The moral implications of placebos

Steve was a builder, working on a busy site. One day, part way through his shift, he was in a hurry and jumped down from the wall he had been sitting on without looking properly first. As he landed, he saw it. A 13 inch nail had impaled his foot, sticking right out of the top of his boot. Pain coursed through his body and he began screaming in agony, and yelling at his colleagues to call an ambulance. Steve was inconsolable, and was in so much pain that he had to be sedated on the way to hospital. He arrived, and the doctors gingerly began trying to remove his boot. When they finally managed, they looked at him in shock.  The nail had passed between his toes, causing no damage at all to his foot. So why was he in so much pain? Simply the thought of that nail through his foot was enough to trigger pain that was real, even if the injury wasn't.

This story may be apocryphal, but it is an example of a real phenomenon - the placebo effect, or perhaps more specifically its negative counterpart, the nocebo effect. The placebo effect has long been known to play a role in healthcare, but how exactly it works isn't well understood. At the Cheltenham Science Festival, three experts came together to discuss the topic, and the big question - is it ever ethical to lie to patients, if it is for their own good?

The panel was made up of health psychologist Felicity Bishop, surgeon and medical researcher Andy Carr and Andrew George, who is an immunologist and lead author on a report about placebo surgery. Andrew started by defining a placebo as a treatment that has no direct biological or therapeutic effect. But even that definition has its problems - the whole reason we are interested in placebos is because they do have a therapeutic effect, even if it isn't through the mechanisms expected. But how can an inert pill cause some people to experience real improvements in their symptoms?

Felicity explained that most modern medicines work by stimulating the brain to release chemicals. Placebos can work in the same way, causing the release of the same chemicals, so do have real biological effects. There are two main mechanisms by which this can happen - the first is all down to expectation. By thinking that a treatment will work, our brains can respond to that treatment in a way that actually causes it to work. But learning mechanisms are also important. If you have previously taken a pill of a certain colour and size and experienced its effects, taking that pill again can cause your body to respond in the same way.

Felicity gave an example she had experienced herself; she recently gave up caffeine after many years of drinking a lot of coffee, and found that when she drank a cup of decaf, the taste and smell was enough to make her feel more alert. There was no (or at least little) active substance in the drink, but her past experiences of feeling more alert after drinking a similar tasting liquid were enough to have an effect. Interestingly, expectations and learning effects are of differing importance for different conditions. Pain, for example, seems to respond better to expectation effects while allergies are impacted more by learning mechanisms.

So can, and should, we use placebo treatments in regular medicine? Research shows that many doctors already do. While only 12 per cent of UK doctors admit to having used traditional placebo such as a sugar pill, 97 per cent say they have used a placebo-like treatment, such as giving such a low dose of a drug that it is unlikely to have any pharmacological effect, or sending patients for unnecessary scans or investigations.

The moral implications of doctors prescribing placebos are complex. Some, including Andrew, argue that is it always wrong for a doctor to lie to their patient. Others take a more utilitarian standpoint, believing that a lie can be moral, if it is for the patient's own good. I find myself on leaning towards the second viewpoint - I would be quite happy to be prescribed a placebo if there was no other treatment that would work, and if there was a chance it would help. But this thought makes many uncomfortable. A handful of studies have suggested that placebos may work even if the patient knows they are a placebo, raising the intriguing possibility that doctors could prescribe inactive pills withoutlying, but more study is needed to ensure that this approach would reliably work.

It is also vital to remember that placebos are not a magic bullet. They don't work on everyone and they don't work in all circumstances. Some conditions will never respond to placebos. But many of those that do, such as chronic pain, have few alternative treatments. If a sugar pill can help someone who suffers from such bad back pain they can't work, and for whom all other treatments have failed, surely it is in everyone's benefit to give provide access to them? I certainly think so.

Where the moral implications become even more complicated, however, is when we come to more invasive procedures than simple drug delivery. Recent work has suggested that the placebo effect may even influence surgery. Andy explained that the more dramatic the intervention, the bigger the placebo effect. So a capsule is better than tablet, an injection is better than a capsule and surgery is even better than an injection! But the concept of placebo controlled trials for surgery is a new, and extreme sounding idea. The thought of giving someone anaesthetic, cutting them open, and then not doing anything (except sewing them up again afterwards) sounds crazy, but recent studies suggest it is vital we start routinely introducing placebo trials into surgery, at least for conditions known to be linked to the placebo effect.

These studies are beginning to be done, and the results are hugely surprising. It seems, for example, that the effects of a common treatment for knee pain, known as an arthroscopy, may be entirely down to the placebo effect. The group who received sham surgery reported just as much pain relief as those who received the real surgery. And even more worryingly, in a study using stem cells in the brains of patients with Parkinson's, the group who received the surgery but had no stem cells inserted actually did better than those who had the stem cell treatment!

These findings suggest that we may be subjecting people to surgeries which only help them via the placebo effect. The obvious downside to this is the fact that surgery is dangerous - people can, and do, die because of the trauma or bad reactions to the anaesthetic. Andy argued that this means we shouldn't carry out surgeries that only trigger the placebo effect.  But if someone's knee pain is unbearable, and all other treatments have been unsuccessful, could it be ethical to carry out some kind of surgery, even ifit's only effect is placebo? Just because it is placebo doesn't make the reduction in their pain any less real, and perhaps it should be down to the patient to decide whether they are willing to take the risk.

So far, I have discussed placebos as if they are an alternative to active treatment, but it isn't that simple. Placebo and active treatments can co-exist, and interact, and we may be able to harness their power to make other medicines more effective. Andy told us about a study he had conducted looking at pain in undergraduates in Oxford. He put the unfortunate volunteers in a brain scanner and watched what happened as he placed chilli powder in their belly buttons (something he assured us is extremely painful)! He also asked them to rate their pain on a scale of 1-10. Next, he gave the students morphine, without them knowing, and their pain dropped from a 10 to an 8, showing a real pharmacological effect of morphine, as would be expected. However, he then told them he was giving the morphine, but didn't change the dose. Their pain went down from 8 to 4, purely from the placebo effect of being told they were being given the drug. Even more amazingly, when he told them he was stopping the morphine, their pain went back to 10, even though he left it on.

So perhaps we can use the placebo effect not just on its own, but alongside mainstream medicine, to boost its power. Perhaps by simply changing the shape or colour of a tablet we can increase its effectiveness. Studies show that the reason treatments like homeopathy can sometimes work has nothing to do with the pills themselves, but everything to do with the consultation, the care shown and the time taken to understand the person's symptoms and reassure them that the treatment will help.

Some people think that allowing doctors to prescribe placebos would be a slippery slope, leading to a loss of trust in the medical profession. But as we learn more about how the brain and body interact and influence each other's health, this is a discussion that medicine needs to have. It may be that our current narrow focus on biochemical pathways isn't enough. If by changing what goes on around the prescribing of a drug, like the way a doctor interacts with her patients, we can produce better outcomes, shouldn't we try it? There are a number of conditions that cause huge impacts on people's lives but that we don't fully understand, and don't have effective treatments for. While we search for cures, maybe we could use placebos to reduce their symptoms and relieve their pain. And isn't that something worth aiming for?

- Ginny Smith is a science communicator. Find her on Twitter.

The Cheltenham Science Festival runs until 12 June, including British Psychological Society events.