Staying proper with your personal protective equipment: How to don and doff

Abstract

Introduction

The global COVID-19 pandemic highlighted the importance of protecting frontline healthcare workers from novel respiratory infections while also exposing the limited instruction that medical students receive on proper donning of personal protective equipment (PPE) and more importantly the safe doffing of contaminated PPE to minimize their risk of nosocomial infection. The best methods of providing this kind of instruction have not yet been determined.

Methods

Anesthesiology interns and CA-1 residents were trained on proper PPE donning and doffing for AGPs using a methodology based on Miller's pyramid and following a “knows-knows how-shows-does” progression. Participants donned PPE without instruction and were sprayed with Glo Germ® to identify areas of contamination, after which they received both video and in-person instruction on best practices for donning and doffing PPE for AGPs. Following instruction, they again donned PPE and were sprayed with Glo Germ® to identify areas of contamination.

Results

54 participants completed the study. Before training, overall donning compliance was 60% and overall doffing compliance was 48%. Overall, 70% were contaminated after PPE doffing, with 46% having multiple sites of contamination. After training, donning compliance increased by nearly 30% (P < 0.001), doffing compliance increased by over 20% (P < 0.001), and overall contamination decreased by nearly 30% (P = 0.029), with multiple-site contamination decreasing to only 6% (P = 0.013).

Discussion

While best methods for providing instruction regarding topics such as PPE donning and doffing have not yet been determined, we have demonstrated that the underlying knowledge base from medical school regarding proper donning and doffing for respiratory isolation is insufficient for preventing self-contamination, and that Miller's pyramid-based training using both video and in-person instruction combined with task execution by learners can improve compliance with PPE donning and doffing protocols and more importantly decrease skin contamination among a group of early training anesthesiology residents.

1. Introduction

The novel coronavirus SARS-CoV-2 (COVID-19) is causing a global pandemic and highlighting the importance of protecting healthcare providers, especially vulnerable frontline workers such as anesthesiology residents. As part of their duties, anesthesiology residents perform aerosol-generating procedures (AGPs) while caring for COVID-19 patients, whether confirmed or under investigation, in operating rooms and other settings, including the regular floor, intermediate care units, and intensive care units. At the start of the pandemic, personal protective equipment (PPE) was in high demand and manufacturer supplies were limited. As a result, the availability of the appropriate components to assemble PPE as recommended by the Centers for Disease Control and Prevention (CDC) for those caring for COVID-19 patients was often limited, and substitute items that varied by institution were implemented. Due to the highly specific full-body PPE required while caring for COVID-19 infected patients, specific training for the proper donning and more importantly the safe doffing of contaminated PPE reduces the risk of the occupational hazard of nosocomial infections. Unfortunately, education for the skills required for proper PPE donning and doffing is not routinely part of medical school or graduate medical education curricula. As the pandemic arrived, we reviewed the education of our current first year anesthesiology residents to ascertain if their training had provided sufficient knowledge on how to protect themselves with PPE donning and doffing when providing care for patients with highly infectious pathogens. Although our institution has each incoming house staff member perform a series of orientation activities including objective structured clinical examinations, the components of this curriculum for infection control and prevention focused on hand washing [1]. Recent infection control training has been previously shown to predict improved adherence to recommended barrier precautions with appropriate PPE use [2].

To ensure appropriate education of the house staff who were to provide care throughout the hospital, including in the operating rooms, we designed a simulation for donning and doffing the recommended PPE. Although the best way to educate healthcare workers and to maintain the skills needed for the appropriate use of PPE remains unclear [3], training via written material only or as a traditional lecture resulted in more errors than using face-to-face, computer simulation, and videos [[2], [3], [4], [5]]. We designed an educational session on PPE donning and doffing techniques to train anesthesiology residents who perform AGPs. This simulation involved face-to-face and video training and was evaluated for its effectiveness to educate residents on proper infection control.

