By SUSAN ADELMAN, MD
Since the beginning of the COVID epidemic, most physicians have been struggling to do the right thing when it comes to treatment and prevention. Part of this was because, in the beginning, a terrifying complex of symptoms hit our patients fast, with high lethality, high transmissibility and distressing urgency. We worked hard, under pressure, to diagnose, to figure out immediate treatments, to devise prevention strategies, to understand new vaccines, and, eventually, boosters. Now we have a great deal of information, but most of us in the medical profession still are hard-pressed to be confident in our recommendations about vaccines and boosters for all age groups. Meanwhile, some politicians, schools and employers still are trying to protect their constituents and employees with well-meaning but scientifically shaky mandates.
It is now time for the medical profession to stop, take a deep breath, and begin addressing COVID as it would address any other diagnostic or therapeutic dilemma, using evidence-based medicine. This is a systematized method of grading evidence to assess, say, the effect of an intervention or the results of clinical trials, usually using meta analysis of multiple studies. The quality of the evidence then is rated High, Moderate, Low or Very low. As an example, moderate might mean that further studies could be useful, and low might mean that future studies are quite likely to change the rating.
The phrase “evidence-based medicine” originated in the 1990s, and it was developed to make clinical measurements more objective. From this, a number of statistical instruments developed, including measures of the internal validity of a study design and of the external validity, which means how confidently the initial results can be extrapolated to the general population. Then, along came the creation of national and international databases, and international collaborations. Much of this is explained in an article in the Annals of Medicine and Surgery, Nov. 2021, “Practical guide for the use of medical evidence in scientific publication: Recommendations for the medical student: Narrative Review,” with multiple authorship.
Some classifications are more complex than others. One, from Oxford University, uses a list from 1 to 5, but with sub-levels a,b, and c. Others are simpler. Of interest, more than one table shown in the article explores and rates biases that lead to medical errors. In addition, the article gives a system for rating fallacies that may cause errors in judgment. Imagine how helpful these analyses could be when a doctor searches multiple studies to find out whether a patient who came in that afternoon was at increased risk for cardiac complications from a specific COVID vaccine. What about assessing whether vaccines or boosters should be required for small children when essentially all studies suggest that they are unnecessary in this age group, but certain authorities insist that they be given? The doctor is more than willing to give the right advice, but what if different studies come to conflicting conclusions? What to do?
Here, instead of looking to politicians and employers, an expert consortium should be convened—within the medical profession—to subject all known studies to continuing high-quality analysis. Imagine how much better doctors and patients would feel about whether to take—or recommend—the latest vaccine or booster! Imagine how much better I would feel when friends and family ask me to advise them on the subject. After all, I am just a pediatric surgeon trying to do my best to keep up with the relevant literature!
Anyhow, it never was the tradition in our society to have our government hand down mandates to take a vaccine, nor was it ever our tradition for the medical profession to criminalize medical decisions that result from educated guesses or judgment calls, as long as they fall within the limits of normal practice. It certainly is within the norm for expert panels to evaluate evidence, and that is what we should do for COVID, COVID vaccines and boosters. Now what we need is to systematize what has been, up to now, an entirely ad hoc way of handling the subject.