In the midst of the COVID-19 crisis, much has been said, often disparagingly, about the insistence by physicians and scientists for the use of randomized controlled trials (RCTs.) Why, ask the naysayers, can’t we simply start using treatments, such as hydroxychloroquine, which show early promise? An RCT is a means of testing a hypothesis, such drug efficacy, while randomly selecting a group for intervention (treatment) and an equivalent group of controls, usually an inert placebo or an alternate drug. By establishing equivalent study groups, this method cancels out potential confounding factors, which inadvertently distort the results. The current debate about the use of chloroquine and hydroxychloroquine illustrates the need for RCTs to study these drugs.
The current interest in chloroquines for COVID-19 rests entirely upon a combination of theoretical considerations and uncontrolled clinical observations. While long used for treatment of malaria and lupus, these agents can have significant adverse effects, including sudden death from cardiac arrhythmias. Without valid and well-designed RCTs no one can tell whether a favorable outcome reflects the natural course of the illness (in which the majority of patients recover on their own) or the response is due to the introduction of the drug. We will also not be able to determine whether a drug is useful early in the course of COVID-19 or later, when the disease becomes more severe.
Because clinicians are not immune to drawing erroneous conclusions about drug efficacy, we now practice “evidence-based medicine.” This is simply a way of saying that our therapeutic decisions are informed by the best available scientific data. The evidence-based approach has revamped our use of many medications, procedures, and surgical practices and even routine cancer screening protocols.
When evaluated in this rigorous manner, drugs that “seem to work” may later be found to be ineffective or frankly dangerous. Back in the 1980s, for example, cardiologists had been routinely prescribing drugs to suppress abnormal heart beats in patients suffering recent heart attacks. Since premature ventricular contractions (PVCs) were associated with higher mortality following myocardial infarction. It was widely assumed that “antiarrhythmic” drugs which suppress PVCs would lead to improved survival. However, in 1987 a large international “Cardiac Arrhythmia Suppression Trial” was initiated. After less than two years, the trial was abruptly terminated. Two of the drugs, flecanide and encanide, were associated with unexpected and substantially higher mortality compared with placebo. The use of these agents ceased, and the practice of medicine was forever altered.
Uncontrolled studies such as many of those proposed for chloroquine/hydroxychloroquine are invariably plagued by questions of validity due to unintended bias in the selection of patients as well as bias in the assessment of response to treatment and are thus not reliable for the determination of efficacy. Confounding factors may also include differences in baseline demographic factors such as age, sex, weight, race, or socio-economic status as well as differences in disease severity which can all influence response. Randomized controlled trials (RCTs) avoid these hazards. RCTs are often “double-blinded” with a placebo control so that neither the investigator nor the patient knows whether (s)he is getting the intervention (drug) or placebo. This blinding further reduces the potential for bias and increases the validity of the results.
While RCTs are expensive and time-consuming, they will ultimately deliver the quality of evidence needed to determine if a given treatment is effective. Moreover, since COVID-19 studies have been fast-tracked by the FDA, we should have preliminary results in a matter of weeks or months, instead of the years usually required for such research. Despite the limited evidence supporting its use, the large-scale availability of chloroquine/hydroxychloroquine and the encouragement of clinicians and patients to use it outside of a randomized controlled trial will undoubtedly impair the ability to recruit patients who do not want to take a chance of getting placebo.
Whether looking at the chloroquines, antiviral agents like remdesivir, or the use of convalescent serum (from patients with established antibody responses to COVID-19), without RCTs, we will be flying blind. It’s time for our political leaders to heed the advice of the medical and scientific community. The world needs to wait for the RCT results.