2. Materials and methods

This project received IRB approval for informed consent that permitted filming the participants who agreed to participate. As part of the study, a video of the PPE available within the institution, including correct steps for donning and doffing for an AGP procedure, was produced with the content approved by an infectious disease physician expert from infection control at our institution [6]. All the house staff who participated had N95 mask fittings before the simulation and were instructed to arrive dressed in scrub suit attire, as well as to bring another set of scrub suit attire with their personally fitted N95 respirators provided by UF Health. This study used the standard PPE available to house staff throughout UF Health at the start of the pandemic because anesthesiology residents would be deployed in the operating room and to care for inpatients on the regular wards as well as in intermediate and intensive care units. Clinical assignments would include caring for patients with COVID-19 and performing AGPs. The available PPE options included nonsterile gowns (permeable and nonpermeable options), sterile surgical gowns, towels, clamps, tape, face shields with a single elastic strap, gloves (sterile and nonsterile varieties), various types of surgical masks including those with eye protection, surgical hoods, and surgical caps and bouffant. Initially, the house staff were instructed to don what they thought was appropriate PPE to use while caring for a COVID-19 positive patient and to observe contact, droplet, and airborne infection control precautions without assistance from others. The various types of PPE available in our hospital setting were made available and participants were observed during this process. House staff members were surveyed before the start of the simulation, and none had received formal standardized training on proper donning and doffing to care for COVID-19 patients who required contact, droplet, and airborne precautions for infection control.

Once each house staff member was outfitted in what they deemed appropriate PPE, they were sprayed with a fluorescent marker (Glo Germ®, Glo Germ Company, Moab, UT) to simulate a contamination with COVID-19 from an infected patient. The fluorescent marker was mixed with isopropyl alcohol in an 80/20 mixture and each participant was sprayed from a 3-ft distance once to each of the following areas: face, chest, forearms, and hands. Each individual was then filmed under blacklight in their full-body PPE and instructed to doff, assuming contamination with COVID-19. After doffing, each individual was again filmed under blacklight and the areas of contamination able to be seen via visible fluorescence were identified. Each individual then received verbal feedback from the instructor with a specific focus on contaminated areas after doffing, which included visualizing areas with the blacklight. Individuals were then instructed to remove any fluorescent contamination on their skin and hair and to change scrub suit attire if contamination of scrubs occurred.

Participants then viewed a video regarding how to don and doff PPE [6] appropriately, which was further reinforced with a face-to-face demonstration. Participants were again checked under blacklight to ensure that areas of contamination had been removed before donning their PPE again. They followed the same sequence of donning PPE without assistance as demonstrated in the video before being sprayed again in the same manner with Glo Germ® and then filmed doffing their complete PPE before being checked under blacklight for areas of contamination. The study was completed in 2 groups. Group 1 participants completed their presimulation and postsimulation testing in June 2020 and Group 2 participants completed their testing in February 2021. A small subset of participants from Group 1 were reexamined in February 2021 (8 months later) to evaluate long-term retention.

The videos were then cataloged, categorized, and randomly assigned to an independent observer who was blinded to whether the video was filmed before or after the education training. The independent observer noted and scored noncompliance with the selection of type and use of PPE for donning using a predetermined observational checklist based on the PPE available in our institution. The number of contaminated sites and the presumed risk associated with those areas were noted.

Data were analyzed using JMP Pro 16 (SAS Institute Inc., Cary, NC). For each area within donning and doffing, the compliance percentage (%) was calculated and an overall compliance percentage score across all areas was calculated. These were summarized by mean % and standard deviation (SD). The difference between presimulation and postsimulation training was evaluated using Wilcoxon signed rank tests (for paired analyses). Contamination was first coded as any or none for 1) overall body and 2) within each site evaluated. For overall contamination, the occurrence of multiple contaminated sites (2 or more) was also assessed. The contamination difference between presimulation and postsimulation training was evaluated using McNemar's test. Exploratory analyses were performed to examine longer-term retention in a small subset of participants using similar approach detailed above. Secondary analyses examined the effects of gender, hair length, and facial hair on initial donning and doffing compliance and contamination using Kruskal-Wallis tests and chi-square tests. P < 0.05 was considered statistically significant.

3. Results

54 participants were enrolled in the study (Supplemental Table 1). A majority of our participants were men (n = 34, 63%) and the majority (n = 28, 70%) were interns (PGY-1). Hair and facial hair characteristics were recorded via observations from videos. Over half (n = 32, 59%) had hair above ear or not visible under cap, 6% (n = 3) had hair below ear but above shoulder, and 35% (n = 19) had hair below shoulder, all of whom were women. Only n = 7 had visible facial hair, all of whom were men.

