For people with no knowledge of the Placebo, I will start with what Wikipedia has to say on the subject:
“A placebo is a simulated or otherwise medically ineffectual treatment for a disease or other medical condition intended to deceive the recipient. Sometimes patients given a placebo treatment will have a perceived or actual improvement in a medical condition, a phenomenon commonly called the placebo effect.”
A small variation of this test is the generally accepted standard of testing for any new treatment or drug, and is known as the double-blinded placebo test. To summarize it (from India Uncut):
“The standard test in medicine for seeing whether a treatment works is a double-blind placebo-controlled test. In this, patients are randomly divided into two groups, one of which is given the treatment being tested, and the other is given placebo—such as pills that look like real ones, but are actually inert. Neither the patients nor the doctors know which group is getting the treatment and which the placebo (that’s why it’s ‘double-blind’), thus eliminating psychological biases on their part. The mere belief that they are being treated often helps patients, so the true test for a treatment is if it can do better than placebo.”
This appears to be a fair and effective method to weed out non-performing drugs – and it is. This has been used for many many decades, even centuries, and so has passed the test of time. Drugs from many alternative systems of medicine have failed these tests repeatedly – due to which, these tests have served as the primary basis for these systems’ criticisms. These tests imply that the drug is to be considered ineffective if it cannot outperform the placebo. In other words, a patient needs to be getting better because he has taken the drug in question – not because he has taken just ‘some medication’ that he has been told will cure his ailment (in this case the placebo). My question now is not targeted at the veracity or the fairness of these tests. My question, however, concerns the implications of these tests when they are taken to the other extreme.
Consider a situation where a patient is given a medicine – from any system of medicine – for a particular ailment that he is suffering from. This drug is made specifically for the patient taking into account ALL necessary information needed as per the corresponding system of medicine to treat him. The medication is taken by the patient at the intervals the doctor/practitioner has recommended and the full dose is completed. However, there is a small difference. In this situation, the patient is completely unaware that he is taking ANY medication at all.
I am not going to get into ‘how’ the above mentioned situation can be executed in reality. I am also not going to get into a debate about the ethics and responsibilities at play here. My interest in this is purely that of inquiry and so if it means that ethics are broken during the potential implementation of this situation, I am fine with that.
What I AM interested in, however, are the results of a series of such potential tests on many patients. A slight variation of this test could include some patients being administered a known placebo – again, without their knowledge. But perhaps the best situation would be what I have outlined below:
For the sake of argument, let us assume that these tests are to be conducted on 100 randomly selected patients with no particular pattern in their medical history – but who are suffering from the same ailment. These patients are then divided into three equal sets. (Well OK smartass, one group will have 34, happy?). Set 1 will have the corresponding drug administered, as tailored to meet each particular patient’s needs, without any knowledge of consumption of ANY medication. Set 2 will have a known placebo administered into them, again without any knowledge of the same. Set 3 will have absolutely NOTHING done to them AND they will not be aware that nothing is deliberately being done to them. The full recommended dose of the medicine/placebo (or the lack of it) will be administered for all patients. The condition of the patients are monitored continuously and the progress (or lack thereof) is noted.
At first glance, administering a known placebo without a patient’s knowledge can appear to be ridiculous. But it is necessary to measure and compare the effects of the drug itself against a similarly induced placebo. Similarly, the absence of a treatment is also necessary to measure a situation where the patients get better ‘anyway’ and to see if the drug was needed at all.
Now let us look into all possible result scenarios and the corresponding implications for the drug and the placebo effect:
Scenario 1: Set 1, for whom the medicine was administered, show a general improvement and/or curing of the ailments. Set 2 and Set 3 do not show any signs of improvement over the same measured time frame. This is perhaps the most expected situation of a drug that actually works. One does not expect the unknowingly-administered placebo patient to recover without any of the subjectivity involved in the knowledge of having taken the placebo. Set 3’s results also go as per expected.
Scenario 2: Set 1 do not show any marked improvement in their ailments as compared to Sets 2 and 3. This is the other extreme and clearly shows that the drug is completely ineffective as compared to those taking the placebos and those having no treatment at all.
