Bioethics Forum Essay
Clinical Trials vs. Right to Try: Ethical Use of Chloroquine for Covid-19
The president of the United States has famously touted the drugs chloroquine and hydroxychloroquine as potentially life-saving for Covid-19. They could be “one of the biggest game changers in the history of medicine,” he erroneously claimed. “So, it’s tested in the sense that you know it doesn’t kill you.” He has variously claimed it has “tremendous promise,” is “very effective,” and “can be taken safely.” His hyperbole hides the fact that chloroquine and its hydroxy variant have been around for a long time and have a reasonably safe record when used for other purposes, such a lupus and malaria prevention. I took chloroquine for two years while living in Nigeria from 1962 to1964 in the Peace Corps and didn’t think too much about it. There is also at least very preliminary report that a handful of Covid-19 patients may have benefited from the drug.
That being said, it has some risks. It can cause cardiac arrhythmias. According to the National Library of Medicine’s medlineplus.gov, it is associated with headache, nausea, loss of appetite, diarrhea, upset stomach, stomach pain, rash, itching, hair loss, seeing light flashes and streaks, blurred vision, reading or seeing difficulties (words disappear, seeing half an object, misty or foggy vision), difficulty hearing, ringing in ears, muscle weakness, drowsiness, vomiting, irregular heartbeats, convulsions, difficulty breathing, mood or mental changes, decreased consciousness or loss of consciousness, and thinking about harming or killing yourself, On the other hand, most of these effects are minor and rare. This leaves us with the question of whether the president, if sometimes muddled, over-enthusiastic, and just plain wrong about the drug, may nevertheless have a case to make. Should those suffering from Covid-19 or merely worried about being exposed have the right to use the drug?
I unqualifiedly accept the position that double-blind randomized clinical trials are the gold standard for answering the scientific question of whether a drug produces any effect, positive or negative, in Covid-19 patients. My training in research neuropharmacology leaves me biased in favor of insisting on such trials to answer the scientific questions.
There is a second, completely independent question to be addressed, however. It is the question of whether it is rational for a patient to choose to try a drug such as chloroquine for Covid-19 outside of a trial. Some patients may correctly hold that they have little to lose by trying it. Being killed by the drug might be seen as no worse or even better than weeks on a ventilator with a high probability of death (88%, according to one recent report). This question of individual patient self-interest is legitimate even if it leads to a different answer than the research question. It is sadly the case that sometimes trying the drug is in the rational interest of a patient who can’t wait for the results of a clinical trial.
To enter a subject into a randomized clinical trial ethically, the condition known as “equipoise” must be present. The concept of equipoise, first made well-known by a former Hastings Center intern, the late bioethicist Benjamin Freedman, incorporates the notion that there must be more or less indifference between the two arms of a randomized trial for a randomization to be ethical. There has been a lot of controversy over exactly who has to be more or less indifferent (i.e., “equally poised”) and how close to complete indifference must exist. The insight begins with the observation that if one were completely indifferent between two arms of a trial (between an active compound and a placebo, for instance), the only way one could choose between them would be by a flip of the coin. The theory of moral justification of randomized trials begins with the insight that, if one is completely indifferent and would flip a coin, there is no moral conflict between flipping the coin in a randomized trial and the interest of a patient. In early use of the concept of equipoise it was held that the research community must be at least relatively indifferent between the two treatment arms for a trial to be justified.
That justification ignored the possibility that an individual patient may be indifferent between treatment arms while researchers were not or—to consider the more problematic case—a patient may have a preference for one arm while researchers are legitimately and honestly indifferent. This is not a problem of patients being irrational. A patient may have a unique package of values that would lead him or her to prefer one of the treatment arms while researchers, with their values, really are indifferent. It is possible for a patient to correctly perceive an advantage from an experimental agent even if most researchers believe there is no net advantage from the experimental agent compared to a placebo. For him or her, the treatment is rationally linked to the preferred outcome. For instance, a patient may be terribly risk-averse and want to avoid the experimental treatment arm while researchers with more typical risk aversion consider the risks to just about equal the potential benefits. That patient today has the absolute right to get what he or she prefers (by refusing to consent to being in the trial), and it is reasonable for him or her to refuse the offer to be randomized.
