Article Text

Download PDFPDF

Original research
Bacterial contamination of endoscopist and assistant face visors during gastrointestinal endoscopy: a pilot study
  1. David Nylander1,
  2. John Leeds1,2,
  3. John Perry3,
  4. Manjusha Narayanan3,
  5. Chris Dipper1,
  6. Mel Gunn1,
  7. Manu Nayar1,
  8. Ally Speight1,
  9. Kofi W Oppong4,5
  1. 1 Department of Gastroenterology, Newcastle Upon Tyne Hospitals NHS Foundation Trust, Newcastle Upon Tyne, UK
  2. 2 Population Health Sciences Institute, Newcastle University, Newcastle upon Tyne, UK
  3. 3 Department of Microbiology, Newcastle Upon Tyne Hospitals NHS Foundation Trust, Newcastle Upon Tyne, UK
  4. 4 HPB Unit, Newcastle Upon Tyne Hospitals NHS Trust, Newcastle Upon Tyne, Newcastle upon Tyne, UK
  5. 5 Clinical and Translational Research Institute, University of Newcastle upon Tyne, Newcastle upon Tyne, UK
  1. Correspondence to Dr David Nylander, Gastroenterology, Newcastle Upon Tyne Hospitals NHS Foundation Trust, Newcastle Upon Tyne NE1 4LP, UK; david.nylander{at}


Background and aim During the COVID-19 pandemic, health workers’ facial exposure to pathogens has been brought into focus. In this study, we aimed to determine the occurrence and degree of facial contamination to both endoscopists and their assistants during endoscopic procedures to help inform future safety measures.

Methods Non-sterile visors worn by endoscopist, assistant and room control visors from 50 procedures were swabbed post procedure for culture. Procedure type, therapy, duration and evidence of visible visor contamination were recorded. After 48-hour incubation, all bacterial colonies were identified using matrix-assisted laser desorption/ionisation time-of-flight mass spectrometry. Organisms were classified into skin/environmental, oronasal and enteric.

Results A total of 104 visors were available for assessment (93 staff; 11 control). In worn visors, skin/environmental flora were isolated from 70, oronasal flora from 8, and enteric flora from 3 with an average colony count of >9.5. Notably, bacteria of enteric origin (Escherichia coli and Enterobacter cloacae) were isolated from three worn visors. In room control, skin/environmental flora were isolated from seven and oronasal flora from one with average colony count of five. No room control visors grew enteric flora. Overall, 9.1% room control and 10.8% worn visors were contaminated with organisms that could possibly have originated from patients. However, enteric flora were only obtained from worn visors. No visors were visibly contaminated.

Conclusion This pilot study demonstrates risk of contamination to faces of endoscopists and assistants. Larger studies are required to determine degree of risk and to give guidance on facial protection during gastrointestinal endoscopy.

  • diagnostic and therapeutic endoscopy
  • intestinal bacteria
  • gastrointesinal endoscopy

Data availability statement

All data relevant to the study are included in the article or uploaded as supplementary information. All data used in study are uploaded.

This article is made freely available for personal use in accordance with BMJ’s website terms and conditions for the duration of the covid-19 pandemic or until otherwise determined by BMJ. You may use, download and print the article for any lawful, non-commercial purpose (including text and data mining) provided that all copyright notices and trade marks are retained.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


  • The facial mucous membranes of endoscopist is at risk of contamination during gastrointestinal (GI) endoscopy procedures. Although there has only been indirect evidence, a recent study has shown increased colony count on the face visor of endoscopists at the end of a session of multiple GI endoscopy procedures.


  • This study is the first to investigate facial bacterial contamination in GI endoscopists and their assistants after individual procedures in which the organisms have been typed in an attempt to clarify their origin. The study has demonstrated the presence of oronasal and enteric flora in a proportion of endoscopists and their assistants.


  • This study demonstrates a risk of contamination to GI endoscopists and their assistants’ faces during procedures and provides additional evidence to help in formulation of guidelines to improve their safety during procedures. However, larger studies which include assessment of mechanisms of contamination are required to give more robust recommendations.


