Appropriate times to measure the credibility of blinding in placebo-controlled acupuncture trials

Background – unsuccessful blinding

In a recent randomized controlled investigation, acupuncture with indwelling intradermal needles reduced pain and accelerated mobilization in patients undergoing elective Caesarean section (CS). ¹ In this investigation, bilateral acupuncture at four traditional auricular acupuncture (AA) point locations and six traditional acupuncture point locations on the body (n = 60) was compared to placebo acupuncture (n = 60). ² Both indwelling acupuncture needles (attached using adhesive tape) and non-penetrating placebo devices (adhesive tapes alone) remained attached for at least 2 days following surgery. Only 11 patients (20%) from the placebo group versus 25 (43%) from the acupuncture group believed that they had received real acupuncture (p = 0.02; chi-square test). The unsuccessful blinding of patients in this case can probably be explained by at least three things: (1) the participants (young curious women) were fully informed about the investigation and had access to Internet-connected gadgets; (2) postoperatively they received only medications that did not compromise their cognitive function; and/or (3) they were asked their opinion regarding group allocation on the second day following CS, by which time they would have had enough time to look under the adhesive tapes.

More unsuccessful blinding

The same result of insufficient blinding of participants was observed in a randomized controlled trial (RCT) of auricular stimulation for exam anxiety in medical students. ² This trial used a similar design/comparison – AA with indwelling fixed needles embedded in adhesive tapes versus a placebo procedure using only adhesive tapes without needles. In this study, even more participants in both groups correctly identified the intervention that they had really received; 34 participants (77%) from the acupuncture group versus 7 (15%) from the placebo group believed they received acupuncture (p < 0.001; chi-square test). Again, both indwelling fixed needles (used for AA) and placebo needles were attached in the evening of the day prior to an anatomy examination and removed in the afternoon the next day (the day of the examination). Again, a plausible explanation for the insufficient blinding is that the young, curious (and in this case partially medically educated) participants had had enough time to look under the adhesive tape and to distinguish the real intervention from the placebo. Also, the crossover design of this investigation added to the risk of unblinding.

Successful blinding

We have been asking participants their opinion about the group allocation (as a standard instrument to measure the credibility of blinding in our RCTs) since 2005. Interestingly, in all our previous RCTs on AA for perioperative analgesia,^{3 –6} the blinding had always been successful (see Table 1), probably because: (1) the participants were older and received medications with sedative properties; and/or (2) for the control procedure/sham acupuncture we used penetrating needles, but inserted them at sites not corresponding to traditional acupuncture point locations (usually the helix of the auricle).

Table 1.

Details of the randomized controlled trials where acupuncture using indwelling fixed needles was compared to placebo acupuncture and credibility of blinding was measured.

RCT year/reference	Condition	Age of participants (years)	Type of acupuncture	Type of placebo	Blinding successful (Y/N)
2005	POP after hip surgery	67 ± 11	AA	P	Y
2005	POP after knee arthroscopy	41 ± 14	AA	P	Y
2007	POP after knee arthroscopy	43 ± 13	AA	P	Y
2011	Intraoperative analgesia during hip surgery	67 ± 9	AA	P	Y
2016 2	Exam anxiety	23 ± 3	AA	NP	N
2022 1	POP after caesarean delivery	31 ± 5	AA and body acupuncture	NP	N

RCT: randomized controlled trial; Y: Yes; N: No; POP: postoperative pain; P: penetrating; AA: auricular acupuncture; NP: non-penetrating; SD: standard deviation.

Data are mean ± SD.

Suggestions for future research

Guidelines on acupuncture research do not specify the time point at which participants should be questioned in order to determine the success of blinding. ⁷ Bearing in mind that the creation of a credible placebo control and blinding in acupuncture research is extremely difficult, and taking into account the points above, we suggest that the STandards for Reporting Interventions in Clinical Trials of Acupuncture (STRICTA) guidelines should be revised to include a recommendation about the appropriate time point for the measurement of credibility of blinding. In order to generate reliable unbiased data on the opinion of patients regarding their group allocation (especially in case of indwelling fixed needles that remain in situ for a long time), we believe it is important to ask the participants at least twice: (1) shortly after placing the needles or placebo devices; and (2) around the time of the measurement of the primary outcome. We should expect a tendency for unblinding in the group that receives the more effective intervention. This could be evaluated by examining for a difference between the first and the second blinding evaluations, although this difference will also be influenced by the tendency and opportunity to be curious about the intervention, as highlighted above.