International Registry of Otolaryngologist‒Head and Neck Surgeons with COVID-19.

Abstract

I read the article by Sowerby and colleagues with great interest and commend the authors for ambitiously attempting to quantitate the impact of the coronavirus disease-2019 (COVID-19) pandemic on otolaryngologist‒head and neck surgeons.1 As of this writing, we are entering the fifth month of the pandemic, with >16 million infections worldwide, many of these diagnosedin health-care workers (HCWs). As Sowerby et al point out, clinical care in otolaryngology‒head and neck surgery requires intimate patient contact and instrumenting of the aerodigestive tract, which is known to harbor the severe acute respiratory syndrome‒coronavirus-2 (SARS-CoV-2). Although this setting could potentially place our specialty at greater risk of infection, one could argue that, with appropriate mitigation and use of personal protective equipment, this concern is largely theoretical and that evidence to support such an hypothesis is lacking. Patel and colleagues were the first to formally ring this “alarm bell” and did a service to our profession by bringing this topic into focus, both within the scientific literature and among administrators and epidemiologists within our respective practices and hospitals.2 Nonetheless, that correspondence was based largely on anecdotal evidence, and potentially raised more questions than answers. We are living through an unprecedented pandemic. In these unsettled times, it is essential that we as a profession rely on the best available evidence to make decisions for our patients and for ourselves. When such data are not available, we must actively seek it, relying on hypothesis-driven research, statistics, and expert consensus. My concern is that the report by Sowerby et al may again leave us with more questions than answers. First, the authors reference reports of higher infection rates among otolaryngologists in Wuhan and Italy. However, on close inspection, the reference to infections in China is from a media report of largely anecdotal observations from a single individual at one Chinese hospital. None of the references mention higher rates of infection among otolaryngologists in Italy. Second, the true aerosol-generating potential of otolaryngology practice and procedures remains controversial,3-7 as do early reports of potential super-spreader events putatively attributed to endoscopic endonasal procedures.8 Finally, the authors highlight 6 “instructional cases” and outline “takeaway lessons,” based on the description of these events. However, at least half of the example cases occurred in middle to late March, when little was known about COVID-19 viral transmission and country-specific recommendations for personal protective equipment (PPE) utilization had yet to be implemented and, consequently, these reported experiences may not be as applicable to the point of the pandemic in which we find ourselves today. Furthermore, details of each case are vague and the methodology for verifying patient-to-provider transmission, or even formal COVID-19 testing, is likewise ill-defined. Nonetheless, data specific to otolaryngologists are clearly lacking and the reasoning for pursuing this query is obvious and badly needed. Thus, the authors sought to collect “real data on the potential risk for our specialty, which procedures may be higher risk, and which PPE may or may not be protective.” To do this, they contacted members of the Young Otolaryngologists‒International Federation of Otolaryngologic Societies, who then served as intermediaries to collect data from their individual countries. Data were collected through e-mail surveys and anonymous questionnaires. Additional data were collected from “the grey literature and social media channels.” Unfortunately, the nature of the study design essentially assures an unrepresentative sample and is at very high risk of selection bias. Likewise, early published statements by some of the authors with respect to the subject matter at hand raises the risk of confirmation bias in both data gathering and interpretation.2 Given the lack of data specific to the otolaryngologist‒head and neck surgeon, we can at least start by analyzing available data from health-care workers (HCWs) collectively and for those that may be on “the front lines” or with routine interactions with COVID-19 patients. Lombardi et al examined infection rates among 1573 HCWs in Milan, Lombardy, Italy, between late February and late March 2020, at the time a major epicenter of the SARS-CoV-2 virus.9 Of the tests, 8.8% were positive, but it should be noted that two thirds of HCWs tested were asymptomatic. No statistically significant difference in infection rate was identified between different occupations, among them HCWs with no clinical contact. In a study cited by the authors, Lai et al assessed infection among HCWs at a tertiary hospital in Wuhan, China, between January 1 and February 9, 2020.10 The infection rate was only 1.1% and, interestingly, there was no difference in the infection rate between frontline HCWs and HCWs in areas with no patient contact. A higher incidence of infection among HCWs was identified in the United Kingdom, but again no difference in infection was noted between patient-facing, non‒patient-facing, and nonclinical roles, suggesting that nosocomial viral transmission from patients to HCWs was not a significant mode of spread.11 Similarly, an analysis of HCWs in The Netherlands showed that, of 86 of 9705 HCWs infected with COVID-19 in 2 hospitals, only 3% reported exposure to an inpatient with a known diagnosis.12 Studies by this group also did not support widespread nosocomial infection of HCWs and instead suggested acquisition in the community as the primary source.12, 13 Ultimately, the key issue is whether health-care occupations put individuals at greater risk for COVID-19 infection and, if so, whether this relative risk can be reduced through appropriate personal protection. Schwartz et al sought to answer this question in a large cross-sectional study in Canada that included 4230 HCWs diagnosed with COVID-19.14 This amounted to 17.5% of cases, and HCWs were diagnosed at a rate 5.5-fold higher than the general population. These numbers alone are at first frightening and, in isolation, would suggest an unacceptable risk of infection among HCWs. However, nosocomial infection could only be confirmed in 3.1% of cases, and HCWs were more likely to be diagnosed when asymptomatic or presymptomatic. Mortality rates were also vastly different (0.2% among HCWs vs 10.5% among non-HCWs), suggesting a substantial difference in populations and testing frequency. The authors noted that testing bias was likely a major contributor to the difference in infection rate between HCWs and the general population and found that the discrepancy decreased as testing capacity improved. HCWs have been prioritized for testing since the early weeks of the pandemic, and differences in age and undertesting in the general population further add to these reported discrepancies. Serologic testing from Lombardy, Italy, supports this bias, with 23% of HCWs testing positive for COVID-19 antibodies, compared with 62% of the general population.15 These examples do not clarify the relative risk of COVID-19 infection among otolaryngologists, but do point to the multiple sources of bias that can result from reporting individual cases of infection or tallies that do not include a denominator. Sowerby et al concluded that their findings “confirm a significant incidence of COVID-19 within our field.” However, incidence is a measure of the probability of occurrence of a given medical condition (within a given period of time) and requires that one know both the total number of cases and the total population at risk. Given the inability of the authors to objectively determine either of these, this statement is simply not supported. None of this is to say that we as a profession should ignore the risk of infection as we care for patients, resume activities in the operating room, or carry out our daily lives. We must continue to follow best practice guidelines, utilize PPE, and reduce relative risk. However, we should avoid relying on anecdotal evidence whenever possible and ensure that the decisions we make are evidence-based and ultimately appropriate both for us and for our patients.