Ernst rightly emphasizes that personal contact between the patient and the physician and a manual therapy personally carried out by the physician produce a pronounced therapy-supportive effect on patients (as an added placebo effect). For that reason, all the German Acupuncture Trials used an equal length of treatment and attention to patients in the 2 acupuncture arms. The mean treatment duration in our chronic low back pain study1 was 30.5 minutes in all 3 therapy groups. The presence of the physician, however, was restricted in the 2 acupuncture groups to the time required to insert the needles (mean time, 8 minutes). The patient spent the remaining 20 minutes lying undisturbed on a bed in a separate room. In contrast to this, the physiotherapist in the standard therapy group often spent the full 30 minutes with the patient, at least if the patient obtained manual therapy. As described in our article,1 the patients received a combination of various treatments as prescribed by the current guidelines, including not only pharmaceutical therapy (95% of all patients took analgesics) but always, in addition, physiotherapy and/or massage and/or heat therapy, for example. It is of course possible that the delegation of therapeutic activity from the physician to a physiotherapist leads to a reduction in a placebo effect linked to the physician (point 3 by Ernst), but in compensation for that, the physiotherapist often treated the patient for up to 30 minutes, while the physician in the acupuncture groups did not spend even 10 minutes. With respect to point 1 by Ernst, we refer to our comprehensive discussion in the “Comment” section, in which we also examine the possibility of a “super-placebo” effect triggered by acupuncture but at the same time point out that the standard therapy is based on a genuine physiological action. This makes it difficult to attribute the large difference between acupuncture and standard therapy solely to a super-placebo effect of acupuncture.
Endres HG, Molsberger A, Haake M. Acupuncture Ineffective, Attention Effective?—Reply. Arch Intern Med. 2008;168(5):551–552. doi:10.1001/archinte.168.5.551-a