Fig. 1 reports the compliance percentages for donning and doffing before the simulation training across specific areas and overall. For donning compliance before simulation training, areas associated with eye protection (100%), isolation gown (99%), surgical mask (93%), and N95 mask (83%) had the highest compliance rates, whereas the area associated with neck cover (12%) had the lowest rate. Overall donning compliance was 60%. For doffing compliance before simulation training, the highest compliance was for the area associated with removing outer gloves and gown (83%), and low rates of compliance were reported for areas associated with hand hygiene both at the beginning (2%) and end (2%) of doffing as well as for removing gloves (16%), neck towel (14%), and surgical hood (6%). Overall doffing compliance was 48%. Secondary analyses examined differences due to gender, hair length, and facial hair. Overall, there were not robust differences in compliance and contamination due to these factors. For donning, the only significant difference occurred for neck cover, in which those with facial hair (29% ± 49%) showed higher compliance than those without (4% ± 19%). For doffing, the only significant differences were in removing neck towel for gender (P = 0.027) and hair length (P = 0.018). Women (29% ± 47%) had better compliance than men (6% ± 24%) and those with longer hair (31% ± 48%) had better compliance than those with shorter hair (6% ± 24%). Interestingly, there were no significant differences for donning (P = 0.860) and doffing (P = 0.725) N95 masks due to facial hair. Finally, for contamination there were no statistically significant differences (P > 0.05) observed for gender, hair length, nor facial hair.

Fig. 1.

Changes in Donning (A) and Doffing (B) Compliance (%) Presimulation to Postsimulation Training for Specific Areas as Well as Overall Compliance. P-values from Wilcoxon signed rank tests. * indicates p < 0.05.

Fig. 2 reports the participants' contamination occurrences after donning and doffing before they went through the simulation training. The highest rates for contamination were for areas associated with the hair/head (31%) and neck (30%); the lowest rates were associated with the forearm (9%) and hands (13%). The overall contamination rate (at any site) in doffing before the simulation training was 70%, with 46% having 2 or more sites contaminated.

Fig. 2.

Changes in Contamination Occurrence (%) Presimulation to Postsimulation Training for Specific Areas as Well as Overall. P-values from McNemar's test * indicates p < 0.05.

Postsimulation assessments were completed in n = 35 participants (Supplemental Table 1). Fig. 1A shows the changes between presimulation and postsimulation training for donning areas. Statistically significant increases in compliance percentage were observed for areas associated with head covering (P = 0.044), gloves (P < 0.001), and neck covering, (P < 0.001). Overall donning compliance also increased by nearly 30% (P < 0.001). Fig. 1B shows the changes between presimulation and postsimulation training for doffing areas. Statistically significant increases in compliance percentage were observed for areas associated with hand hygiene at the beginning (P < 0.001) and end (P < 0.001) of doffing, as well as for removing surgical hoods (P < 0.001), removing face shields/googles/outer masks (P = 0.004), removing neck towels (P < 0.001), and removing gloves (P < 0.001). Overall doffing compliance also increased by over 20% (P < 0.001).

Fig. 2 shows the changes between presimulation and postsimulation training for contamination occurrence. Overall contamination occurrence rates decreased by nearly 30% (P = 0.029). The occurrence of contamination at multiple sites decreased by 40%, from 46% presimulation to 6% postsimulation (P = 0.013). Of note, 1 individual's hair was contaminated and the individual was unsuccessful in removing the fluorescent marker from the area of hair. This was noted during examination with the blacklight before the second doffing and was determined to have remained the same size; therefore, it was not counted as another area of contamination for the second round of doffing. Fig. 3 shows an example of contamination with fluorescent marker, and Supplemental Fig. 1, Supplemental Fig. 2 illustrate the differences in participants' presimulation and postsimulation donning.

Fig. 3.

Examples of Fluorescent Markers for Contamination Measurement.

Supplemental Fig. 1.

Example of Personal Protective Equipment Donned Presimulation.

Supplemental Fig. 2.

Example of Personal Protective Equipment Donned Postsimulation.

Exploratory examinations of long-term retention were assessed in a small subset of participants (n = 8, Supplemental Table 1). Fig. 4 shows the changes in donning (A) and doffing (B) total compliance, as well as contamination in each participant across the 3 time points. For donning, total compliance for retention was significantly increased from presimulation (P < 0.001) and was not significantly different from postsimulation (P = 0.484). Conversely, for doffing total compliance for retention was not significantly different from presimulation (P = 0.773) and was significantly decreased from postsimulation (P = 0.008).

Fig. 4.

Individual Changes in Donning (A) and Doffing (B) Compliance (%) Presimulation to Postsimulation Training and Long-Term Retention in Subset of n = 8 Participants.

4. Discussion

This study demonstrates, using a before and after model, the effectiveness of PPE education and training for PPE donning and doffing adherence among anesthesiology residents who perform AGPs while caring for COVID-19 patients. The pandemic's risks of nosocomial infection have raised the needs for training and compliance with infection control procedures to protect healthcare workers and prevent self-contamination and spread to others. The first educational issue involves ensuring that the proper PPE is being used according to recommended guidelines. For COVID-19, the recommendations for PPE are gloves, masks, goggles or face shields, long-sleeved gowns, and medical masks; further recommendations for AGPs include a N95 respirator [3,[7], [8], [9]].