The SARS-CoV-2 pandemic brought into sharp focus transmission of pathogens within hospitals especially those spread via aerosols and by contact with contaminated surfaces.1 In endoscopy units, there have been reports of infection outbreaks from endoscope contamination.2 Guidelines now exist on automated re-processor decontamination. These policies have protected patients by preventing micro-organism transmission by endoscopes,3 4 which can become contaminated with secretions from the gastrointestinal (GI) tract and oropharynx.

Prior to the pandemic, there was little international consensus and published guidelines on use of personal protective equipment (PPE) to protect health workers performing GI endoscopy procedures from organisms carried by the patient. This is pertinent especially with the concerns raised by transmission of organisms like SARS-CoV-2. It is recognised that during GI endoscopy, there is risk to the endoscopist (including the face and eyes) from direct splash and droplets. Also, conjunctivitis and systemic infection can occur by the endoscopist touching their eyes with contaminated fingers.5 Only the American Society of Gastrointestinal Endoscopy gives recommendations for facial protection. That guidance states that because of the potential for splash exposure to the face, units should develop policies for wearing face and/or eye shields or masks.6

In a survey of endoscopists at three facilities in Pittsburgh, USA, prior to the pandemic, 16.7% of ‘attendings’ would not consider and 66.7% would consider using facial protection.7

In our institution, prior to this pandemic, it was not a standard practice for endoscopists or assistants to use facial protection when performing GI endoscopy procedures in patients who do not have highly transmissible serious pathogens such as HIV and Mycobacterium tuberculosis.

Our aim was to do a pilot study to estimate exposure to and degree of bacterial contamination to endoscopists and assistants’ faces by assessing bacterial contamination of facial visors worn during GI endoscopy procedures.


The study was carried out at Newcastle Hospitals NHS Trust, UK. Seven endoscopists participated and procedures reflected the spectrum performed during the study period: 3 June–7 August 2020 (in the early ‘recovery phase’ after suspension of endoscopy services). During this period, procedure numbers were significantly restricted. Facial visors were used as part of enhanced PPE for staff: face visors and surgical masks for all procedures plus aprons for lower GI procedures and full-length water-resistant gowns for upper GI procedures. Records were kept of the personnel who wore the visors and their role (ie, endoscopist or assistant), procedure date, type and duration and visible contamination of the visor during the procedure. The duration of each procedure was defined as the time between endoscope insertion and end of withdrawal. Those procedures in which there was repeated instrumentation via the biopsy channel with repeated piercing of the biopsy ‘bung’ were defined as therapeutic procedures (eg, endoscopic retrograde cholangiopancreatography (ERCP) and sphincterotomy/stent, endoscopic mucosal resection, polypectomy and fine needle aspiration during endoscopic ultrasound (EUS)). This excluded cases in which only biopsies were taken. Each procedure was given a study number to match procedure to culture swabs.

Endoscopist/assistant visor swabs

One new non-sterile full length visor (Citisen face shield; Synectics Medical, Enfield, UK) was used for each procedure by each of the endoscopists and assistants. This covered the entire face (figure 1). The plastic covering the outer surface was removed from the visor when patients entered the room. Visors were donned by endoscopists and assistants ensuring that the outer surface of the visor was not touched. Swabs were not taken and visor discarded if:

  1. The outer surface was touched by the endoscopist/assistant or any object.

  2. There was a change of endoscopist or assistant during a long procedure (eg, switch from trainee to trainer).

Figure 1

Full face shield worn by endoscopists and assistants.

After each procedure, visors of the endoscopist and assistant were swabbed in a standardised manner using full aseptic technique (vertical strokes, left to right to sample entire visor surface) as shown in figure 2.

Figure 2

Visor and swab sampling pattern.