On the other hand, another patient may aggressively value taking the risks of the experimental treatment; he or she may say that life isn’t worth living unless I am cured of my condition or that without the experimental agent my future would be so terrible that it is worth taking the risk of the new drug, For such a person, access to the experimental treatment is rationally in his or her interest. The patient would rather die from the experimental agent than live in the condition that will result without the drug.
At such a point, where the patient has a rational preference for the treatment while the researchers are legitimately indifferent, there is a conflict between answering the research question in the best way possible and the right of the patient to try something that really is in his or her interest.
This could be the unfortunate situation we find ourselves in with chloroquine and hydroxychloroquine. It is misleading, indeed false, to say that a patient has nothing to lose by trying the drugs or that they can’t kill someone with Covid-19 just because they almost never kill people with lupus or those trying to avoid malaria. On the other hand, it may well be true that the risks of not using one of the chloroquines are greater for some Covid-19 patients than the risks of using it. That was particularly true in the early weeks of the chloroquine discussion when we had some very tentative evidence that it may have helped a handful of patients and had not harmed any of them. The benefit-harm calculus to a given patient changes daily as a result of new drug information and his or her disease progression. The patient facing the decision to go onto a ventilator in the days when evidence of harm was lacking might have plausibly looked favorably on chloroquine. On the other hand, recent reports from a nonrandomized clinical trial of hydroxychloroquine suggesting that those getting it had higher mortality than those who did not could change the calculation in the other direction. With some sets of daily-changing facts, it may be rational for a patient to prefer the risks of chloroquine while at the same time the research question can best be answered by a randomized trial.
At this point we have two options: 1) require that patients only get access to the drug through a randomized, carefully controlled trial, thus riding roughshod over the rational interest of some patients in trying the drug or 2) permit patients to choose the course they prefer based on their values (including the option of permitting those who rationally prefer the drug to have access to it). If we are clever, however, we might be able to satisfy both goals at the same time,
The conservative patient who prefers not to risk taking the drug until it is proven effective for Covid-19 has the absolute right to his or her preference. Once we realize that it is logically possible for a patient to prefer the unproven drug, we could also give that patient access. If we are nimble in our planning, there will be a time when researchers are legitimately indifferent between the drug arm and the placebo arm It is plausible that some patients are likewise indifferent. We can do the randomized trial using those patients who are close enough to being indifferent that they don’t mind flipping a coin.
We can, with moral justification, complete a clinical trial even after researchers begin to have a preference for one of the arms if we realize that, at least for a while, some atypical patients may drift into indifference while other patients and researchers begin to have a preference for one of the arms. A month or so ago, when there were rumors of possible benefit of hydroxychloroquine with scant evidence of serious harms, some conservative patients might nevertheless have preferred not trying the drug while some slightly less conservative patients might have been indifferent to whether they got the drug or not. The latter patients could ethically have been randomized. Meanwhile some patients more willing to take risks might have preferred trying the drug and should have had a right to get it.
A day before this writing, an unpublished report leaked that a Veterans Affairs study was finding no benefit and possible higher mortality in the group getting hydroxychloroqine. That information might make some patients change their minds. Patients who would previously have preferred the drug to a placebo might become indifferent and willing to be randomized; somewhat conservative patients who would have previously been willing to be randomized might now refuse to enter the trial.
In the end, whether hydroxychloroquine is beneficial or harmful depends not only on the results of trials, but also on the unique value systems of individual patients. It is those patients who are close enough to indifference that they are willing to be randomized who permit us to ethically answer the science question. If we do it right, we can answer the science question while simultaneously recognizing the right of patients to choose the therapeutic strategy that will best fit their idiosyncratic values. Those who prefer the standard (nonchloroquine) treatment already have the right to get it. Those who prefer the chloroquine should get it as well (at least as long as granting access does not deprive those who need the drug for more traditional accepted uses). The patients who are ambivalent about getting chloroquine can ethically be randomized. As long as there is an open question in the scientific community about whether chloroquine or its hydroxy variant helps patients with Covid-19, there should be patients close enough to indifference that they are willing to volunteer to be randomized. Especially at the present moment, there should be plenty of patients in all three categories. We can simultaneously honor patients’ preferences and do gold standard research.
Robert M. Veatch, PhD., is Professor of Medical Ethics Emeritus and Senior Research Scholar at the Kennedy Institute of Ethics, Georgetown University. A Hastings Center fellow, he was the associate for medical ethics and then senior associate at The Hastings Center from 1970 to 1979. Twitter: @veatchr
45% of The Hastings Center’s work is supported by individual donors like you. Support our work.
