You may see the terms “placebo effect” and “placebo response” used interchangeably under the umbrella term of “placebo”, but there are some nuances there worth teasing out. Originally, “placebo effect” would have described any improvements in symptoms in a group administered an inert treatment (e.g. pill with no medicine, unplugged machine), which would be accounted for by multiple factors, including regression to the mean, bias, co-interventions, “real
placebo” responses, etc.. On the other hand, a “placebo response” is restricted to those benefits resulting from a psychophysiological response to an inert treatment (Benedetti 2014).
To give an example, if we had a group of people with shoulder pain who were administered an inert treatment of a supposed painkiller (a pill with no medicine in it) and we found that they improved over 6 weeks, we couldn’t say the whole improvement was a “placebo response”. The improvement would also result from regression to the mean over time, other activities they may have engaged in during that time, and more, along with the placebo response. The actual
placebo response would be the benefits derived from the psychophysiological response i.e. what element of the improvement could be put down to the manipulation of the patient’s expectations and the psychosocial context.
Therefore, when thinking about what I mean by “placebo”, I would like you to consider it to be those improvements resulting from the psychosocial context. Here are some examples of what might come under that umbrella in the real world:
Senses – smell, sight, touch.
Words – positive or negative suggestions.
Meaning/Stories – how the suggestions fit in with what you already understand about medical treatment.
Expectations – what sort of expectation is being set by the clinician (e.g. you will feel much better after this).
Social Learning – what sort of treatments are considered effective within your social environment (e.g. expectation of improvement following manual therapy, perception of the operator model of rehabilitation).
Relationship/Trust – the relationship between you and the clinician, how they have treated you, how much you trust them, etc.
Therapeutic Ritual – the act of a needle being placed under your skin, putting a pill in your mouth, or a therapist pressing hard on an area.
There are a lot of variables like these to which we are often blind. We assume that the improvement in a symptom/condition must be specific to the treatment, but this is far from true. When painkillers are administered covertly, the effects are significantly diminished. Likewise, when painkillers are stopped (e.g. morphine) but the patient is told they are still being delivered the medication, they continue to have reduced pain, despite the fact they are no longer
receiving any medication (Benedetti 2014).
This extends even further to surgery, where there is a large theatrical component that is part of the ritual, and often a very large expectation of improvement. We have seen this in multiple surgeries that have been demonstrated to be no better than “sham” surgeries, where the surgeon makes a cut but doesn’t actually perform the surgery. In a cohort of patients with knee osteoarthritis, this was tested by the subjects being randomised into one of two groups
(Sihvonen et al. 2013):
Arthroscopic Partial Meniscectomy – A surgical procedure where an incision is made into the knee and part of the degenerated, torn meniscus is removed.
Sham Surgery – Replication of the surgical procedure (e.g. still bringing the patient to the operating room, getting all of the instruments, turning on all of the surgical equipment throughout to replicate the noise and environment).
Both groups improved, but there was no difference between them. This sounds far fetched, but if you think it through, there are multiple placebo effects at play, from the expectation to the suggestions to the environment, all the way to the post-surgical care and rehabilitation. It’s certainly something we should pay attention to, especially when we consider how biologically plausible the surgical procedure would seem to be, with most other treatments (including
those manual therapies discussed) seeming far less plausible in comparison.
Therefore, we shouldn’t be surprised when we see improvements in pain with therapeutic machines that aren’t even turned on. Crothers et al. (2016) investigated the efficacy of Graston technique, spinal manipulation, and a “sham”/placebo treatment for the treatment of non-specific thoracic pain. The sham
treatment was an ultrasound machine that was unplugged (so the patient would expect to be receiving ultrasound), yet there were no differences in the outcomes at any time-point between 1 week to 1 year.
Therefore, all groups improved over time, as is often the case, with neither Graston (IASTM) nor manipulation leading to better outcomes than an unplugged ultrasound machine, indicating a lack of efficacy of both treatment modalities.
When you are dealing with physios, chiropractors, and others claiming to "heal your pain", you have to keep all of this stuff in mind. You can learn more in the full article on our site.
- Gary McGowan
