It’s time to tackle a tricky subject. Everyone has heard of the placebo effect but there are a lot of misconceptions out there about what it really is and whether or not it’s ‘real’. Let’s start with a basic description of what it is: when a person is given a fake treatment for an illness, like an empty injection or sugar pills, it is common for them to feel better and for there to be a subjective improvement in their symptoms; this is the placebo effect. It works whether or not the person is aware of the effect and is also a component of the improvement seen when someone is given a real treatment.
One of the key things that I would like to get across is that the improvement is subjective, i.e. people will often feel better but, by definition, a placebo treatment cannot actually make someone better. And there is evidence to back this up, but I shall come to that later.
Another main point I want to make is that there is a huge amount of really bad information out there on the placebo effect. Unfortunately, even organisations like the NHS cannot be trusted to relate reliable information to the public. Take this page for example. It’s a mess; a mixture of reasonably accurate information and abject nonsense. Perhaps we shouldn’t expect anything more given that this is a page about the benefits of complimentary and alternative medicine (CAM) when there isn’t any such thing. There is mainstream, evidence based medicine and then there is quackery; CAM falls into the latter category. The main stumbling block that that NHS page and many others trip over, even those that mean well, is that it oversimplifies what is a complicated topic. For that reason I’m afraid this has to be a bit of a long post, I’d be doing you a disservice if I just fudged it.
Let’s pick apart, then, exactly what placebos can and can’t do. This is complicated somewhat by the fact that the placebo effect isn’t just one thing. It is a mish mash of responses some of which are real physiological responses and some of which are illusions caused by nothing more than regression to the mean, confirmation bias or reporting bias.
For indications that are completely subjective, like chronic pain or nausea, a placebo can have a significant response. People given a placebo treatment often report being completely cured (think of the curative power of kissing a child’s grazed knee) but, then, nausea and pain can be brought on psychosomatically and so it should come as no surprise to us that a psychosomatic treatment can cure them. In this case a placebo is no more effective than watching your favourite film, getting a hug from a loved one or taking a bracing walk. All of these things can and do make us feel better. What can commonly happen, however, is that after such an intervention we don’t report our symptoms in the same way. Someone might be feeling sick or even vomiting just as much but they’re not feeling so down about it and so they shake it off a bit faster and, if you asked them, they’d say that they’re feeling better even though nothing has technically changed. This is reporting bias.
So for symptoms that it is impossible for an objective observer to accurately measure it could be argued that placebos have a worthwhile effect, but this is putting the cart before the horse. If the placebo treatment is indistinguishable from a film, a hug or a walk then I’d say that you should watch that film, have that hug or go for that walk before you pop a fake pill that isn’t actually doing anything. There’s nothing to be gained from deceiving yourself.
Far more important than what placebos can do is what placebos cannot do, which is pretty much everything. It has become commonplace for practitioners of CAM to conduct studies to try to prove that their particular brand of nonsense; be it homeopathy, acupuncture, iridology, chiropractic, chelation therapy or whatever; does work. If they were proper scientists they would be trying to disprove their theory but we’ll leave that for now. In nearly every study the CAM treatment performs no better than placebo. To any vaguely scientifically literate person this would be conclusive proof that the treatment has not worked, the null hypothesis stands. CAM practitioners, either through ignorance or the will to deceive, take this as evidence of efficacy, however. They say that because there is an effect that the treatment works.
Let me give you a now classic example. In 2011 there was a paper published amongst much fanfare in the New England Journal of Medicine, a very prestigious and high quality journal. In it the researchers took a group of people with asthma and treated them either with the recommended treatment, an albuterol inhaler, or with a placebo inhaler, pretend acupuncture or no treatment at all. I’m going to show you two graphs from the study but don’t panic, I’ll talk you through them. They illustrate perfectly the problem with letting people think that placebos are worthwhile.
What this first graph shows is how much better each group felt after their treatment. As you can see, both the placebo inhaler and sham acupuncture performed very nearly as well as the mainstream treatment for asthma; they all showed an improvement in subjective symptoms of about a half. Impressive stuff. It is worth noting that the ‘no intervention’ group, those who had nothing happen to them, not even a placebo, felt an improvement of more than 20%. This improvement can be explained by things like regression to the mean. Let’s take a look at graph two.
Now we have a very different picture. This graph shows the improvement of lung function with each treatment; lung function can’t be faked, it is an objective measure of the severity of the asthma. When it comes to this objective measure the albuterol treatment clearly has a significant benefit whereas the placebo and the sham acupuncture arms of the study do no better than those who received no treatment of any kind.
This is a home run, you don’t get data much clearer cut than this. Placebos are able to make you feel a bit better but if you want to actually be better than they won’t do anything for you, at all. This is a pattern that repeats itself across the literature but one that CAM practitioners stubbornly refuse to take onboard. Indeed, the author of this asthma study avoided the natural conclusion of his own data and argued that objective outcomes shouldn’t be the benchmark for success. Instead, we should apparently be content with fooling ourselves that we’re better when the reality is that our ailment continues unabated. This is where the danger lies and what brings me to my final take home message.
It is very common for people to ask, what’s the harm? So what if people believe in woo? It’s their choice. And indeed it is their choice, people are free to make up their own mind. That’s the problem. Having been bombarded with nonsense by people who claim to have their best interest at heart, people with no background in healthcare, statistical analysis or sceptical thinking are expected to make difficult decisions about their health.
On the one hand there are the CAM conmen who can spend an hour with them in their plush office space, burying them in testimonials and soothing words, unrestrained by such niceties as efficacy and codes of ethics. On the other you have an overworked, hard bitten doctor who only has ten minutes with you before he has to see the next person, who only has the cold and often uncomfortable facts, who has an obligation of candour to tell you exactly how awful the future might be. It’s not difficult to see why people might choose the CAM path.
But there genuinely is a risk of harm. If someone is given a sham treatment and told that they will get better the evidence shows that they will feel a little bit better. This could stop them seeking conventional treatments that they need. In most cases this will mean unnecessary suffering, in some cases it will mean death. Do not scoff, this happens.
Given that placebos cannot achieve anything of substance there is a strong argument that prescribing a placebo is unethical. As a doctor you may know full well that there is nothing of substance wrong with someone, they are just one of the legion of the worried well perhaps. That is not a reason to deceive them to get rid of them. They need to be told to either go away and stop wasting your time or you need to explain to them that there isn’t a magic wand you can wave to make them better. A hard truth is better than an easy lie.
In summary, we need to do all that we can to disabuse the public of the widespread misconceptions that abound regarding the fabled placebo effect. We are duty bound to inform, educate and help those who need help and, where we can, to protect them from those who would do them harm.