Arch Phys Med Rehabil 2009 (Feb); 90 (2): 366–368 ~ FULL TEXT
Gregory N. Kawchuk, DC, PhD, Rick Haugen, MD, FRCPC, Julie Fritz, PhD, PT, ATC
Department of Physical Therapy,
University of Alberta,
Edmonton, Alberta, Canada.
OBJECTIVE: To determine if short-duration anesthesia (propofol and remifentanil) can blind subjects to the provision or withholding of spinal manipulative therapy (SMT). DESIGN: Placebo control. SETTING: Day-procedure ward, University of Alberta Hospital.
PARTICIPANTS: Human subjects with uncomplicated low back pain (LBP) (n=6).
INTERVENTIONS: In each subject, propofol and remifentanil were administered intravenously. Once unconsciousness was achieved (3-5min), subjects were placed in a lateral recumbent position and then randomized to either a control group (n=3) or an experimental group (with SMT, n=3); subjects received a single SMT to the lumbar spine. Subjects were given a standardized auditory and visual cue and then allowed to recover from anesthesia in a supine position (3-5min).
MAIN OUTCOME MEASURES: Before anesthesia and 30 minutes after recovery, a blinded evaluator asked each subject to quantify their LBP by using an 11-point scale. This same evaluator then assessed the ability of each subject to recall specific memories while under presumed anesthesia including events related to treatment and specific auditory and visual cues.
RESULTS: In either the experimental or control group, subjects could not recall any event while under anesthesia. Some SMT subjects reported pain reduction greater than the minimally important clinical difference and greater than control subjects. No adverse events were reported.
CONCLUSIONS: Short-duration, low-risk general anesthesia can create effective blinding of subjects to the provision or withholding of SMT. An anesthetic blind for SMT subjects solves many, if not all, problems associated with prior SMT blinding strategies. Although further studies are needed to refine this technique, the potential now exists to conduct the first placebo-controlled randomized controlled trial to assess SMT efficacy.
From the FULL TEXT Article:
SPINAL MANIPULATIVE therapy is a clinical intervention used in the treatment of LBP. Most often, SMT is applied as a high-velocity, low-amplitude force to the tissues overlying the spine. Although several investigators have reported SMT to be more effective than alternatives,  other studies report no significant benefit.  Given these conflicting results, arguments have been made that the primary effect of SMT is not biological but rather the result of subject expectations (ie, the placebo effect). 
To separate physiologic effects from those based on subject expectations, investigators must blind subjects to their treatment. In pharmaceutical studies, this is achieved easily through placebo controls, interventions that are therapeutically inert but equal to the active treatment in their mode of delivery and physical nature. By effectively blinding subjects to their treatment, the placebo effect is presumed to be equal between groups, whereas physiologic treatment effects are isolated.
Given the success of drug-based placebos, investigators have attempted to use this approach in SMT studies. Unfortunately, the properties of SMT are substantially different from pharmaceuticals, making it problematic to create an effective placebo.  Specifically, prior SMT placebos are often dissimilar to the physical  experience of SMT (eg, foot massage, nonspecific exercises, traction, detuned modalities), have an entirely different mode of delivery (eg, educational material), or may not be inert (eg, light touch, posteroanterior force). Given these circumstances, conflicting reports exist as to the effectiveness of SMT placebos for blinding subjects. 
Given the difficulties in creating an effective SMT placebo, investigators have attempted to achieve subject blinding by studying subjects who have no prior knowledge of SMT.  Because these subjects are thought to have no expectations regarding SMT, the need for a placebo that mimics the delivery mode and physical nature of SMT is alleviated. Although a promising approach, the effectiveness of naive subjects as an SMT blind has not been determined,  whereas the practicality of recruiting naive subjects and maintaining their naivety is questionable.
Placebo controls and/or naive subjects are needed to balance subject expectations in studies in which subjects are conscious and therefore aware of the treatment they are receiving. If subjects are never aware if treatment has been provided or withheld (ie, they are unconscious), placebo treatments or naive subjects become unnecessary. Given all of this, the objective of this study was to determine if short-duration, low-risk general anesthesia can blind subjects to the provision or withholding of SMT.
