I haven’t posted in a while. The current global Ebola panic, spread mostly by social media and the media and not so much by the actual global threat of Ebola, has spurred me from complacency. Specifically, a WHO ethics panel today unanimously authorized the use of unproven, untested, experimental Ebola drugs in the field. But missing from the public discussion of this move is a key question: what is the scientific benefit (or downside) of this decision? In this short post, I hope to address the issue of using unproven drugs on Ebola patients from the perspective of the scientific method.
Today, a World Health Organization (WHO) ethics panel unanimously voted to support use of untested, unproven, experimental drugs in the fight against the current Ebola outbreak. Of course they did so unanimously – to have done otherwise would have brought the ethical ire of the world down upon the WHO. Why should privileged Americans get the drugs, but not the poor and suffering West African population? These are serious ethical issues, not to be taken lightly at all. But in the kerfuffle over ethics, the most important thing in the deployment and use of medicine – science – seems to have been utterly forgotten.
What do I mean? Well, these drugs, by all media accounts (c.f. [1]), have never been through human trials. There is no gold-standard proof that these drugs actually treat Ebola infections in humans. I need to be careful here. What do I mean by this?
- By “treat,” I mean that the drugs lessen the symptoms, or fight directly the Ebola virus, better than a placebo treatment, given under the exact same conditions as the test drug, and resembling identically in every way (physical appearance, delivery, etc.) the test drug.
The drug most mentioned in the news (and now, as of today, mention of a Canadian Ebola vaccine), ZMapp, has never been through a gold-standard medical trial in humans (or, of course, any replication of that trial): a double-blind, randomized, placebo-controlled study. How might that look for Ebola, since so few people are infected by Ebola every year?
- An Ebola drug gold-standard trial might look as follows. With consent of those being treated in an outbreak zone, patients in the same facility will randomly be selected to receive placebo + supportive care, the test drug + supportive care, or just supportive care. Those administering the treatment to patients won’t know whether it’s the real test drug or the placebo drug. Data would then be gathered on each patient to determine the severity of symptoms and outcome (e.g. duration and nature of suffering, survival rate, or death rate). After the outbreak has concluded, the data would be analyzed without knowledge of which patients received placebo, the drug, or neither. Correlations would be searched for. Only after data analysis has concluded would the actual relationship between patient and test condition be revealed.
You might be recoiling in disgust right now. “That’s terrible!” you might say, “Because shouldn’t you just give the drug to everyone and measure the survival rate over previous outbreaks? Wouldn’t that be better?”
Scientifically, the answer is no. That is useless. Here is why.
- The Placebo Effect [3]: if a person believes they are being given a strong treatment, the evidence says they will get better faster even if they are receiving a FAKE treatment – a sugar pill, a saline injection, or an elaborate but meaningless ritual. This must be disentangled from the actual effects, positive, negative, or otherwise, of the test drug. The placebo effect confounds poorly constructed studies. Since three patients in the current outbreak have been given the drug in addition to supportive care, and 2 have so far lived while one has died, it is impossible to know whether or not the people who lived because of the supportive care, the drugs, the belief that they were getting better treatment, none of these (e.g. may have just been their immune system), or some combination of the above.
- Without giving the drug to some and not to others, but allowing people to believe (or not) that they are all getting the treatment (or not), there is no way to disentangle the effects of the drug from that of placebo. In many independent studies, the strength of the placebo effect has varied in unpredictable ways – sometimes 30-50% of people get better for no reason at all (they never actually received the treatment).
Our best information about medical practice comes from such gold-standard studies; a lack of such studies in areas like “alternative medicine” has led to a proliferation of questionable, useless, and even dangerous treatments touted as “medicine.” Science leads to reliable information. Its opposite – pseudoscience – leads to suffering, death, confusion, pain, and misery.
But, you might ask, won’t the drug manufacturers benefit from this ability to skip normal protocols for trials and go straight to their market? The simple and short answer is: no. Here is why.
- The drug maker has now an ethical obligation, under the WHO ruling, to provide the drug whether it works or not. They will learn nothing usable from the data they will gather, since no randomized, double-blind, placebo-controlled methods are in place. They will deplete their stock for no new knowledge about the efficacy of their product, and have no supplies with which to conduct trials for some significant time.
So there is a great ethical tragedy in this ruling by a panel of ethicists, ethicists who seem to have little or no knowledge of actual scientific methodology. By ruling that experimental drugs can be given to patients, they are depleting to zero the stock of unproven drugs while providing no useful data on the test outcomes. The WHO ruling has, if anything, set back Ebola drug research by months or years.
Worse, they have opened the doors to hucksters, liars, and snake oil peddlers. Anybody with an unproven treatment that sounds legit can now hand it out to the sickest and most vulnerable people on the planet. Those poor souls, in turn, will be sold a bill of goods. Some, or many, might forgo the established isolation protocols that are the best, proven ways of treating the infection. Some doctors, given unproven vaccines, might lessen their isolation protocols and inadvertently promote the spread of the disease.
Allowing scientifically unproven drugs to be given to sick and vulnerable people carries a tremendous scientific and ethical negative burden that will, no doubt, make worse and more complex the current Ebola outbreak in West Africa.
[1] https://ca.news.yahoo.com/ebola-experimental-drugs-vaccines-early-211930082.html and http://www.npr.org/2014/08/12/339878597/ethics-panel-endorses-the-use-of-experimental-drugs-to-slow-ebola?utm_medium=RSS&utm_campaign=allthingsconsidered
[2] https://secure.flickr.com/photos/niaid/8425030684
[3] For a number of resources on the amazing and fascinating placebo effect, checkout our lecture notes on the subject from CFB3333/KNW2333/PHY3333 (Introduction to the Scientific Method) at SMU: http://www.physics.smu.edu/pseudo/Placebo/
