The COVID-19 pandemic has had a profound effect on emergency care in Canada and around the world. Emergency departments (EDs) have been reconfigured to meet the new demands of the pandemic [1]. Some of the reconfigurations include increased capacity, physical divisions into identified risk zones, changes in care protocols, redefinition of staff tasks, and development of protective measures against COVID-19 nosocomial transmission [2,3].

In order to prevent transmission at the peak of the pandemic, personal protective equipment (PPE) was required for all ED care providers (eg, physicians, nurses, and nursing assistants) in contact with patients with proven or suspected COVID-19. PPE encompasses all clothing and other equipment used to protect health care workers from injury or infection [4], and is considered the last step in the hierarchy of controls in infection prevention [5,6]. With respect to the COVID-19 pandemic, the term PPE refers primarily to two options (see Multimedia Appendices 1 and 2) [7]: (1) standard PPE, which includes procedural masks, isolation gowns, nonsterile gloves, and ocular protection; and (2) enhanced PPE which includes respirators (eg, N95), water-repellent gowns, long-cuff nitrile procedure gloves, and face shields. Since the beginning of the pandemic, the understanding of the COVID-19 modes of transmission and the most effective equipment to protect against it have greatly evolved. In many jurisdictions, enhanced PPE was initially reserved for aerosol-generating procedures (eg, endotracheal intubation), that is, aerosolized particles of <5 μm that can remain suspended in the air for extended periods of time and circulate over long distances on air currents [8]. In Canada, in 2021, the use of enhanced PPE with a respirator has gradually become the norm when caring for patients with confirmed or suspected COVID-19 cases, even in the absence of aerosol-generating procedures [9-11].

The rules for PPE use in health care settings have fluctuated greatly and have been a source of uncertainty [8,12]. Often developed as a compromise between incomplete scientific evidence [7,8,13], uncertain PPE supplies [14-16], and staff concerns [12,17], these rules were sometimes so confusing for health care workers that their trust towards the infection control teams and managers was diminished [17]. Although the benefits of PPE for protecting health care workers against SARS-CoV-2 have been demonstrated [7,13,18], the level of evidence is relatively low and its use is more or less influenced by subjective preferences and personal values [19].

Although the achievement of a certain level of immunity in the population provides significant protection against severe or chronic symptoms of COVID-19 [20], important questions remain; what types of PPE will be needed to protect health care workers as SARS-CoV-2 becomes endemic and circulation in the community continues unabated? Somewhat less discussed since the beginning of the pandemic, the societal consequences of PPE use must also be considered. What will be the opportunity cost for public institutions to pursue the expanded use of PPE [21-23]? What is the ecological cost of continuing on this path [24-27]? More than ever, evidence is needed to evaluate PPE use and to provide recommendations that ensure the clinical, economical, and ecological efficiency of its use.

The protocol for this review is registered in the PROSPERO database of systematic reviews (CRD42022302598) and we will follow recommendations of the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analysis) guidelines for its reporting [39].

What effect does wearing PPE have on health care workers’ risk of becoming infected with SARS-CoV-2, being secondarily absent from work, admitted to the hospital or to the intensive care unit (ICU), or die? Table 1 shows the PICOS framework for the systematic review.

We will search for studies published since December 2019 (identification of SARS-CoV-2 in Wuhan) comparing health care workers’ use of disposable or reusable, standard or enhanced PPE to no protection used or a different combination of PPE components. We will use Medline (Ovid), Embase, Cochrane Library, CINAHL, Epistemonikos, ClinicalTrials.gov, MedRxiv, and Web of Science search engines, without language restrictions. The search strategy was developed with the assistance of a qualified librarian and is available in the appendices (Multimedia Appendix 3). This strategy will be executed three times: (1) at the start of the systematic review (May 2022); (2) at the end of the prepublication data extraction (June 2023); and (3) before phase 4 (January 2024) to update the meta-analysis to adequately inform our focus groups.

Using the Covidence systematic review software (Veritas Health Innovation), 2 independent reviewers will select abstracts, read the articles, and determine their eligibility. All experimental or observational studies (cohort, case-control, or cross-sectional studies) comparing the use of PPE with the absence of PPE with regard to health care workers’ risk of SARS-CoV-2 infection, hospitalization (ward or ICU), death, or absence from work, will be included in the review. Studies evaluating the effect of PPE when worn by nonhealth workers, narrative reviews, editorials, modeling studies without original clinical data, and practice guidelines will be excluded. Systematic reviews will also be excluded but will be consulted to identify original articles potentially missed by our search strategy. Bibliographies of selected articles will be reviewed in the same manner.

