Bad Science by Ben Goldacre

(1978) gave patients a sugar pill before a dental injection, but the doctors who were handing out the pill gave it in one of two different ways: either with an outrageous oversell (“This is a recently developed pill that’s been shown to be very effective…effective almost immediately…”) or downplayed, with an undersell (“This is a recently developed pill…personally I’ve not found it to be very effective…”). The pills that were handed out with the positive message were associated with less fear, less anxiety, and less pain.
    Even if he says nothing, what the doctor knows can affect treatment outcomes; the information leaks out, in mannerisms, affect, eyebrows, and nervous smiles, as Gracely (1985) demonstrated with a truly ingenious experiment, although understanding it requires a tiny bit of concentration.
    He took patients having their wisdom teeth removed, and split them randomly into three treatment groups: they would have salt water (a placebo that does “nothing,” at least not physiologically) or fentanyl (an excellent opiate painkiller, with a black-market retail value to prove it), or naloxone (an opiate receptor blocker that would actually increase the pain).
    In all cases the doctors were blinded to which of the three treatments they were giving to each patient, but Gracely was really studying the effect of his doctors’ beliefs, so the groups were further divided in half again. In the first group, the doctors giving the treatment were told, truthfully, that they could be administering placebo, or naloxone, or the pain-relieving fentanyl; this group of doctors knew there was a chance that they were giving something that would reduce pain.
    In the second group, the doctors were lied to; they were told they were giving either placebo or naloxone, two things that could only do nothing or actively make the pain worse. But in fact, without the doctors’ knowledge, some of their patients were actually getting the pain-relieving fentanyl. As you would expect by now, just through manipulation of what the doctors believed about the injections they were giving, even though they were forbidden from vocalizing their beliefs to the patients, there was a difference in outcome between the two groups. The first group experienced significantly less pain. This difference had nothing to do with what actual medicine was being given or even with what information the patients knew; it was entirely down to what the doctors knew. Perhaps they winced when they gave the injection. I think you might have.
    Placebo Explanations
     
    Even if they do nothing, doctors, by their manner alone, can reassure. And even reassurance can in some senses be broken down into informative constituent parts. In 1987, Thomas showed that simply giving a diagnosis—even a fake “placebo” diagnosis—improved patient outcomes. Two hundred patients with abnormal symptoms, but no signs of any concrete medical diagnoses, were divided randomly into two groups. The patients in one group were told, “I cannot be certain of what the matter is with you,” and two weeks later only 39 percent were better; the other group was given a firm diagnosis, with no messing about, and confidently told they would be better within a few days. Sixty-four percent of that group got better in two weeks.
    This raises the specter of something way beyond the placebo effect, and cuts even further into the work of alternative therapists, because we should remember that alternative therapists don’t just give placebo treatments; they also give what we might call placebo explanations or placebo diagnoses: ungrounded, unevidenced, often fantastical assertions about the nature of the patient’s disease, involving magical properties, or energy, or supposed vitamin deficiencies, or “imbalances,” which the therapist claims uniquely to understand.
    And here it seems that this placebo explanation—even if grounded in sheer fantasy—can be beneficial to a patient,