I concur with Dr. Leonard that a more ideal study design would be one in which the control group subjects receive sham treatment—though they may think they are receiving the true therapy—and both groups are blinded as to whether they are receiving the true or sham treatments. This is easily accomplished with medications and is therefore the standard for pharmaceutical studies.
In the design phase of this study, my colleagues and I considered various sham posture instructions and exercises to attempt to control for the well-known "Hawthorne Effect" but thought that any interaction emphasizing posture might cause the subjects to develop a greater postural awareness. This could potentially provide a TMD therapeutic effect, negating its benefit as a control group. We concluded that the best design for this study was the design selected.
Dr. Leonard’s suggestion of using low-back massage as the sham treatment could be considered for future studies, but the lower back is so far away from the masticatory system that I believe subjects would suspect they had been assigned to the sham treatment group.
Because of the well-established relation between the neck and shoulder musculature and TMD, massaging the upper shoulders may provide some beneficial effects on the TMD. In the present environment of complete disclosure of experimental methods in the informed consent document provided to potential study patients, the knowledge that a sham or placebo treatment is to be employed can make it difficult to develop an appropriate sham treatment in which subjects believe they may be receiving the true therapy while not affecting the disorder being evaluated.
I agree that some subjects find sham treatment beneficial, but the degree of this benefit is generally quite small. A limitation of this study is that we did not control for the attention effect of sham treatment. In my opinion, however, the treatment group had such a large treatment effect that it was far beyond what would typically be seen with the placebo effect.
Once the subjects completed their participation in this study (four weeks), the subjects received the traditional TMD therapies that had been recommended at the initial TMD evaluation: occlusal orthotics, daytime habit-breaking strategies, relaxation and so on.
Some of the treatment plans were revised due to the improvements obtained from the posture exercises. Since additional TMD therapies were provided on completion of this study, we did not consider telephoning the subjects six months later, which on hindsight would have provided interesting information.
I would like to thank Dr. Leonard for his thoughtful suggestions on our methodology.