Room control visor swabs

Room control swabs were randomly taken as follows. The plastic covering was removed from the visor which was then hung 2 m high on a drip stand. This was then placed approximately 2 m behind the endoscopist. The visor was then swabbed in the manner described above at the same time as the endoscopist/assistant visors.

Position of endoscopists/assistants

All procedures were performed in the standard manner. Patients had topical throat anaesthetic or sedation (midazolam±opioid) depending on preference. No patients had general anaesthesia or propofol. Endoscopists maintained their usual position with bed at usual height during procedures. For upper GI procedures (gastroscopy, ERCP and pancreaticobiliary EUS), the assistant is defined as the person ‘at the head’ of the patient ensuring that airway was clear and giving reassurance. For lower GI procedures (colonoscopy and flexible sigmoidoscopy), the assistant was person closest to the point of insertion of the scope, assisting with biopsies and therapy. (see figure 3).

Figure 3

Relative positions of personnel in rooms for oral and anal procedures E=endoscopist; A=assistant; p=patient

Sampling methods

Swab sampling was performed using sterile polywipe; (Medical Wire and Equipment, Corsham, UK). These are sterile sponges (10 cm × 5 cm) moistened with phosphate buffer designed for microbiological sampling. We sampled visors only after procedures (rather than before and after) to avoid any effect of the phosphate buffer deposit on bacterial adhesion to visors. After sampling, each polywipe was sealed in a sterile bag and transported to the laboratory for processing the same day. Culture was performed by aseptically removing the polywipe using sterile forceps and imprinting the surface used for sampling onto a Columbia blood agar (CBA) plate. This was achieved by pressing the entire surface of the polywipe onto the surface of the agar plate 10 times using sterile forceps. After inoculation of CBA, the process was repeated (using the same polywipe) to inoculate a second culture plate containing fastidious anaerobe agar (FAA). Both media were prepared in house from ingredients obtained from Oxoid, Basingstoke, UK. CBA was incubated in air at 37°C for 48 hours. FAA was incubated in an anaerobic chamber at 37°C for 48 hours.

After incubation every different colony type (assessed by size, shape, colour and haemolysis) was characterised to species level by matrix-assisted laser desorption/ionisation time-of-flight mass spectrometry using a Bruker Biotyper (Bruker, Coventry, UK). A colony count was recorded for each species within a particular sample. Organisms were broadly classified into skin/environmental and oronasal and enteric.

Statistical analysis

Continuous variables are reported as mean and SD if normally distributed and the median and IQR otherwise. Comparisons of continuous variables between groups were made using analysis of variance if the values followed the normal distribution, and the Kruskal-Wallis test otherwise. Categorical data were compared between groups using the χ2 test. MedCalc Statistical Software V.20 (MedCalc Software, Ostend, Belgium;; 2021) was used.


Swab samples were obtained from 50 procedures: 19 gastroscopy, 11 ERCP, 10 colonoscopy, 5 flexible sigmoidoscopy and 5 oral EUS. A total of 93 clinician (endoscopist and assistant) visors were sampled. Seven swabs were excluded from three ERCPs and four colonoscopies. The reasons for this included a trainer taking over from a trainee (n=1), an assistant removed the visor incorrectly (n=1) and a change of assistant (n=5). Eleven random room control visor swabs were taken.

Details of the procedures and personnel facial swab samples, together with procedure duration, are presented in table 1. None of the visors had visible contamination.

Table 1

Procedure details and number of swabs obtained

In the worn visors, there was no growth on 18, skin/environmental flora grown on 70 and oral/enteric flora on 10. Some visors had organisms from more than one suspected origin. The organisms cultured and category allocated were as presented in table 2. Organisms from more than one category were grown from some visors.

Table 2

Full classification of bacterial species isolated from visors

Of the 11 room control visors, there was no growth on 4, growth of skin/environmental flora on 6 and potential oral flora (Streptococcus salivarius) on 1. There was no significant difference in the proportions of growth between the worn visors and the environmental visors (p=0.28). There was also no significant difference in the proportions of contamination between the endoscopists’ visors compared with the assistants’ visors for either any growth (p=0.31) or oral/faecal bacteria (p=0.64).