Anesthesiology residents and other healthcare workers need to know the appropriate PPE to use and how to don and doff it. The first step when educating providers involves knowing the PPE required for procedures that carry a higher risk of infection, such as AGPs. For anesthesiology residents, this commonly includes AGPs such as endotracheal intubation/extubation and other procedures including but not limited to bronchoscopy and cardiopulmonary resuscitation. Even with the use of appropriate PPE for AGPs, probably the highest risk for infection of healthcare workers involves inappropriate removal of the PPE [10], which can result in self-contamination. To provide training and education on these issues, anesthesiology residents first chose from all forms of PPE available at our institution to don what each considered appropriate PPE. After being sprayed with fluorescent marker, they then doffed their PPE. This provided participants an opportunity to evaluate specific areas of contamination via blacklight to what they believed was fully protective PPE following donning and doffing the contaminated PPE.

This technique identified preexisting knowledge gaps. Survey-based work in a large teaching hospital has previously demonstrated significant knowledge deficits concerning standard infection control and isolation precautions that correlated with knowledge and training [11]. In addition, focus group studies during the SARS outbreak revealed that healthcare workers did not think that knowledge deficits represented a major barrier to compliance with infection control practices [12,13]. In 1 study, the rate of contamination significantly improved after demonstration of the CDC-recommended gown removal [14].

In our educational model, after the baseline donning and doffing, participants then watched a video followed by an in-person demonstration of donning and doffing appropriate PPE for an AGP procedure. This PPE education involved identifying gaps in knowledge and task performance using Miller's pyramid, during which the residents had to “show how” they would don and doff the PPE followed by immediate feedback via an assessment displaying areas of contamination. The next step of education involved a video illustrating appropriate technique followed by a face-to-face demonstration. The anesthesiology residents then had the opportunity to “show how” they perform donning and doffing with this new knowledge by repeating the processes to document improvement as measured by fewer areas of contamination.

A recent Cochrane Review that focused on PPE equipment for preventing the nosocomial spread of highly infectious diseases due to contaminated body fluids indicated that the best ways to train and maintain the skills remain unclear [3]. There are many challenges to designing a model education program with no accepted standard, recognizing that healthcare providers have different backgrounds and training experiences. We also recognize that PPE is not standardized across healthcare institutions, and the PPE with which providers were familiar may differ from that which is available. Acknowledging these challenges, the decision was made to use >1 technique for training. The first we employed was to actively assess current knowledge gaps by performing baseline donning and doffing for an AGP procedure using only options available in the participants' healthcare system and identifying areas of contamination after the donning and doffing sequence. Video training and face-to-face demonstration then followed. The learners then had the opportunity to address their identified gaps in knowledge by donning and doffing again followed by visual inspection of areas of contamination.

In this study, we examined subjects for contamination in a before-and-after model with each participant serving as their own control. There was no standard definition for contamination, with noncompliance measured as the number of participants not using appropriate equipment, failing to follow the correct protocol order, or omitting elements [[15], [16], [17], [18]]. In other studies, contamination involved identification by using a sprayed fluorescent marker that included the areas of the body contaminated or the number of people or areas of the body contaminated [[19], [20], [21]].

During this training, we emphasized the use of face shields, as an epidemiology study showed that not wearing a face shield during an AGP on patients with infectious respiratory diseases resulted in a 3-fold higher risk of nosocomial infection [22]. Contamination rates are also determined by the type of PPE; face shields with a single Velcro or elastic strap that require only 1 hand to remove are the easiest to don and doff [23] and were available for this study.