Brilliant analysis…should be mandatory reading for the Federal CV Task Force.
Queries:
1. Is there a controlled randomized study for hydroxychloroquine being planned at this time? When was lr will it be launched, and how long will it run?
2. Once the study population is filled (including participation by those who prefer and thus who are allowed to be in the control variable arm as suggested here), what is the ethical response to others who might prefer to take the drug? Must they wait for the results of the trial, or is it ethical for them to be allowed to take the drug?
My understanding is that randomized trials are being planned. Moreover, since chloroquine is on the market for other uses (lupus and malaria prevention), it is legal for a physician to prescribe it for “off-label” use. Even if the experimental drug were not available for other uses, it can be made available for “compassionate use.” Since any access is on prescription only, the issue becomes one of persuading one’s physician that it can be rational for a well-informed patient to choose to use such drugs. The point of my Forum commentary was that such use can be rational, even if the results of trials are not in.
I agree with the possibility of a patient to be treated with a medication such chloroquine, but what’s the way to act with others drugs such as anti inflammatory monoclonal antibodies. If a patient is invited to participate in a protocol in which he or she could be assigned to a placebo arm. She or he has the right, in a COVID situation, to require to be treated with such drug that was approved for other desease?
Thank you for this interesting essay Professor Veatch. I agree with your conclusion and want to emphasize the importance of high quality informed consent as a means to this ethically desirable end.
It is true that any patient has the right to freely choose a treatment if they take it or not, but this option is reasonable if it is raised on accredited treatment. While a treatment is the subject of investigation (by means of a clinical trial or other procedure), it is not reasonable for the patient to choose its intake, outside the study procedure.
It is not reasonable that patients can choose unaccredited treatments (or discredited if we are guided by the first results), with significant potential toxicity as an element of free autonomy. There can hardly be valid decisions without adequate information, and now we have fears, doubts, hypotheses … but little verified information.
I don’t agree that it is unreasonable for a patient to want to try a drug while it is still under investigation. Obviously, such a decision requires careful, informed choice (as indicated in the previous comment). My point was that benefit-harm estimates are subjective judgments. A well-informed patient may make these subjective judgments differently from the investigators or other pharmacological experts and, when they do, they are not objectively wrong. When the expected outcome without use of an experimental agent is very bleak–as in terminal cancer as well as in patients with covid-19 who face a bleak prognosis at the point where they require mechanical ventilation–it becomes more reasonable to want to try an experimental agent even if one knows the treatment may fail or even produce a harmful outcome. The real problem is not that such a choice is irrational, but that, if enough patients make such a choice and gain access, it may become impossible to get enough patients to volunteer to be randomized to complete the study. “Fortunately,” for covid-19 there are thousands of potential research subjects. Even if some choose to refuse to enter the trial because they want standard treatment and some more refuse to enter the trial because they prefer the experimental agent, as long as researchers legitimately consider their subjective estimates of risk-benefit profiles to be about equal for the two options (that is, if they are in equipoise), it is likely that quite a few patients will agree and volunteer to be randomized. It is contradictory for researchers to claim that taking the drug on a nonrandomized basis is too dangerous and simultaneously to claim that they are in equipoise in their belief that they don’t know which agent is better.
Thank you for this! It is extremely insightful, as one would expect. This post acknowledges how quickly our knowledge is changing, which of course makes for serious challenges for consent.
The position put forward here seems especially applicable to the current pandemic, which involves a very large number of patients and an illness that progresses quickly. It might have more limited application to conditions for which there is less cost to delaying treatment, or conditions affecting smaller numbers (so that if many go the route of right to try then it becomes harder to recruit for a proper study.)
Now substitute “injected (low-dose, non-fatal amounts of) bleach,” or “exposure to UV light” or “sacrificing a goat to St. Dymphna” for “chloroquine” and this argument still works. That’s the problem. Asserting that people who want to try something of unknown benefit and risk because they’ve heard about it on TV should be able to is like saying “let’s sacrifice the gullible for medical research.”
