Science is a demanding enterprise. One of its core demands is humility. In the scientific sense, humility is not about geniality or deference. It is about submission to the discipline of the scientific method. It is about defending yourself from your own biases and assumptions that can obscure deeper, non-obvious truths that the scientific method is designed to reveal.
Scientific humility is like a Ulysses contract, a voluntary agreement to be restrained from the seductive influence of seemingly persuasive, but potentially misleading information. When Ulysses directed his crew to lash him to the mast of his ship and keep him tied up when he fell under the spell of the Sirens — no matter how much he protested — he was recognizing his own susceptibility to false promise.
The president’s “really good feeling” about hydroxychloroquine (HQC), and his reckless propaganda about the drug’s potential value for COVID-19, present a textbook argument for the central role of humility, or the recognition of our susceptibility to false promise, in science. In clinical research, humility is baked into the logic of clinical trial design, including the data safety and monitoring boards that perform independent reviews of accumulating clinical trial data precisely to defend clinical scientists from their own susceptibilities.
There are thousands of examples of the seductive power of false promise in science. One that I often use in my teaching about the regulation of research with human beings is the introduction of lithium chloride as a table salt substitute in the 1930s and 1940s when sodium chloride — regular table salt —was first implicated in a number of metabolic disorders and hypertension. How different could the two salts be? Someone almost certainly had a “very good feeling” about lithium chloride. But although it seemed safe enough at moderate doses, it turned out to have deadly toxicity at doses just slightly higher than a few shakes of a salt-shaker. Lots of people died.
A recent review of “what every clinician should know” about the use of HQC for COVID-19 in the Annals of Internal Medicine outlines serious and sometimes deadly cardiac side effects of HQC, even in its approved use for malaria. The review also emphasizes that HQC is a highly effective drug, “a cornerstone of therapy”, for lupus, an auto-immune disease. It goes on to say that “landmark clinical trials have demonstrated that the withdrawal of HQC can lead to flares of disease, including life-threatening manifestations, such a lupus nephritis.” The president’s relentless advocacy for HQC, over the muted protests of his own COVID-19 Task Force, have already resulted in shortages for lupus patients and will inevitably lead to bad outcomes for some of them.
The president’s other persistent idea that there is “nothing to lose” by trying unproven drugs on people just because they are critically ill is simply ludicrous. Recent data from the Intensive Care National Audit and Research Centre in the United Kingdom show a 50/50 survival rate for COVID-19 patients admitted to Intensive Care Units (ICUs) without severe co-morbidities. Let’s imagine that a very sick person in the ICU is given HQC and gets better. There is no way to know whether they would have gotten better without the drug, i.e., whether they would have been part of the 50 percent to survive to be discharged. And, more importantly, if someone is given HQC in the ICU and then dies, there is no way to know whether they might have recovered if they had not been given the drug, i.e., whether the drug might have contributed to their death. We call this inability to know the alternative outcome the “counter-factual” problem.
Careful scientific observation of individual patients, alone, does not solve the counter-factual problem in the way the president supposes. And, for precisely this reason, the randomized, double-blinded clinical trial (RCT) has arisen as the best way we have yet devised to determine the safety and efficacy of new, or, in the case of HQC, re-purposed drugs. “Control” is the key. Patients receiving the HQC need to be compared to those who do not receive the drug. HQC would be added to the “standard” supportive therapy for COVID-19 for one group of patients, selected at random, while another group would be given a placebo in addition to standard therapy. And neither the doctors caring for the patients nor the researchers running the trial would know which patient got HQC or placebo until after the trial is over.
Properly designed and rigorously executed RCTs would solve the most fundamental problem facing the evaluation of HQC for COVID-19, by imposing humility on the cheerleaders for the drug who have been unable, for whatever reasons, to arrive at that humility independently.
Comparing patients with similar clinical profiles, at the same dose and duration of therapy, with similar patients receiving placebo, would allow us to see if a single patient’s outcome — good or bad — was an outlier, or part of a clear and measurable pattern of a drug’s action.
The world’s main registration site for clinical trials — www.clinicaltrials.gov — currently lists 68 trials of HQC, alone or in combination with other drugs, for the treatment of COVID-19. It is a virtual certainty that the number of trials would have been lower without the president’s promotional appeals. On April 10, Sanofi, the company that manufactures HQC, announced that it would donate 100 million doses across 50 countries. But these donations will be made only “if the ongoing clinical studies demonstrate its safety and efficacy in COVID-19 patients.” All fair-minded people should hope that HQC is effective for COVID-19, but be prepared for disappointment.
The president claims his apparent goal of promoting the widespread, unregulated clinical use of HQC for COVID-19, outside the structure of RCTs, is based on his “common sense.” But such an approach would almost certainly cause harm and even hasten the death of many patients, even if it did help others. It would also simultaneously deprive us of the kind of evidence that doctors and regulators need in order to fulfill their ethical obligations to patients. This evidence can only come from RCTs.
Scientific humility — the Ulysses contract — requires us to reject the president’s “common sense” siren song for HQC and rely, instead, on the accumulated wisdom of clinical research. The opportunity costs from the lost evidence would be simply unconscionable.
Jim Lavery is professor and Conrad N. Hilton Chair in Global Health Ethics at the Hubert Department of Global Health, Rollins School of Public Health, and Faculty, Center for Ethics at Emory University in Atlanta, Ga. Follow him on Twitter @LaveryJim