Randomized controlled trials (RCTs) are considered the most rigorous form of research and are commonly used to evaluate the effectiveness of medical interventions. An RCT compares an intervention to either a placebo, gold standard, or both. Selecting the optimal placebo group is essential for valid comparison between groups and to measure the effectiveness of an intervention. A placebo treatment should be believable to ensure expectation of benefit, yet not provide a true treatment effect [1,2]. Thus, to a subject in a pharmaceutical clinical trial, the placebo treatment should appear indistinguishable from the active treatment.
Several definitions have been used to define placebo. The strictest traditional interpretation of the definition would limit the use of a placebo to only pharmaceutical trials using sugar pills as the placebo treatment. Such a definition is not useful in clinical trials of non-pharmaceutical interventions such as manipulative treatment. One definition of placebo that attempts to incorporate non-pharmaceutical interventions is "a substance or procedure that has no inherent power to produce an effect that is sought or expected" . McQuay and Moore more broadly define placebo effect as "the effect that we observe when patients are given a placebo," or "the effect caused by placebo" . Hróbjartsson suggests there are three main elements of placebo effect including: change after placebo medication (pre/post change in placebo group), effect of placebo intervention (treatment experience), and effect of patient – provider interaction . The term placebo response is often used interchangeably with placebo effect.
There are two prominent theories for why the placebo effect exists. The conditioning theory suggests that when pairing a neutral stimulus with an unconditioned stimulus (such as the active drug) the neutral stimulus elicits a response, resulting in a conditioned response . In contrast, the expectancy theory is based on the patient expectations. The response to a stimulus depends on what response is expected from the stimulus. These patient expectations may not account for all of the placebo effect, but they are the most significant factor of the expectancy theory .
A common misconception about placebos is that a third of the population will demonstrate a placebo response. This is not always the case. McQuay and Moore found that placebo response varies from 7% for pain treatment of migraines to 49% for pain treatment of diabetic neuropathy . The response to placebos in clinical research trials differs according to the length of the trial, the medical condition studied, the placebo used, and various other factors . Patient-physician interaction plays a role in placebo response. For example, a physician's attitude (enthusiastic versus doubtful) towards an intervention can significantly influence a patient's health outcomes such as pain, psychiatric illness, hypertension, and obesity .
The informed consent process also affects the placebo response. In one clinical trial, cancer patients not needing major analgesics were given either naproxen or a placebo for pain. Some of the patients were informed that they were participating in a clinical trial and would receive either naproxen or a placebo, while others were not informed about the trial. The patients who were informed of the possibility of receiving a placebo and actually received the placebo had greater pain relief than those patients who were not informed of the trial and were given naproxen . Additionally, the active drug to placebo treatment ratio of a trial may affect placebo response. Diener et al. examined the use of placebo in migraine trials. They found that participants in clinical trials with a greater active drug to placebo treatment ratio had a greater placebo response due to a higher expectation of receiving treatment .
Choosing an appropriate placebo treatment, sometimes referred to as ~sham~treatment, is an obstacle when conducting clinical trials with manipulative treatment. It is difficult to develop a placebo that mimics osteopathic manipulative treatment (OMT) or chiropractic treatment and produces the expectation of benefit. A variety of placebo treatments such as light touch , sham manipulation , and sub-therapeutic ultrasound  have been used in manual therapy clinical trials. Vernon et al. reported that after receiving a cervical sham manipulation, study participants did not report clinically significant changes in range of motion or tenderness. This study, however, included 20 subjects of which only three had no previous experience with chiropractic treatment . Hawk et al. reported improvement in subjects with subacute or chronic low back pain in both placebo and active manipulation groups. There were no differences in the amount of improvement between the groups, even after controlling for prior chiropractic experience and initial treatment expectations . Furthermore, experts do not necessarily agree on what constitutes the active component of a treatment or placebo. In a recent study, a list of 10 placebo manipulative techniques (including a description of each) was developed and sent to experts in Australia and New Zealand. At least one of the 16 respondents replied that each technique had an active component, and none of the techniques were considered an appropriate placebo by at least 50% of the respondents . To determine the credibility of light touch and sub-therapeutic ultrasound as compared to High-Velocity Low-Amplitude (HVLA) and standard of care, Slicho conducted a survey of the general population. Survey respondents more strongly agreed with HVLA as a way to logically treat low back pain, but responses did not differ in other aspects of treatment expectation after reading a description of the treatments .
Understanding the placebo effect and developing the best placebo is vital to studying the effectiveness of OMT. This current study was designed as a continuation of the research conducted by Slicho . Clinic patients were asked to view a video of three types of treatments (one active and two placebos) for chronic low back pain, and responses for expectation of benefit from the treatments were measured. The primary purpose of this research study was to determine attitudes towards different types of treatments commonly used in OMT clinical research trials. A secondary question was whether or not these attitudes are different if a person has previously received OMT or chiropractic treatment.