The median total colony counts were 1.0 (IQR: 1.0–3.0) for the endoscopist visors, 1.0 (IQR: 1.0–4.0) for assistant visors and 2.0 (IQR: 0.75–5.25) for the room control visors (p=0.39). This distribution is shown in figure 4.

Details of procedures and personnel wearing visors are presented in table 3. There were incidents of colony counts of >100 in 2 ERCP and 2 oral EUS cases (see figure 5). Enteric organisms were isolated in three visors, one from a colonoscopist and one each in ERCP and flexible sigmoidoscopy assistants. There were no significant differences in median procedure time for visors with no growth (10.0 min, IQR: 6.0–26.0 min), skin/environmental bacteria (13.0 min, IQR: 9.75–29.0 min) or oral/enteric bacteria (15.0 min, IQR: 10.0–50.0 min) (p=0.23). There were no significant differences in overall visor contamination rates between upper or lower GI procedures (upper: 47/67 vs lower: 18/26, p=1.0).

Figure 5

Culture plates: left plate shows light contamination of skin flora and dish on the right shows heavy faecal flora growth in ERCP assistant visor. ERCP, endoscopic retrograde cholangiopancreatography.

Table 3

Details of procedures in which enteric or oronasal bacteria were isolated

However, therapeutic procedures (7/18) were significantly more likely to have visors contaminated with oral/faecal bacteria than diagnostic procedure (3/32; p=0.024).


There have been suggestions that endoscopists are at minimal risk of facial contamination performing GI endoscopy procedures, so face shields are not required.8 A study comparing contamination of endoscopist face shields during upper GI endoscopy in patients in left lateral versus prone position showed greater contamination of face shields in the left lateral position, but contamination in the face shields was deemed not clinically meaningful.9

Our study is the first to investigate potential contamination of the faces of both the endoscopist and assistant during individual endoscopy procedures. We have demonstrated facial bacterial contamination of visors of both the endoscopist and assistant despite no visible visor contamination in a broad range of procedures. Visor contamination appeared to be independent of procedure type or duration although the numbers are small. Enteric organism contamination occurred in a small number of procedures all of which involved biopsy channel instrumentation.

Johnson et al investigated facial contamination of endoscopists’ facial visors7 by taking swabs from the visors of the endoscopists before and after a 4–5 hour morning session containing different procedures. In that study, they used a count of greater than 15 colony forming units as evidence of exposure but did not identify which organisms had led to visor contamination. In their study, 5.3% of post-endoscopy facial swabs of endoscopists demonstrated bacterial contamination.

We did not swab visors before and after each procedure to avoid any impact of the swab phosphate buffer deposit on visor contamination.

Visor contamination can happen by a number of mechanisms (not investigated in this study).

One possibility is direct aerosolisation of droplets to the face of the endoscopist or assistant. Data as to whether GI endoscopy is aerosol generating are conflicting. In a proof-of-concept study, upper GI endoscopy led to an increase in particle count above that seen in the baseline count immediately prior to the procedure.10 Similarly, Sagami et al showed an increase in aerosolised particles during endoscopy when compared with controls using a novel box around head of the patient.11 There have been other studies investigating transmission of SARS-CoV-2 during procedures and only endotracheal intubation was consistently associated with this.12

The second possible mechanism is by splash contamination. Blood and bodily fluid splashes to the face and body of health workers have been demonstrated in open laparoscopic and endosurgical procedures.13–15 Splash contamination is also recognised in GI endoscopy16 even though it has been less well studied. The association of increased risk of especially enteric organism contamination with procedures in which there is instrumentation suggests that the mechanism may be related to instrumentation through a biopsy channel ‘bung’ where an aerosol might be generated. In a recent study,17 a model was created to quantify droplets emerging from biopsy channel in vitro and during GI endoscopy. The study concluded that the biopsy channel may be a source of infectious droplets especially on removal of instruments from it.