In total, 2 independent reviewers will collect data from the selected articles using a previously tested extraction grid. From each article, they will extract, when available, the following variables: authors; title; date and scientific journal of publication; country where the research was conducted; study design; population studied; sample size; age; gender; sex; vaccination or immune status; and comorbidities of the participants, study outcomes, and results. The quality of the selected studies will be assessed by 2 reviewers using the Cochrane Risk of Bias Tool for randomized studies [40] and ROBINS-I for observational studies [41].

Disagreement between reviewers will be resolved through discussion and consensus. If a disagreement persists, a third reviewer will mediate to reach a final consensus.

The results will be presented in tabular and narrative form and will compare the benefits of standard or enhanced PPE to the absence of PPE use. Based on previous systematic reviews, we know that a random effects model meta-analysis is feasible and will then be conducted. According to the original data available, we will calculate estimates of relative risks (RR) or odds ratios (OR) on the risk of transmission of SARS-CoV-2 and of the other outcomes evaluated. The necessary numbers of PPE that must be used (number needed to treat) to prevent one SARS-CoV-2 transmission, admission, or death among staff will be calculated secondarily. Randomized studies will be analyzed separately from observational studies, as recommended by the Cochrane Handbook [42,43]. If studies are in sufficient numbers, we will conduct subgroup analyses by analyzing separately (1) the studies conducted in EDs; and (2) the studies conducted after the start of vaccination in the country where the research took place. Heterogeneity between studies will be assessed using the I², a statistic that estimates the percentage of variation in results between studies that are not explained by chance. Thus, an I² value between 0 and 40% will be considered to represent a low level of heterogeneity, while values between 30% and 60% represent a moderate level, between 50% and 90% a substantial level, and between 75% and 100% a considerable level [44]. The causes of heterogeneity will be examined by stratifying our outcome measures by (1) the country where the studies were conducted and (2) the clinical departments or units from which the data originated (eg, COVID-19 unit and ICU). We will also conduct a sensitivity analysis excluding studies with a high risk of bias. Finally, the quality of evidence for each outcome (high, moderate, low, and very low) will be assessed using the GRADE (Grading of Recommendations Assessment, Development, and Evaluation) approach [45].

The cost of PPE use goes beyond the cost of purchasing its components. Wearing PPE means donning and doffing by care providers who, during this time, cannot engage in direct care, which is a cost to the health care system. Using PPE also incurs costs for material disposal (eg, garbage collection) and overhead (eg, supply management). Therefore, we will estimate the mean cost of wearing PPE in the CHU for contact with a possible case of COVID-19 using a time-driven activity-based costing method [46,47]. This method accounts for all expenditures incurred during direct patient care (eg, staff salaries and medical costs), consumables (eg, masks), and management costs (eg, disinfection service). It uses the duration of care processes to estimate the associated cost; the longer a process (eg, donning PPE) takes to be performed, the higher the cost. It provides a simpler, more accurate, and more reliable way of estimating the cost of health services than other methods frequently used in research or management, such as diagnosis-related group methods (DRG), the level of relative intensity of resource use (NIRRU) or the conventional activity-based costing method (ABC) [48-50]. Time-driven activity-based costing was previously used in many health care settings [48,51-56] and our team successfully adapted it for use in the ED [57-59]. For this project, we will apply this costing method to the care pathway of patients suspected with COVID-19 in the ED. We will analyze the processes involved in PPE use and disposal. The disposal component will include both trajectories of completely disposable PPE versus reusable gowns, as one of our sites used reusable gowns at the beginning of the pandemic. Cleaning of the reusable gowns was done by a subcontractor (Partagec), as such, the cleaning cost will be estimated based on the weight of the gowns used as stipulated in the contract with Partagec.

The preceding steps will estimate for each category of ED care providers (eg, nurses and physicians) the mean cost incurred during the pandemic for one standard and one enhanced PPE use, which includes donning, doffing, disposing, and disinfection or cleaning. A set of subanalyses will then estimate (1) the minimum and maximum costs of PPE use based on the minimum and maximum supplier unit costs of the PPE components purchased by the CHU in 2020-2021; (2) the weighted mean cost of PPE use based on the weight of participation of each category of providers in the care pathway, estimated using the mapping of a typical care pathway for COVID-19 in the ED (Multimedia Appendix 4); (3) The total cost of PPE use in the CHU EDs based on the number of suspected COVID-19 cases that consulted since the start of the pandemic, the weighted mean cost of using PPE, and the number of contacts with PPE per episode of care estimated with the mapping; and (4) mean costs based on component costs in the year prior to the pandemic (financial year 2019-2020) to estimate the effect of the shortage on equipment selling prices.