I agree that double-blind randomized trials are the gold standard–there’s a reason. The risk-taking nature of those who might opt for a new/untried treatment may well translate into other aspects of their lives and serve to make the groups less truly random, yes? The “rational interests” of people sick with a disease that may kill them are hardly freely assessed by those patients–in which case it’s not the less risk-averse but the more fearful who make those “free” choices, yes? Allowing people into non-randomized experiments can also make it harder to fill those gold-standard trials, thus slowing the progress to a verified treatment, at a time when we desperately need good information as fast as we can get it. (An exception can and should be made for emergency use protocols under the aegis of an IRB–but those processes already exist.)
Finally, I think this piece misrepresents the interest level of people who wish to be randomized. They are not necessarily indifferent or close enough to indifferent–equipoise applies to researchers, not to participants. They may in fact be passionately interested in the outcomes of the trials in which they participate. They may also be willing to be generous enough to participate in a gold-standard trial so that rationally-supportable information can be gathered from their risk. Why is this risk less laudable than the risk of people willing to take something a TV huckster told them would work?
You make good, important points. It is true that granting nonrandomized access to agents currently in randomized clinical trials can–at least sometimes–make it more difficult to complete the trial. My point was that there can be legitimate conflict between the goal of producing the best scientific information and the goal of permitting rational patients to have access to the course that best serves their interest. Fortunately, one good thing about the Covid-19 crisis (perhaps the only good thing) is that there are many thousands of potential subjects for research so that, even though some patients would rationally prefer the “standard” treatment and others would rationally prefer the experimental treatment, if researchers are legitimately in equipoise, then many patients, after they are well-informed should be also and they can be recruited ethically for a randomization. I disagree, however, about the application of the concept of equipoise to patients as well as researchers. Both patients and researchers can make benefit-harm calculations. When patients are adequately informed, they may still hold different values that lead them to be in equipoise at a different point from researchers. When a patient ends up out of equipoise, it is reasonable that they should not be forced to enter a randomized study in order to get a 50% chance of getting what is in their interest. There are exceptions (if, for instance, there is not enough supply of the investigational agent to complete the study if those wanting the agent on a nonrandomized basis are also given access), in the normal case, those wanting to try, should have a right of access if a randomized trial is defensible. One cannot argue that it is too dangerous to give such access and simultaneously hold that researchers are in equipoise, that is, that they honestly believe the risk-benefit profiles of the two arms of the study are about equal. As you correctly point out, this does not imply that patients have a right to anything they want to try. For now at least, I apply the argument only to cases where there is an approved trial, which means that the investigators have convinced authorities (such as IRBs) that their equipoise position is reasonable. In those cases, patients should have access to either arm on a nonrandomized basis, but, since many patients will reach the same equipoise position as the investigators, there should be subjects available willing to be randomized.
I make one additional point: If investigators are more or less in equipoise at the beginning of a trial and there is time enough to have interim looks at data (in the covid-19 case because the patients are surviving at dramatically higher rates, for instance), then it becomes ethically impossible to complete the study since (at least using Bayesian logic) they should no longer be in equipoise just before the stopping boundary is reached unless some subjects who were originally not in equipoise are driven into their zones of indifference and thus can be ethically randomized.
The headline of this article is unfortunate and appears to tacitly endorse access to investigational drugs via the Right to Try pathway. Right to Try began as an anti-regulatory movement and has been opposed by clinical societies, consumer protection groups, medical ethicists, and patient advocacy organizations. The Federal Right to Try Act of 2017, enacted two years ago and touted heartily by President Trump, removes both FDA and IRB oversight of individual patient access to investigational drugs. As many ethicists have argued – some in these pages – this weakening of oversight exposes patients to considerable harms. The importance of FDA expertise and the crucial input of IRBs have been brought into stark relief during this devastating pandemic. A more apt, and less politically fraught, term for the headline could be “patient choice,” “individual autonomy,” or “patients’ right to decide.” Using the term “right to try” in the context of Dr. Veatch’s article may cause confusion and could inadvertently lend respectability to a law that hasn’t earned it.
I agree with everything you say. I tried to make clear that President Trump’s uncritical views of the chloroquines cannot be accepted, but I look for grains of truth in even the most irrational positions. The suggested terms such as patient choice and patient’s right to decide are good terms, perhaps better than “right to try,” but, in the end, they mean about the same thing. My conclusion holds that, at least for cases where an agent is in an approved randomized clinical trial and the agent is not scarce (see my replies above), those patients who, after being informed, prefer the experimental agent should have a right to it.