THE PLACEBO EFFECT BO EFFECTS IN PHYSICAL THERAPY

to be in any way unethical. Implicit in much of this research is the idea that placebo effects do not operate in isolation; they may ‘sit’ above or alongside many pharmacologically active treat- ments. Thus, even when a pharmacologically active treatment is given, an associated placebo effect may make that treatment more effective (we return to this topic later).

Below we review some of the medical research that has demon- strated placebo effects. In part two, we propose some strategies that therapists may adopt to better utilise such psychological processes in their practice. As was the case with our recent arti- cle on neuro-immune links (issue 3, sportEX dynamics), much of the evidence presented below is taken from the medical literature. However, we hope that its implications for sports rehabilitation practice are clear.

PLACEBO EFFECT RESEARCH The placebo effect, not surprisingly given its somewhat counter- intuitive nature, is the subject of much debate. Two recent meta-analyses of research illustrate this point well; Kirsch and Sapirstein (4) meta-analysed data from 2,318 patients in 19 double-blinded clinical trials, and demonstrated statistically that 75% of the effect of active drugs was in fact placebo, that is, only 25% of the therapeutic benefit being attributable to the pharma- cological properties of the drug. However in another recently published meta-analysis by Hrobjartsson and Gotzsche (5), the authors claim that in their analysis of 130 medical trials there was little evidence of the placebo effect. These authors in fact suggest that outside the realm of clinical trials there is no justification for the use of placebos.

However there is a substantial body of evidence suggesting that the placebo effect is a significant factor in therapeutic outcomes although there is much uncertainty about the mechanisms involved, for example, whether certain personality types or cer- tain conditions are more placebo responsive than others.

EVIDENCE FOR THE PLACEBO EFFECT In an often cited study, Dimond et al (6) investigated a curious phenomenon related to a surgical procedure called internal mam- mary artery ligation. The procedure was designed to reduce the painful symptoms of angina pectoris by blocking unhealthy blood vessels with the aim of re-routing the blood flow through health- ier vessels. Thousands of such procedures were carried out suc- cessfully in the 1950s. However, in most cases in which later investigation was possible, despite success in terms of symptom alleviation among patients, expected post-surgical anatomical outcomes were not evident. That is, in these cases, the blood sup- ply had not in fact been successfully re-routed. Dimond and his co-workers sought to investigate why patients had reported that the operation was a success while the surgeons knew that, tech- nically and anatomically, it shouldn’t have been. They performed the correct surgical procedure on one group of patients and a sham procedure (the whole procedure except for the ligation of the blood vessel) on another group. Participants in both groups believed that they had received the full surgical procedure and

www.sportex.net

both reported similar levels of pain reduction. The reduction in pain among patients in the sham operation group was a placebo effect (on this basis, it may also be argued that the reduction in pain in the surgical experimental group, and by implication all previous operations, was also a placebo effect). Evidence that similar placebo effects are present in current angina surgery has been presented (1).

In a study examining the injection of a placebo pain killer, Benedetti et al (7) used a group of post-operative thoracotomy and lobectomy patients. Patients were given an injection of a pharmacologically active pain killer at a dose resulting in a patient report of reduced pain. The following day, when reported pain had returned to pre-injection levels, one group of patients were given a placebo injection while the other were given no injection. During the following hour, patients in the placebo group reported significant reduction in pain while those in the no- injection group reported an increase in pain.

In a study examining ultrasound therapy, Hashish et al (8) placed three groups of post-operative dental patients in either an ultra- sound, placebo ultrasound (ie. machine positioned as per usual but not actually operative) or control (ie. no ultrasound) condi- tion. Data demonstrated that patients exposed to the placebo ultrasound reported similar pain reduction to those exposed to the real ultrasound and that both these groups experienced sig- nificantly more pain reduction than those in the control group. Over and above the reduction in pain, both real and placebo ultra- sound groups also experienced a reduction in post-operative inflammation that was not evident in the control group. Van Der Windt et al (9) in their review of ultrasound therapy for acute ankle sprains found compelling evidence that much of the thera- peutic effect of ultrasound is in fact placebo with no difference between true and sham treatments.

Voudouris (10) studied the effects of a placebo pain killing cream. In an initial trial, these authors used a device to stimulate localised pain among two groups of participants and recorded the pain tolerance of each. They informed participants that a cream they would apply to their skin was a potent pain killer, it was in fact a placebo. In a subsequent trial, the same pain-induction procedure was repeated with the placebo cream applied. However, in this procedure, the intensity of induced pain was reduced among one group (the placebo group) while it remained at a high level in the other (ie. the placebo group experienced reduced lev- els of pain due to the machine being ‘turned down’ rather than because of the cream). It was assumed that this process would cause the placebo group to associate the cream with reduced pain. A third trial, again with the application of the cream, returned both groups to the same high level of pain as in the first trial. Subsequent analysis demonstrated that, when the pain induction procedure was returned to similarly high levels for all participants, participants in the placebo group experienced less pain than the other group. Participants in the placebo group had in fact learned, through lower pain sensation upon the applica- tion of the placebo cream in trial two, that the cream was a pain

9