Andonian 2019.

Study characteristics
Methods	Study design: RCT Study grouping: parallel group For simulation study, what was used for the exposure (virus, fluorescent fluid etc): 1. The optimised fluorescent slurry consisted of fluorescent powder (Glitter Bug, Brevis Corporation, Salt Lake City, UT; 75 mg/mL) in a viscous suspension of grape‐seed oil and water (1:6 oil‐to‐water ratio) 2. Fluorescent 2‐μm polystyrene latex bead (PLSs) (G0200, Thermo Fisher Scientific, Waltham, MA) diluted in water. PLSs are commonly utilised in aerosol research and were used to simulate pathogens Exposure simulation: 1. Fluorescent tracer mixture was applied to PPE using 1000 mL in a pesticide hand sprayer (RL Flo‐Master, Lowell, MI; 2000 mL capacity) and 5 sweeping passes of sprayer from head to feet on the front and back of the HCW. 2. A PLS suspension (25 mL) was aerosolised using a 3‐jet Collison nebuliser (Mesa Laboratories, Inc, Butler, NJ) for 4 min of continuous aerosol generation while the HCW turned 90° every 60 s
Participants	Baseline characteristics 48 participants were included in the study Enhanced doffing protocol Male %: not reported Age (m ± SD): not reported Occupations: 13 HCWs and 13 doffing assistants Employment duration: not reported CDC doffing protocol Male %: not reported Age (m ± SD): not reported Occupations: not reported Employment duration: not reported Overall Male %: not reported Age (m ± SD): not reported Occupations: total 48 participants Employment duration: not reported Inclusion criteria: not reported, but study authors included: adults (male/female) with no prior experience doffing enhanced PPE Excluded criteria: not reported
Interventions	Intervention characteristics Enhanced doffing protocol: doffing with extra instructions Intervention aim: to mitigate the risk of self‐contamination during PPE doffing Content of the intervention: participants received approximately 2 h of training prior to doffing PPE. The curriculum for both the treatment and control groups included a basic introduction to germ theory, modes of pathogen transmission, types and purpose of PPE, and basic tenets of infection prevention. Both control and intervention groups were shown the PPE components they would doff during the study. The intervention group participants watched a video about teamwork concepts and their application in healthcare. The training included information about potential risks in the doffing process, the benefit of teamwork in PPE doffing, and the roles and responsibilities of the doffing team members. Participants were instructed on teamwork strategies including use of verbal and nonverbal communication (e.g. closed‐loop communication); developing, maintaining, and updating situational awareness (e.g. monitoring inadvertent contact of the HCW with other team members or room surfaces); mutually supporting team members; and the importance of verbalising safety concerns. They were then shown a video that demonstrated the intervention package doffing process. The intervention package addressed various components of the doffing process, including tools/technology (e.g. PPE selection), people (e.g. roles, teamwork), task (e.g. technical aspects of PPE removal), and environment (e.g. doffing room characteristics). PPE consisted of surgical gown, isolation gown, inner and outer gloves, PAPR, PAPR hood, tape on sleeves and boot covers. The intervention group had, in addition, examination gloves. The boot covers differed between intervention and control group. CDC doffing protocol Intervention aim: same goal Intervention duration per session: not reported Intervention frequency per week: once only Intervention duration (months): N/A Provider of the intervention: CDC Content of the intervention: after training, the control group participants watched a video that highlighted general facts about respiratory etiquette and the importance of covering your cough to prevent the spread of respiratory infections, followed by a video that demonstrated enhanced PPE doffing based on the 2015 CDC recommendations
Outcomes	How the outcome was measured: from the fluorescent tracer slurry ‐ detection was by direct visualisation in a dark room using ultraviolet light. (1) The number of body sites contaminated and (2) the extent of contamination at each site were recorded. PLS detection was performed by (3) counting via epifluorescent microscopy and (4) quantifying the number PLSs per cm² of skin or per m3 of sampled air. (5) Teamwork dynamics were assessed via video and coded using a task analysis of the process sets and subsets (checklist). (6) The National Aeronautics and Space Administration (NASA) Task Load Index (NASA‐TLX) questionnaire assessed perceptions of workload during doffing (7) The Team Strategies and Tools to Enhance Performance and Patient Safety Teamwork Attitudes Questionnaire (T‐TAQ) assessed attitudes toward teamwork Body sites with fluorescent marker Outcome type: dichotomous outcome Reporting: fully reported Unit of measure: body site Direction: lower is better Data value: endpoint Body sites with PLS Outcome type: dichotomous outcome Reporting: fully reported Unit of measure: body site Direction: lower is better Data value: endpoint
Notes	Outcomes Median and IQR of 22 possible contaminated sites reported. For Fluor Marker: intervention 1 (1‐2) control 5 (2‐5) For PLS out of 12 possible contaminated sites: intervention 4 (2‐5) control 5 (5‐8). These were transformed to means and SDs for use in the data tables.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Quote: "randomly assigned to the control or intervention condition and then to the role of HCW or DA." Judgement comment: method of random assignment not reported
Allocation concealment (selection bias)	Unclear risk	Judgement comment: method of allocation concealment was not reported
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Quote: "role as either HCW or DA." Quote: "After training, all participants were given the opportunity to ask questions and were informed about their randomly assigned" Quote: "Study participants were blind to their group assignment." Judgement comment: group assignment (intervention and control) was blinded; role was blinded to participants until after training and before the doffing intervention
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Quote: "The contamination forms were deidentified and assigned randomized numbers for scoring purposes. Two IPs, blinded to experimental assignment, independently scored each form" Judgement comment: infection preventionists were the outcome assessors and were blinded to intervention group
Incomplete outcome data (attrition bias) All outcomes	Low risk	Quote: "Forty‐eight study participants (35 females, 13 males) were randomly assigned to the control (n = 22) or intervention group (n = 26)." Quote: "Participants in each study arm were randomly assigned to the role of control HCW (n = 11), control DA (n = 11), intervention HCW (n = 13), or intervention DA (n = 13). For the fluorescent tracer, 11 HCWs (84.6%) in the intervention group and all 13 control HCWs (100%) contaminated at least 1 body area." Quote: "Coding and scoring of teamwork behaviors exhibited in the videotaped doffing sessions were completed for 10 intervention and 11 control teams. Technical difficulties resulted in missing videotapes for 3 intervention teams." Judgement comment: main outcomes were listed within the methods (but scattered and hard to find). All recruited participants completed the interventions and outcomes were collected. 1 typographical error (I assume) in reporting fluorescent tracer contamination (they reported 13 control HCWs but there were only 11). 3 sets of teamwork behaviour outcomes recorded in videos from the intervention group were lost. However, despite the missing data, there was a plausible difference in median (IQR) between groups that may not have impacted the observed effect size.
Selective reporting (reporting bias)	Low risk	Judgement comment: the availability of the study protocol is not reported in the paper, but it is clear that the published report includes all expected outcomes for this type of study.
Other bias	Low risk	Judgement comment: no other bias detected