Over the Christmas break, I found myself thinking about experimental design – specifically how to design a trial to determine categorically whether or not the widely-accepted placebo effect is a genuine phenomenon. Don’t ask me why!
The more I considered this issue, the more I felt that many of the principles of Good Clinical Practice cannot rigidly apply – particularly the principle of blinding and informed consent.
The placebo effect has been the subject of considerable research since it was first widely postulated back in 1955. Some researchers observed measurable physiological and psychological changes from the use of placebos – from physiological changes in the brain (involving alleviation of pain, depression, anxiety, fatigue, and some symptoms of Parkinson’s) to changes in heart rate and blood pressure.
But placebos are based on deception – they only work because the patient believes they will work. Understandably, the idea of systematically lying patients does not sit well with many doctors.
To test whether or not this deception is key to the placebo effect, I found myself reasoning like this…
Clearly, there needs to be a group of volunteers who meet rigidly-defined criteria and who are given the recognised treatment for a particular condition (Group 1).
Normal trials would include a blind element and so Group 2 would be randomly chosen to receive what appears to be the active treatment but in fact contained no active ingredient. To be truly comparable the “control” medicine should be taken in an identical manner to the active substance. Usually double-blinding is used, where neither the researchers nor members of Group 1 and Group 2 know whether a particular participant is receiving the active drug.
Many would probably argue that the above two groups are all that you need to test the placebo effect – if Group 2 does better than would be expected for untreated patients, the placebo effect must be at work.
I disagree.
Supposing any apparent improvement (above normal expectation) of those receiving no active ingredient is a response to the regular medical care regime of the trial itself – rather than the body being fooled into believing it is receiving an active drug?
Numerous, well-publicised studies have implied that productivity in businesses increases when almost any procedural change is introduced – even when the change is to re-introduce a procedure that was stopped in a previous productivity-boosting change.
This is believed to be partly because of the added enthusiasm engendered by being part of something new and the inevitably increased attention that individuals receive. Why should the same not apply to participants in medical trials?
To test this, the trial would need a further two sets of participants – and this is where the principles of GCP would be strained…
Group 3 would be asked to do everything that Group 2 does (including to participate in all of the medical examinations etc). The only difference being that they do not actually take any medication – active or otherwise. They would be completely aware that they would be receiving no medicine but they would also be aware that, despite this, they play an important part in the trial.
Some studies into the placebo effect included both “placebo” and untreated groups – I think this is essential but I wonder if it goes far enough.
I propose a 4th Group that comprises essentially unknowing participants. This group meets the precise inclusion criteria for the other three groups but, uniquely, they believe they are not part of the trial. Instead of being regularly monitored, they will merely be assessed at the beginning and end of the trial to provide a true baseline against which the other three groups can be measured.
I would be really interested to know what you think!
|
|
 |
|
|
Share on LinkedIn |
