Workers in healthcare and nonhealthcare settings are at risk to acquire severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) and other respiratory viruses. 1–Reference Porter, Ramaswamy, Koloski, Castrodale and McLaughlin5 To reduce the risk, the Centers for Disease Control and Prevention (CDC) has recommended several measures, including the use of face coverings, improving building ventilation, and maintaining physical distancing of ≥1.8 m between individuals. 1 In situations where distancing cannot be maintained, it is recommended that barriers such as plexiglass shields be installed. 1,Reference Herstein, Degarege and Stover4,Reference Porter, Ramaswamy, Koloski, Castrodale and McLaughlin5 If an opening in the barrier is necessary, it is recommended that it should be as small as possible. Although barriers are commonly used in workplaces, limited information is available on their effectiveness in decreasing the risk for viral transmission. Therefore, we conducted simulations to assess the effectiveness of plexiglass barriers in reducing exposure to an aerosolized benign virus, and we assessed the characteristics of barriers used in retail businesses and a hospital.
Methods
Characteristics of barriers used in a hospital and in retail businesses
For a convenience sample of retail businesses and 1 hospital, we assessed the presence and characteristics of barriers used in situations where workers and customers or coworkers were unable maintain physical distancing. The height and width were measured, and the presence of openings was assessed.
Simulations using bacteriophage MS2
We conducted simulations to assess the effectiveness of barriers in reducing exposure to an aerosolized benign virus in a 140 m3 room with ˜6 air changes per hour. A mannequin wearing a face mask was positioned at a table. For each simulation, an Aerogen Solo nebulizer (Aerogen, Galway, Ireland) was used to release 1 mL of droplets containing 108 plaque-forming units (PFU) of bacteriophage MS2 over 3 minutes. Moreover, 70%–80% of particles generated by the Aerogen Solo are ≤5 µm. Reference Bowling, O’Malley, Klimstra, Hartman and Reed6 The nebulizer was positioned at the height of the mannequin’s head 91 cm (3 feet) or 241 cm (7.9 feet) away and was directed toward the mannequin. Supplemental Figure 1 shows the setup with the nebulizer 91 cm from the mannequin and the barrier configurations.
Simulations were conducted with no barrier and with 3 barrier configurations: (1) plexiglass barrier (width 61 cm and height 76 cm with a 40 × 15 cm or 20 × 15 cm opening; Luxor, Egypt) placed on the table 30 cm from the nebulizer and 61 cm from the mannequin; (2) Plexiglas barrier with opening covered; and (3) Plexiglass barrier with opening covered and horizontal extension doubling the width to 122 cm. For configurations 1 and 2, simulations were conducted with a desktop portable high-efficiency particulate air (HEPA) cleaner (HPA020B Tabletop Air Purifier, Honeywell, Charlotte, NC) positioned on the tabletop between the opening and mannequin.
To assess the effectiveness of the barriers, surface samples were collected using premoistened cotton-tipped swabs. The sites sampled included a 5 × 5-cm area on the tabletop and the face and face mask as well as the neck and chest of the mannequin. Quantitative cultures were performed for bacteriophage MS2. Reference Alhmidi, Jones, Pearlmutter, Cadnum, Silva and Donskey7 The simulations were repeated in duplicate or triplicate for each configuration.
Simulations with larger particle size and visualization of contamination
Additional simulations were conducted using fluorescent polyethylene microspheres (Cospheric, Santa Barbara, CA) with 212–250 µm diameter expelled toward the mannequin with a nebulizer with plastisol bulb (3M, Saint Paul, MN) and ultraviolet (UV) blue water soluble dye (DayGlo, Cleveland, OH) 0.1 mg/mL dispensed using a Preval sprayer aerosol-based spray system (Nakoma Products, Bridgeview, IL). The microspheres were chosen to be similar in size to larger (100–1,000 µm) respiratory droplets released with coughing. Reference Bourouiba8 A black light UV flashlight (uvBeast) was used to visualize the microspheres and dye.
Data analysis
For a composite of contamination on the tabletop, face and face mask, and neck and chest, we estimated a regression model comparing log10PFU recovered for the barrier configurations in comparison to no barrier, adjusting for sample site. Estimated marginal means and 95% confidence intervals were calculated for each group. Data were analyzed using R version 4.0.3 software (The R Foundation for Statistical Computing, Vienna, Austria) with functions from estimated marginal means (emmeans).
