Uncertainty and the Risk of Occupational Exposure to Severe Acute Respiratory Syndrome-Coronavirus-2.

The pandemic of 2020 is characterized by a lack of good and useful information. The dearth of evidence-based data creates anxiety and stress, causing caregivers to grasp at theoretical straws and to seek anecdotes, instead of science, for hope. As a result, there is much speculation about the therapeutic value of off-label uses of various substances for mitigating the effects of coronavirus disease 2019 (COVID-19). The world waits for hard data on prevention, detection, and treatment.

In our specialty, our job is not to treat COVID-19 but rather to sustain our practices under its pandemic waves. The information we seek is an understanding of how the presence of COVID-19 threatens our patients and ourselves in our practices. Our greatest threat is the occupational exposure to the virus that causes COVID-19: severe acute respiratory syndrome–coronavirus-2 (SC-2). The purpose of this article is to outline a model that anyone can use to estimate their risk of occupational exposure to SC-2.

In conversation, we tend to use the words “uncertainty” and “risk” synonymously. In fact, they mean different things. Uncertainty is a qualitative expression of an unknown outcome. We express uncertainty with round words, such as high, moderate, or low risk. In contrast, risk is a quantitative estimate of an unknown outcome. Risk is numeric, for example, you have 1 chance in 6 of guessing the next roll of the die. When you know the risk of an outcome, such as, occupational exposure to SC-2, you can make an informed “go–no go” decision.

When we are uncertain as to an outcome, we are inclined to make conservative decisions. During the early days of COVID-19, uncertain as to the future impact of the disease, we dumped equity holdings, saw a 30% drop in the Dow, and witnessed unprecedented “negative prices” for oil futures. As more data became available, investors could better estimate risk and began reinvesting in equity markets.

When SC-2 first invaded Washington State, I drew on my previous experience facing a new pathogen, human immunodeficiency virus (HIV). Back in those days, after much fumbling, we developed and implemented the concept of universal precautions, whereby we treated all patients as if they were HIV positive. Following that model, and being uncertain of the risk of SC-2 exposure, I recommended that our outpatient facility at the School of Dentistry treat all patients as if they were SC-2 positive by applying universal precautions. Unlike HIV, however, which is a blood-borne pathogen, SC-2 is an aerosol transmissible disease (ATD) pathogen in the same category as influenza, tuberculosis, and Ebola, and requires different considerations for protection and containment.

In preparing for managing suspected or actual COVID-19 patients, I spent time reviewing “The California Workplace Guide to Aerosol Transmissible Diseases.” This was a game changer. Treatment and containment guidelines for ATDs go far beyond universal precautions. Few, if any, academic or community-based oral surgery offices could reasonably comply with these guidelines. We would go out of business.

A couple of examples should serve to demonstrate the near impossibility of implementing these guidelines at the average oral-maxillofacial surgery practice to operate safely on patients with suspected or identified ATDs. “Engineering controls must be used in workplaces that admit, house, or provide medical services to people known or suspected to have. . . an ATD.” Examples of engineering controls include isolated treatment rooms with negative pressure environments, use of high-efficiency particulate air filtration, or ultraviolet light. These resources are not found in most offices and would require significant capital investment to upgrade to this level of isolated treatment room. Next come the decontamination procedures, which as described would require a highly trained workforce—a literal SWAT team of environmental cleaning specialists trained to decontaminate the work area between each patient.

After review of the guidelines pertaining to ATDs and decontamination of the operatory, it was apparent that treating all patients as suspected or actual COVID-19 patients was a nonstarter. However, if we could make an informed decision about which patients we could treat safely using standard precautions—surgical cap, mask, full-length gown, and eye protection and conventional room decontamination procedures—could we then continue to serve at least a subset of patients in our offices and clinics? To do this, we would need to estimate the risk of occupational exposure to SC-2.

In late March 2020, UW (University of Washington) Medicine's surgical leadership made the recommendation that patients who were asymptomatic and SC-2 test negative could be treated using standard personal protective equipment and environmental cleaning, just as we did in the pre–COVID-19 era, regardless of whether it was an aerosol-generating procedure. This seemed like a good starting point.