Propofol and remifentanil are used globally to achieve brief unconsciousness for noninvasive or minimally invasive procedures. Consistent with our results, these drugs also prevent implicit and explicit memory.  Given these effects, the use of propofol and remifentanil as an anesthetic blind for SMT may solve many, if not all, problems associated with prior SMT blinding strategies. In addition, the superior performance of an anesthetic blind does not come at the expense of subject safety. The safety record of propofol and remifentanil is excellent,  and their safety in a study such as this that uses short durations of unconsciousness in relatively healthy subjects who are not undergoing invasive procedures should be optimal.
Should a true blinding strategy be developed for SMT, it would be possible to investigate many long-standing questions related to SMT such as the number of applications needed to achieve a clinically significant effect and if different forms of SMT vary in their efficacy. Although the results of this study suggest a promising new approach, larger studies are needed before short-duration anesthesia can be considered to be an effective blinding strategy for SMT.
Specifically, it must be confirmed that propofol and remifentanil are inert. In this study, 1 control group subject reported a decrease in pain. Although this decrease was less than the minimally important clinical difference and the brevity of the period of anesthesia seems unlikely to produce a therapeutic effect, further studies are required to confirm that this anesthesia protocol is therapeutically inert. Should propofol and remifentanil be noninert, they could still be used as an SMT blinding strategy because their influence would be global and could be accounted for in all subjects. Comparatively, this is not true of placebos that must be noninert and cannot be provided to those receiving active treatment.
In addition, future work is needed to determine if SMT provided during short-duration anesthesia is similar to that provided clinically to conscious subjects. Because propofol and remifentanil cause decreased muscle tone, it may be possible that SMT applied in this situation has unique effects, either negative or beneficial. Negatively, a lack of muscle tone may leave spinal tissues unprotected and susceptible to SMT-related damage.
Although larger studies are needed, the possibility of SMT-related injury because of reduced muscle tone caused by anesthetic blinding appears remote for 3 reasons:
(1) the absence of adverse effects reported in this study;
(2) the absence of adverse events reported in related procedures ; and
(3) best clinical practices that use positioning, stretching, massage, and other techniques to reduce muscle tone before SMT.
Positively, a reduction of muscle tone caused by an anesthetic blind may act to enhance any beneficial effect of SMT. This speculation is plausible given that deeper, longer duration anesthesia (ie, medication-assisted spinal manipulation) has been used in some subjects with the assumption that it will reduce pain and muscle spasm in those who do not respond to traditional manipulation.  Unfortunately, convincing evidence that medication-assisted spinal manipulation results in outcomes that are significantly different from any other form of SMT is lacking.  Given this, it is unlikely that there is a significant therapeutic advantage to SMT provided in conjunction with the shallow, short-duration anesthesia used as a blinding strategy in subjects able to receive traditional SMT.
Given that the objective of this study was to determine if short-duration, low-risk general anesthesia can blind SMT subjects, measures of pain intensity were obtained secondarily to determine if subjects who were blinded successfully could guess their group assignment (ie, SMT or control group) based on self-assessment of their symptoms. Given that 4 of the 6 subjects guessed correctly, larger studies are needed to ascertain the validity of self-assessment and how correct/incorrect guessing may alter subjective reports of improvement. Note that any trend reported here regarding changes in pain intensity after SMT should be regarded as preliminary.
Finally, the usefulness of an anesthesia blinding procedure in clinical research may not be limited to SMT alone. We foresee its use in generating mechanistic knowledge for other physical treatments (eg, massage, acupuncture) whose understanding has been impeded because of a lack of an effective subject blinding procedure.
A short-duration, low-risk general anesthesia can create effective blinding of subjects to the provision or withholding of SMT. Given our results, an anesthetic blind for SMT subjects may solve many, if not all, problems associated with prior SMT blinding strategies. Although further studies are needed to refine this technique, the potential now exists to conduct the first placebo-controlled randomized controlled trial to assess SMT efficacy.