It is also of interest to note that in our study, the single room control visor contamination happened in a procedure where there were multiple biopsies taken. For diagnostic procedures, the biopsy bung is often not pierced. It possibly acts as a guard to aerosol escape from the biopsy channel, so there is no risk of splash contamination that can happen on withdrawal of instruments.

This pilot confirms the risk of contamination to the face of GI endoscopists and assistants during upper and lower GI endoscopy. In the absence of a visor, there is risk of pathogens reaching the facial mucous membrane and potentially causing transmission of infectious agents from the patients.

There are some limitations to this study. First, only bacterial cultures were performed as we were exploring the possibility of contamination to the faces of endoscopy healthcare workers from the patient. Such contamination to the faces of both endoscopists and assistants had not previously been studied for single procedures. This study was exploratory and bacterial contamination was considered to be a good surrogate for all micro-organisms. Also, importantly in this study, growth on visors underwent full classification and species identification to give suggestion of origin. Second, no cultures were taken directly from the patients or endoscopy staff to confirm the origin of the bacterial strains. However, as swabs were taken only from the outer surface of visors of healthcare workers who also wore oronasal masks, we assumed the oronasal or enteric organisms grown were most likely from the patient. Third, no air samples were taken which would have been useful in determining the mechanism of contamination between aerosol and fluid splash. No visors were visibly contaminated during the study, however. It should be noted that all endoscopy rooms used conformed to the HTM 03–01 (specialised ventilation for healthcare buildings) 2020 standards and as per those standards, all endoscopy rooms were at 10 air changes per hour (ACH) and at positive pressure to the outer corridors.

Finally, only the endoscopist, assistant and a control visor were studied, limiting the understanding of distance and direction of contamination. It would be useful in future studies to determine whether contamination occurs evenly or whether certain areas around the patient are at higher or lower risk. It would also be useful to understand whether there are differences in rates and degree of contamination between bacteria and viruses. In line with the ‘green endoscopy agenda’, future studies should probably use reusable visors rather than the single use visors used in this case.

In conclusion, this study shows a significant rate of visor contamination to the endoscopist and assistant of around 10.8% with organisms of oronasal and faecal origin. This suggests that GI endoscopists should consider the routine use of facial protection during GI endoscopy procedures. Further studies are needed to determine the mechanisms by which this occurs. More detailed investigations are required before such protection can be incorporated into infection control guidelines.

Data availability statement

All data relevant to the study are included in the article or uploaded as supplementary information. All data used in study are uploaded.

Ethics statements

Patient consent for publication

Ethics approval

This study involves human participants. This study was approved by Caldicott guardian, Newcastle upon Tyne Hospitals (20 May 2020; ID: 7581). As per United Kingdom National Health Service research ethics guidance, ethical approval from an institutional review body was not required for this study. Institutional authorisation to hold a prospective patient database for use for service evaluation was obtained. Procedure-specific written informed consent was obtained from all patients prior to their procedure exempted this study. Participants gave informed consent to participate in the study before taking part. Study registration in Hospital clinical effectiveness register: 19 May 2020 (Nno.: 10100).Approval by Caldicott guardian: 20 May 2020 (ID: 7581).


We acknowledge the endoscopy nursing staff of Newcastle Hospitals who carried out the aseptic swabbing of visors.



  • Twitter @dlny50, @drallyspeightccc, @KofiWOppong

  • Contributors DN came up with concept of study and wrote initial proposal and was endoscopist and lead in writing the manuscript and is the guarantor. DN, KWO and JL designed the study. MNarayanan and JP planned and carried out microbiological and bacteriological analysis and report and also provided critical feedback in bacterial results reporting on manuscript. JL carried out ad wrote statistical analysis of the results and wrote that section of the manuscript. KWO, JL, MG, CD, AS and MNayar all carried out endoscopy procedures and provided critical feedback in data analysis and the writing of the manuscript and were involved in finalising the manuscript for submission.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.