Acupuncture for analgesia in the emergency department: a multicentre, randomised, equivalence and non-inferiority trial

In reply:

We dispute Wenden’s assertions that the patients who chose to participate in our trial1 are statistically more inclined to experience placebo benefit from acupuncture and that a 62% participation rate is a “significant flaw” that introduces “critical bias”. Clinical emergency department (ED) trials have specific challenges,2 as patients arriving in pain to an ED do not expect to be asked to enrol in a clinical trial, and many do not have the capability or patience to hear, understand and agree to the implications of participating in a trial, irrespective of their opinion of the intervention being tested. Moreover, our participation rate is comparable with or better than other ED studies,3-5 and while the real world, pragmatic nature of our trial meant that we did not have the luxury of eliminating all potential sources of bias, the fact that the majority of patients chose to participate means that selection bias is unlikely to be a significant factor.

With respect to Wenden’s comments about a “broad interpretation of non-inferiority”, we formally tested for non-inferiority at one hour and have reported on this. What we observed afterwards does not necessarily render the acupuncture-alone option up to one hour unacceptable, even though there was a 25% versus 15% versus 15% difference in rescue therapy at one hour between the treatment groups (P = 0.016); however, this was taken into account according to our per protocol analysis. Patients with migraine were given minimal discussion because of power issues. To suggest that pain score at one hour is a more important outcome measure than patient satisfaction is an opinion that does not align with current evidence on pain management in the ED. It has been shown that pain scores do not correlate with satisfaction,6 and it remains to be seen which is a more accurate reflection of adequacy of analgesia.

While it is true that we did not demonstrate diminished opioids use, the follow-up was only for 2 days and it is known that patients often leave EDs with scripts for opioids and that many of the conditions we studied have a high opioid addiction risk. Any effective alternate analgesic option should therefore be considered and offered in their place. In addition, providing acupuncture in an ED does not require any diversion of resources. Emergency physicians are adept at picking up new skills (ultrasound being a far more complex example), and training them to provide this relatively quick and simple intervention would require no additional staff or resources, as equipment costs are negligible.