Results
Characteristics of barriers used in retail businesses and healthcare facilities
Of 30 retail businesses surveyed, 29 (97%) had barriers between workers and customers and 3 (10%) had barriers between coworkers. All barriers were made of plexiglass and extended to ≥1.8 m from the floor. The width ranged from 45 to 180 cm. Also, 22 (73.3%) had openings with a wide range of sizes. Supplemental Figure 2 shows examples of barriers. In the hospital, barriers were placed between personnel and patients or visitors in the cafeteria and lobby. In multiple areas, workspaces were spaced <1.8 m apart with no barrier.
Simulations using bacteriophage MS2
With no barrier present, bacteriophage MS2 released at 91 or 241 cm from the mannequin resulted in recovery from all 3 culture sites, including tabletop, face and face mask, and neck and chest (Fig. 1). In comparison to no barrier, MS2 contamination after release 91 or 241 cm from the mannequin was not significantly reduced when the barriers had openings of 40 × 15 cm or 20 × 15 cm (P > .17), but significant reductions occurred for all other groups (P < .01), including when the opening was present but a portable air cleaner was used. No contamination was detected for the extended barrier and air cleaner groups when aerosol was released at 91 cm.
Fig. 1. Contamination with bacteriophage MS2 during simulations with no barrier versus different barrier configurations with or without a portable high-efficiency particulate air (HEPA) cleaner. Aerosol containing bacteriophage MS2 was released 91 cm (A) or 241 cm (B) away from a mannequin that was positioned 61 cm behind the barriers. Marginal means for a composite of tabletop, face and face mask, and neck and chest samples with 95% confidence intervals are shown.
Simulations with larger particle size and visualization of contamination
Ultraviolet blue dye was visualized on the mannequin and tabletop when there was no barrier and when there was an opening (Fig. 2). With the opening closed, dye was visualized on the tabletop but not the mannequin. No dye was visualized on the tabletop or mannequin with the horizontally extended barrier with the opening closed. The fluorescent microspheres were detected on the mannequin (≥7 microspheres) and tabletop (≥151 microspheres) with no barrier and with the opening but not with the other configurations.
Fig. 2. Contamination with UV blue water soluble dye during simulations with different barrier configurations. Aerosol containing dye was released 91 cm away from a mannequin, which was positioned 61 cm behind the barriers.
Discussion
Plexiglass barriers with no openings were effective in reducing exposure to an aerosolized benign virus, but barriers with openings were ineffective. In Cleveland area retail stores, plexiglass barriers were commonly used, but most had openings for transfer of items that could substantially increase the risk for contamination. These results provide support for the recommendation that barriers be installed in workplace situations where distancing cannot be maintained but highlight the importance of barrier design. 1 Our findings also suggest that barriers may be underutilized in some settings because hospital personnel often worked in close quarters without barriers.
The CDC recommends that openings in barriers should be as small as possible. 1 However, openings were often quite large in retail businesses, and even relatively small openings (20 × 15 cm) resulted in increased risk for contamination. Options to address this issue might include eliminating openings or designing adjustable openings that remain closed unless items are being transferred. In addition, when an opening was present, a desktop portable air cleaner positioned between the barrier and mannequin significantly reduced contamination. Lindsley et al Reference Lindsley, Derk and Coyle9 similarly found that portable air cleaners reduced exposure to simulated aerosol particles.
Our study has several limitations. Simulations cannot replicate all factors that affect transmission in real-world settings. For example, air currents created by fans or heating/cooling systems may facilitate dispersal of viral particles and warm air could carry particles above barriers. Reference Jones, Chan and Zabarsky10 Only 2 sizes of barrier openings were tested, and the room had relatively good ventilation. Finally, we did not assess the efficacy of barriers in comparison to or in combination with measures such as face masks, increased ventilation rates, and upper room ultraviolet light. Reference Jones, Chan and Zabarsky10
Supplementary material
To view supplementary material for this article, please visit https://doi.org/10.1017/ice.2021.383
Acknowledgments
Financial support
This work was supported by a merit review grant (no. CX001848) from the Department of Veterans’ Affairs to C.J.D.
Conflicts of interest
C.J.D has received research grants from Clorox, Pfizer, and PDI. All other authors report no conflicts of interest relevant to this article.
