Blog | Friday, April 17, 2020

A face shield strategy to reduce COVID-19 nosocomial transmission

In previous blog posts, I mentioned the implementation of face shields to prevent nosocomial COVID-19 infection. Over the past few days, I have received many questions from people across the country, so I thought it would be useful to pull everything together in a single post and add some details.


As we began to prepare for the pandemic, we assessed our inventory of personal protective equipment (PPE). However, knowing current inventory levels alone is not useful. It's deceiving when you see PPE inventory levels of several hundred thousand items. How could we ever run out, right? This is why your inventory levels must be evaluated in the context of normal demand. Your supply chain folks should be able to tell you how many PPE items are normally used on a daily or weekly basis.

Next, you need to determine your expected demand for the outbreak. There is no right answer here as there are too many unknowns. You'll just need to make an educated guess. We determined that our critical level of each PPE item was 12 weeks (84 days) at four times normal demand, although you could argue that this is an underestimate. Next our supply chain group developed an interactive spreadsheet with each row being a PPE item, and columns showing current inventory; normal demand per day; and days of stock at normal demand, two times normal demand, and four times normal demand. The final column (days of stock at four times normal demand) is color coded as follows: red ≤84 days, yellow 85-111 days, green ≥112 days (16 weeks). Once this is done, you will likely be surprised to find that what seemed like an abundance is really not so. When evaluating your levels, you also need to consider that some items are on allocation and you can expect to receive periodic shipments, while others are simply stocked out with no promise of future deliveries.

After reviewing all of the above, the most worrisome thing for us was an inadequate supply of face masks (<84 days at four times normal usage). My biggest fear was that we would overuse them early in the outbreak when few COVID-19 patients are hospitalized, then have none after the surge of COVID-19 inpatients arrived. Many hospitals had extended the use of face masks beyond a single patient encounter, which is a reasonable decision in this time of shortage, but we know that face masks lose their effectiveness when they get wet. Some began to use cloth masks, which is also a problem. So I began to wonder whether face masks could be replaced by face shields.

Why face shields?
• They provide greater facial surface area coverage than face masks by protecting all the facial mucosal surfaces from infectious droplets.
• Given that the eyes are protected, we can eliminate the need for goggles when a face mask is worn. And we know that healthcare workers are really bad at wearing eye protection.
• They prevent you from touching your face. One of the major drawbacks of face masks is that some people will touch their faces even more to adjust the mask and this poses a risk for autoinoculation by contaminated hands.
• Face shields are durable, can be cleaned after use, and reused repeatedly.
• Many people (myself included) find face shields more comfortable than face masks.
• Communication is better with shields than with face masks as your face is visible to patients and coworkers.
• If all of your health care workers are shielded, social distancing becomes less important.
• And importantly, this is a device that is diversified across other industries. There is greater availability since the medical supply chain is so stressed at this time.

Are there any disadvantages compared to face masks? The only one I can think of is the possibility of a droplet coming in an upward trajectory going under the bottom edge of the shield. Although the probability of this is small, this can be minimized by having the shielded healthcare worker flex their neck when standing over the patient (for example, when performing a physical exam), bringing the bottom edge of the shield closer to the HCW's torso. Moreover, when doing a procedure that normally requires a face mask, we recommend that a mask be worn under the shield anyway.

A few people have asked what is the evidence that face shields can replace face masks, and those particularly inclined toward methodolatry (the profane worship of the randomized clinical trial as the only valid method of investigation) continue to demand that face shields not replace face masks. Do I have evidence? No. To me, this is just plain common sense—we have a product that is reusable, cleanable, covers more of your face, decreases the risk of autoinoculation, and keeps us from burning through our mask supply. We have hospitals in the U.S. where nurses are using bandanas to protect themselves. In this extraordinary time, I can live without a clinical trial.


We quickly found that face shields marketed for medical uses were stocked out. One of our pharmacists went to a local hardware store and found shields used for grinding. We then explored vendors that supply hardware and agricultural products (see more here on the purchasing process). In addition, the University of Wisconsin has a great website that includes diagrams for construction of shields, and Johns Hopkins has a “recipe” available that can be used to create 50,000 shields. We placed the Wisconsin diagrams on the hospital website and several manufacturing firms responded that they could fabricate them for us. In addition, we have had some designed and produced by people interested in 3-D printing. One of our physicians, modified the Johns Hopkins' information and has her kids at home making shields. We also placed on our hospital's website a request for donation of face shields that people have at home, and we received many donations. It has really been a community effort. As our supply of shields grows daily, we deploy them throughout the medical center. At this point, the shields are handed off from one worker to the next as their shift ends, but our ultimate goal is for every person to have one for their personal use.

Here are the instructions we give to our staff on when to use face shields:
• Wear the face shield with every patient encounter (COVID-19 and non-COVID-19 patients). Think of it as a new component of standard precautions—every patient, every time.
• For non-COVID-19 patients, wear the face shield without a face mask.
• For COVID-19 patients (confirmed or suspect), wear the face shield over a face mask. If an aerosol generating procedure is being performed, wear the face shield over an N95 respirator.
• If you are performing a procedure for which a face mask is normally worn (e.g., lumbar puncture, central line insertion, joint injection, procedure in the OR, etc.), wear the face shield over a face mask.

To help introduce the concept to our workforce, we produced this video:

We still have some details to iron out, such as the best product to clean the shield, since some products damage polycarbonate. We have had some of the shields break, so fabricating replacement shields to repair them is ongoing.

Lastly, I have had a number of people who want to purchase their own shield ask me which one I recommend. Over the past two weeks, I have tried many models and have become a face shield connoisseur. If you want to buy your own, I think the best is the Uvex Bionic S8510 made by Honeywell. It is more sturdy than many other models, provides greater facial coverage (extends laterally on the face to your ears), and is comfortable. What really sets it apart is the V-shaped, downward projecting bottom border of the shield. This allows it to sit close to your upper torso, minimizing the risk of upward trajectory droplets. As an added bonus, from a sartorial standpoint, you'll look top-notch in this one! It's the kingdaddy!

This is probably more than you ever wanted to know about face shields. But it's a crazy time and we hospital epidemiologists are doing things I could have never imagined just a month ago. The bottom line is that by employing face shields we are able to protect our workforce while extending the duration of time that we will have face masks available.

Stay safe and be well!

Michael B. Edmond, MD, FACP, is a hospital epidemiologist in Iowa City, IA, with a focus on improving the quality and safety of health care, and sees patients in the inpatient and outpatient settings. This post originally appeared at the blog Controversies in Hospital Infection Prevention.