Scenario 3: All sets show general improvement at similar rates. This is similar to scenario 2 in that the drug has not outperformed the placebo and the non-treatment.
Scenario 4: Set 1 has not shown any signs of improvement whereas Sets 2 and 3 have exhibited improvements in their ailments. This is clearly a sign that the drug is actually inhibiting improvement and/or making matters worse, and should not be used at all.
Other scenarios include those wherein the patients in Set 2, who have had the placebo administered to them, suddenly display symptoms different from those in Set 3 (they could be better or worse). If this were to indeed happen, it would be the most curious case of clinical trials ever studied. Such tests have indeed been administered, albeit with the knowledge of the patients, and so cannot be referenced here.
But the bigger implications of these tests lie elsewhere. Consider Scenario 1 wherein it can be stated with good certainty that the drug does indeed work. Would the converse be true? That is to say, would the same drug pass the standard double-blinded placebo test? Is there a possibility that it would fail the standard placebo test but pass the modified placebo test as described in this post?
Looking at it differently, can this modified placebo test be viewed as both a necessary AND sufficient condition/test for a drug to be declared effective? I would certainly think so. If a drug is working without the patient being aware of its administration, it should certainly work (perhaps even better) if the patient is indeed aware of it. That, in fact, is the crux of the placebo effect. The effects of the knowledge of ‘something’ being consumed that is ‘supposed to’ cure an ailment is what lies at the root of the placebo effect. With that in mind, it follows that drugs passing the modified placebo test should definitely pass the standard placebo test.
If the modified placebo test is accepted to be the necessary AND sufficient condition for a drug to be considered effective, then it also follows that every drug that passes this test must be accepted as effective – irrespective of the system of medicine and without regard to its composition. Now, this is where things can get a little tricky.
For instance, consider Homeopathy. This system of medicine has long been criticized and ridiculed for prescribing drugs that have minuscule and even negligent concentrations of the active ingredient in them. Irrespective of any explanation that may have been put forth by the proponents, the extreme low levels of concentration can make one wonder how something like that can have ANY effect at all. But in my current inquiry, all this is completely irrelevant and beside the point. My only proposition is that ANY drug passing the modified placebo test should be considered effective.
The next question is obvious. What if there are instances wherein Scenario 1 was replicated in reality with Homeopathic drugs? That is to say that a patient was unknowingly administered homeopathic drugs and got better in a relatively short time – as compared to the time he spent suffering from the ailment with no medication. In such a situation, the exact workings of the drug – biological and chemical – become irrelevant and fade into the background. What matters then first and foremost is the fact that this drug indeed works.
Note that this particular illustration is definitely not restricted to Homeopathy. It can be used with ANY system of medicine and the implications will still remain the same. Sure, there will always be questions asked (and mostly answered) about the composition and workings of the drug itself. But in any case, the drug should simply be accepted as something ‘that works’. I am saying that it is OK if one is unable to explain how or a why a drug is working – as long as it passes this modified placebo test. But going back to an earlier observation, what if these very drugs, that have passed the modified placebo test, fail the standard placebo test?
Ideally, they should not. If they do, then the very mechanism and thinking behind the standard placebo test will have to be questioned. Fortunately or unfortunately, we will NEVER have sufficient data to ever make either claim with complete certainty. For one, this modified placebo test will almost certainly never gain approval in the medical community – not with all the ethical issues involved. And secondly, it would be very difficult to actually find a way to administer these drugs (or placebos) without the knowledge of the patient. And finally, and perhaps more importantly in the current scenario, nobody will ever publish such data even if such tests were indeed conducted, due to obvious ethical and professional reasons.
It is a shame that such a test cannot be conducted in the real world. It would answer so many questions and raise so many more. I suppose I can now relate to the pains of economists like Stephen D Levitt when they talk about the need for a controlled randomized experiment to explain the dynamics of this world. (A big smug follows).
In any case, I invite more discussion and differing points of view on this. Please feel free to point out possible misconceptions and overlookings.