The ecological footprint of standard and enhanced PPE will be estimated through a life cycle assessment that will measure the pollutant load of all material and energy resources used (inputs) and residual materials generated (outputs) by their use (Multimedia Appendix 5 for a typical life cycle diagram). The analysis will assess the environmental impact of the consumables (eg, gloves) and processes (eg, disinfection of face shields) required for PPE use in CHU EDs from cradle to grave, that is, from their design and manufacturing to their final disposal. The life cycle assessment methodology, which is well established in the industrial field, has previously been used in the health field as well [62-64].

The results of the analyses of the previous 3 phases on benefits (OR, RR, and number needed to treat), costs, and ecological footprint will be scaled to a single PPE use for contact with a suspected or confirmed COVID-19 case by a care provider in the ED. They will also be secondarily extrapolated for the full duration of the pandemic from the number of suspected COVID-19 cases assessed in the CHU EDs during this period. Using data from the Institut national de santé publique du Québec on the hospitalization rate of COVID-19 cases in Quebec [70] and the cost of an average hospitalization at the CHU obtained by the Cost per Care Pathway and Service (CPSS) system [71], we will estimate the potentially avoided costs by preventing hospitalizations through PPE. We will do the same by estimating the avoided costs related to staff absenteeism using a human capital approach [72] on the basis of the care providers’ salaries identified in phase 2. A cost-consequence analysis will then be carried out by presenting, in parallel, the benefits, costs, and environmental consequences with their 95% CIs in a simple and meaningful disaggregated format. A cost-consequence analysis is a type of economic evaluation that assesses a wide range of costs and consequences associated with an intervention. It can include all types of positive or negative effects of that intervention, including those that are health-related (eg, risk of transmission) and those that are not directly health-related (eg, environmental impact). By presenting the results separately in a nonaggregated manner, it allows stakeholders to determine for themselves the weight and value assigned to the costs and consequences presented for informed decision-making. We believe that cost-consequence analysis is most appropriate for our project because (1) the values and preferences of the stakeholders involved in PPE use are multiple and often divergent, making aggregative ratios or calculations inappropriate (eg the incremental cost-effectiveness ratio [ICER] assesses only one outcome or one benefit) and (2) the economic value of the benefits and consequences of PPE is not clearly elucidated, making cost-benefit analysis (ie, an analysis where all the consequences of intervention are converted into monetary value) problematic.

The cost-consequence analysis will be presented and discussed in 4 focus groups [73] of 8 participants each [74], via 2 one-hour videoconferences for each group led by a facilitator and a note taker. A third videoconference will be held if not all topics have been covered in the first two. Individual interviews will be offered to participants who prefer this approach or who are unable to attend group meetings. The first group will bring together health care managers responsible for budgets; the second will bring together experts responsible for developing recommendations on PPE use in hospitals; the third will gather the perceptions of health care providers, nursing assistants, respiratory therapists, and physicians; and the last group will survey the perspective of patients and citizens. Additional groups of participants will be recruited as needed until the data are saturated. A purposeful sampling strategy will be used to ensure representation of various levels of authority (departmental, regional, local), professions (eg, nurses), regions (urban, semiurban, and rural), and individual characteristics (gender, ethnicity, and age) [75]. Participants will be presented with all the results in a simple, adapted format, after consultation with experts in content (eg, economists and environmental engineers) and information formatting (eg, information designers and knowledge transfer experts). The discussions in each group will be semistructured. The primary objectives of the semistructured discussions will be to (1) analyze the perspective of each group on the study results and (2) identify the cost-consequence analysis results (benefits vs costs vs environmental impact) that participants believe should be prioritized for future recommendations regarding PPE use in the ED. The meetings will be recorded and fully transcribed (verbatim). In total, 2 independent reviewers will code the responses on NVivo 12 (Lumivero) to structure the data and bring out the main themes. Their disagreements will be reconciled by discussion or, if necessary, by the intervention of a third researcher. The analysis will take an inductive approach to capture the different perspectives on PPE use in the ED from the data collected and then develop a theoretical framework.