There are limitations in this study. Despite individuals serving as their own controls, the sample size is small and from a single institution. There was a technical problem with some of the videos which limited the sample size for comparison; as such, only a small subset has completed long-term retention examination. Furthermore, hair length and facial hair were recorded from video, which may have affected accuracy. We elected to standardize to PPE available at our institution, recognizing the availability of many types of PPE with varying components that may be available for use at other institutions. The education was primarily designed to provide protection for our residents using the PPE available and could be replicated at another institution using its available PPE. The limited number of components for PPE may have decreased the number of errors made as there were fewer options from which to choose due to shortages. Participants had different levels of experience, ranging from no exposure to COVID-19 patients to involvement in the care of no >5 COVID-19 patients, although none had received formal education and/or training in PPE selection, donning, or doffing for AGPs or in caring for a highly infectious respiratory disease (such as COVID-19) patient. Practical experience caring for COVID-19 patients was also limited, as the study participants were either interns within their first month of clinical work or first-year anesthesia residents within their first month of clinical anesthesia training. This study used fluorescent markers, which may underestimate contamination as the particle size and weight is likely substantially higher than the microparticles that carry COVID-19, and bacteriophage contamination has occurred without fluorescent biomarker detection in 2 studies [24,25]. An example of gross contamination seen with these fluorescent markers is illustrated in Fig. 3, panels A and B. With donning and doffing procedures, it is unclear if they should be performed with a helper buddy removing part of the PPE [3]. A helper buddy and/or an independent observer is not consistently available in all areas that our anesthesiology residents would care for these patients, so for this education module the decision was made not to employ a helper buddy and/or trained observer to provide feedback during the donning and doffing procedures to replicate a typical clinical setting situation. Using a helper buddy and/or independent observer could potentially decrease performance errors. Repeating this study to see if there is improvement in performance with the use of a helper buddy and/or independent trained observer for donning and doffing represents an area for potential future investigation.

Unfortunately, skills retention after training is understudied and results from studies examining longer-term follow-up are conflicting. Our exploratory analyses on a subset of participants support higher retention in donning compliance but not doffing compliance. These findings may support more emphasis on doffing in educational trainings. Future research is needed to see if any specific areas are related to better or worse retention.

Previous work by Northington et al. instructed EMS workers on the proper donning and doffing of Level C PPE for chemically contaminated environments. This training occurred over a single 90-min session which involved 30 min of didactic training followed by supervised practice until the task could be completed correctly. Follow-up evaluation 6 months later revealed that only 14% of participants could still correctly don and doff their PPE without critical self-contamination [26]. Similarly, West demonstrated that when pediatric nurses are trained in basic infant life support their retention wanes as early as 6 weeks after training [27]. At the other end of the spectrum, Clarke et al. instructed practicing anesthesiologists in invasive subglottic emergency airway access using a simulation-based skill mastery model. This training took place over a single 2.5-h session which involved training on 3 different techniques with a 15-month follow-up to evaluate skills retention on 2 of the techniques. This study found approximately 80% retention at 15 months [28]. With these differing results in long-term retention of skills, it remains difficult to draw widely applicable conclusions except to say that skills retention should be periodically evaluated and retraining provided as necessary to maintain an appropriate knowledge base and practical skillset. This, taken together with the work of Verbeek et al. [3] reporting only approximately 14% of participants demonstrated full compliance at 6 months after training, shows that long-term retention of knowledge is problematic and suggests that future work should investigate to what degree new knowledge gained through this education is retained, and what additional steps or educational techniques can or should be employed to maximize retention.

Additional future studies should specifically evaluate the effects of employing additional personnel to both supervise and guide proper donning and doffing. Verbeek et al. have concluded that using assistants or helper buddies during donning and doffing, although they may not be consistently deployed, and using checklists improves compliance with proper donning and doffing of PPE [3]. While institutions may choose to overlook the deployment of such assistants or buddies as a cost-saving measure, the use of Trained Observers to assist with the proper donning and doffing of PPE is a critical component of CDC recommendations for PPE donning and doffing and is explicitly listed as a necessary element in PPE donning and doffing when caring for Ebola patients [29]. This strongly suggests that institutions have a responsibility to provide such Trained Observers to maximally protect healthcare workers who interact with patients suffering from highly infectious diseases. Perhaps it is time for the Department of Health and Human Services or the Centers for Medicare and Medicaid Services to mandate the use of such Trained Observers to mitigate the risks associated with improper PPE donning and doffing in future outbreaks of highly infectious diseases and overcome institutional reluctance to dedicate resources to this important but often overlooked role.

The following are the supplementary data related to this article.

Supplemental Table 1

Characteristics of Participants at Each Timepoint.

Disclosures

This research did not receive any specific grant from funding agencies in the.

public, commercial, or not-for-profit sectors.

CRediT authorship contribution statement

Cameron R. Smith: Conceptualization, Data curation, Methodology, Writing – original draft, Writing – review & editing. Terrie Vasilopoulos: Methodology, Formal analysis, Writing – original draft, Writing – review & editing. Amanda M. Frantz: Writing – review & editing. Thomas LeMaster: Data curation, Writing – review & editing. Ramon Andres Martinez: Data curation, Writing – review & editing. Amy M. Gunnett: Project administration, Data curation, Writing – review & editing. Brenda G. Fahy: Conceptualization, Data curation, Methodology, Writing – original draft, Writing – review & editing.

Declaration of Competing Interest

The authors declare no conflicts of interest.