Abstract
The attempt by Hade 1 to estimate the risk of an anaesthetist contracting the SARS-CoV-2 virus from a patient results in risk estimates that are unrealistically high.
His starting formula claims risk is the product of (A) the probability of patient carrying SARS-CoV-2 and (B) the probability of transmission. This seems reasonable, as does his comment that in a confirmed case of COVID-19, the value of A is 1 (100%).
For the value of B, Hade cites evidence that the effect of personal protective equipment is to reduce the probability of transmission to 0.22 (22%). 2 Without disputing this, it is clear the product of A and B by Hade’s method is therefore 1 × 0.22, which is 0.22 (22%, or ∼1 in 5). Hade does not state this explicitly, focusing instead on the population of patients that test negative for SARS-CoV-2.
It is already clear that Hade’s estimate of a one in five risk of an anaesthetist contracting SARS-CoV-2 from a confirmed case after a single aerosol-generating procedure (AGP) is incorrect and far too high. The best data are probably from the intubateCovid registry, which reports that one in 10 anaesthetists were symptomatic after 3 weeks of follow-up after an AGP (only one in 30 testing positive, even after several AGPs). 3
As a separate concern, Hade seems to conflate ‘incidence’ with ‘prevalence’ of disease in the population. Hade presents the calculations for ‘incidence’, which is the number of new cases per head of population over a certain period. The incidence only tells us the potential rate of growth of disease. What is needed for his own method is in fact prevalence. This is the net result of the new cases (incidence) and those cases recovering or dying. This is a relatively minor point compared with the issue referred to above.
It is, however, commendable that Hade offers a binomial model to predict the cumulative risk (his last equation). I refer to our recent modelling, 4 based on intubateCovid data. Also using binomial probabilities, we offer a more realistic estimate of risk after AGP in both SARS-CoV-2 test-positive and negative patients, several orders of magnitude lower than Hade’s estimates (we estimate one in 240 and one in 480,000 as respectively the upper and lower bounds).