To estimate the risk of occupational exposure to SC-2, we need 3 parameters: 1) the prevalence of COVID-19 infection, 2) the leakage rate of the surgical mask (1 – filtration effectiveness), and 3) the false-negative rate (1 – sensitivity) of the SC-2 test. In my local environment, the prevalence of SC-2 infection among patients scheduled for an urgent operation was 0.8%. To be conservative, I estimated a 95% confidence interval and chose the upper limit, 3%, as the estimate of COVID-19 prevalence among asymptomatic patients. Next, I estimated the filtration effectiveness of the surgical mask. To be conservative, I chose a level 1 mask that is more than 95% effective at filtering particles of relevant size (>0.3 μm). As such, the leakage rate should be less than 5%. Finally, we need to know the false-negative rate for testing nasopharyngeal samples for SC-2. Locally, this is estimated to be less than 2%. By multiplying these 3 values, 3% × <5% × <2%, the estimated risk of SC-2 exposure if the patients you treat are limited to asymptomatic, SC-2 test–negative patients and you are wearing a level 1 surgical mask is 3 or fewer SC-2 exposures per 100,000 patient encounters. Is this a safe risk of occupational exposure?

The beauty of this model is that you can devise a situation-appropriate estimate of risk based on your own circumstances and local variables (Table 1 ). You can modify any of the assumptions to create a new estimate of occupational risk of SC-2 exposure. For example, the Achilles heel of screening is that asymptomatic patients may be infected with SC-2. So let us change the prevalence of SC-2 infection and see what happens. A recent study estimated that the prevalence of SC-2 infection among asymptomatic pregnant women urgently admitted to the hospital was 13%. Holding the other variables constant, the estimated risk of SC-2 exposure in this sample is 13% × <5% × <2%, or 13 or fewer SC-2 exposures per 100,000 patient encounters. This is a higher number than in my local environment (3 per 100,000), but it is still a small number. Is this safe?

Table 1

Estimation of Risk of Occupational Exposure to SC-2

Scenario	Prevalence of COVID-19, %	Mask Leakage Rate, %	False-Negative Rate of SC-2 Test, %	Estimated Risk of SC-2 Exposure
1	3	<5	<2	≤3 per 100,000
2	13	<1	<2	≤13 per 100,000
3	10	<1	Test not available	≤100 per 100,000

Abbreviations: COVID-19, coronavirus disease 2019; SC-2, severe acute respiratory syndrome–coronavirus-2.

What if testing is not available in your community? What is the risk of SC-2 exposure if I limit my patients to being asymptomatic and I put on an N95 mask (>99% effective in filtering particles of relevant size)? Estimates of COVID-19 among symptomatic patients hover around 10%. Therefore, the prevalence among asymptomatic patients must be lower. To be conservative, however, let us set the prevalence at 10% and mask filtration leakage rate at less than 1%. The estimate for SC-2 exposure in this model is 10% × <1%, or 100 or fewer SC-2 exposures per 100,000 patient encounters. Is this safe?

Other considerations when estimating the consequences of occupational exposure to SC-2 include that 1) not all exposures result in COVID-19 and 2) if a patient is infected with SC-2, the morbidity of COVID-19 ranges from asymptomatic to death. Death is serious and scary, especially when its risk is uncertain. So let us see if we can make an informed estimate of the risk of death in my local environment. Estimates of death from COVID-19 are about 5%. Again, the value of the model is that you can choose what you think the best estimate of risk is in your local community. So if I assume that every SC-2 exposure results in COVID-19, the prevalence of disease is 3%, the mask leakage rate is less than 5%, the false-negative rate is less than 2%, and the chance of death is 5%. Thus, the overall risk of death is 1.5 deaths per 1,000,000 patient exposures. Is this safe?

After spending some time working with this model and evaluating the recommendations of UW Medicine's surgical leadership, I recommended that we limit procedural care to asymptomatic patients who were SC-2 test negative. We limit in-person consultations to asymptomatic patients while wearing masks, both provider and patient. I concluded that it was safe to deliver care using standard, pre–COVID-era personal protective equipment and environmental precautions. That being said, some people are uncomfortable with any risk of exposure. This model will always return a value between 0 (no exposure) and 1 (certain exposure). How one's actions are guided by risk is a personal choice.

As politicians debate what level of business activity can be safely sustained during a pandemic, I take comfort only in quantitative solutions based on available evidence. Being able to “crunch the numbers” brings me some level of comfort and control during this time of uncertainty. I encourage you to use data from your local environment or community to compute an estimate for your risk of SC-2 occupational exposure and make informed decisions as you plan a return to practice.