Just Do IT!: Hydroxychloroquine

Really?

Why is it the strongest, loudest advocates of unproven, proposed drug therapies are never the first in line for a trial or study? The president is telling the nation, “What do you have to lose? Take it” (here). Even Dr. Hannity, prescribing for only himself said, “I would based on all I know, Dr. Hannity for Dr. Hannity, I'm my only patient, I would absolutely use it if I got this” (here). And known truth teller and sage for the ages, Rudy Gulliani, retweeted a conservative activist’s view, “who falsely called hydroxychloroquine 100% effective at treating COVID-19” (here).

Who doesn’t hope hydroxychloroquine works. The drug is here now. Those most seriously affected by the coronavirus, hospitalized and especially those on ventilators, could be pulled back from the brink—if the drug works.

And therein lies the question: Does hydroxychloroquine reduce the adverse effects of the coronavirus? 

The gold standard for answering this question has always been a properly designed experimental trial at appropriate scale. You don’t give it to a small group of ill persons and reach any valid scientific conclusion.

In graduate school they taught us four rules about experimental designs and studies. First, and foremost, the study had to be a true experiment. Some people get the drug. Some people get a placebo instead. Observational studies—studies where there is no control group—are just observations and not true scientific evidence of anything.

Second, never confuse a temporal result with a causal (i.e., as in cause and effect) one. “Coincidence” happens even when people are disinclined to believe in coincidence. There’s also the problem of wanting something to work so badly you become convinced it does when it has not. Like, a person’s fever breaks sooner than others, or the symptoms were not as severe, and so on.

Early wisdom about the coronavirus stated the illness avoided the young. This was an older person’s disease. And, for older seniors, the coronavirus was a guaranteed death sentence. Subsequently, there were credible reports of infant and toddler deaths followed by the deaths of some spring break revelers. Recently, there was a report of a 100+ year old survivor being discharged from a hospital ICU. Early reports are just that—early, premature.

Third, does your drug experiment measure the right things? Have possible confounding “things” been anticipated. Is there a food or drug allergy? Was any other medication or treatment administered? Was an antibiotic also involved? If so, was the antibiotic toxic for a patient’s underlying health status?

Fourth, are the results of the study applicable to the larger population of coronavirus patients. This is why drug treatment trials are complex to design and conduct and often require a large number of participants. It’s also why multiple studies at different locations are designed to replicate the outcomes of other drug and treatment trials.

There are news reports of practicing physicians stating their willingness to self-prescribe, or prescribe for their families, these unproven treatments. Equipped with only their practical experience and a prescription pad and a license to use approved drugs for off-label purposes, these seemingly heroic figures are prepared to violate all four basic rules of experimental designs. AND, to do so on very small groups—even a sample size of one. They don’t control for confounding influences. Their results are unlikely to apply to the general population. It isn’t even clear the drug therapy was the reason a patient improves.

Worse yet, patients who trust these “experimenters” do so at the risk of all sorts of unanticipated interactions and outcomes. It seems a physician who dabbles in ultra-small-scale experimental drug and treatment trials is like the lawyer who has himself for a client.

Not a good thing.

Copyright 2020, Howard D. Weiner