and patients with respiratory complaints.