Inspired by Ferreira et al 2023.
NICE – National Institute for Health and Care Excellencekey to acronyms
SR – systematic review
SNRIs – serotonin-norepinephrine reuptake inhibitors
CI – confidence interval
MCID – minimum clinically important difference
GDG – guideline development group
ACT – acceptance commitment therapy
CBT – cognitive behaviour therapy
MSK – musculoskeletal
Many acupuncturist readers will be aware that NG193, the NICE guideline for chronic primary pain published 07 April 2021, recommended acupuncture, albeit with funding limitations. Some (especially prescribers) will also have joined the general surprise that accompanied the explicit recommendation in NG193 against the use of pain medicines. Antidepressants were the exception. No doubt it was the latter that stimulated this overview of SRs of antidepressants for pain.
26 SRs were included and exactly half of them were Cochrane reviews. They included a total of 156 unique trials and over 25k patients. Eight different antidepressant classes covering 22 pain conditions resulted in 42 distinct comparisons in the overview.
70 of the 156 trials had ties to industry.
No review provided high certainty evidence on the efficacy of antidepressants for any pain condition. Moderate certainty evidence was found for SNRIs in back pain, postoperative pain, neuropathic pain, and fibromyalgia. Remember that we are talking about drug trials here, so there should be no difficulty in blinding (unlike in acupuncture trials). What else can reduce ‘certainty’? Well, my guess here is ‘imprecision’. That is when the CI of the pooled result in a meta-analysis crosses into the minimum clinically important difference (MCID) from the line of no effect. In this case the pooled effect in terms of mean difference did not even reach the level most would define as a MCID (10 points on a 0 to 100 pain score). In other words, we cannot be very confident because the effect is too small to ever be convincing.
Indeed, the best result in terms of effect size for antidepressants was a mean difference of -7.3 (-12.9 to -1.7) over placebo (postoperative pain). In contrast, the effect size for acupuncture in a variety of chronic pain conditions taken from NG193 (Appendix E1 Figure 2) was -14.1 (-21.1 to -7.2) over sham [I converted the figures to match the 0 to 100 scale].
Cathy Stannard, who was the clinical lead on NG193, and Colin Wilkinson, a lay member on the GDG, wrote an editorial in the BMJ to accompany this overview of SRs. I like their subtitle: Compassionate relationships with clinicians, not medicines, are the foundation of effective care.
I was not so happy that exercise and psychological therapies got a mention in the editorial but not acupuncture, particularly when the effect size for acupuncture against sham is considerably more impressive than that for exercise, ACT or CBT against no treatment.
I guess the bottom line is that we don’t have any really effective treatments for chronic pain, and there are no drugs that reach a clinically meaningful level of efficacy over placebo. Conventional non-pharmacological treatments have no evidence of efficacy over shams and variable evidence of effectiveness that rarely reaches clinically meaningful levels. By comparison acupuncture is pretty good, but the orthodox bias against it always seems to relegate it to second or third place, and funding within public services is woefully inadequate.
I saw a new patient the other day who wanted another opinion on this chronic dorsal back and arm pain. He was a professional man in his 40’s with no hard neurological signs and MSK findings consistent with modest facet arthrosis and muscle tension in lower trapezius. He had already been offered cervical spine surgery and pregabalin, two things that are notably absent from NG193.
1 Ferreira GE, Abdel-Shaheed C, Underwood M, et al. Efficacy, safety, and tolerability of antidepressants for pain in adults: overview of systematic reviews. BMJ 2023;380:e072415. doi:10.1136/bmj-2022-072415
2 Stannard C, Wilkinson C. Rethinking use of medicines for chronic pain. BMJ 2023;380:p170. doi:10.1136/bmj.p170
You must be logged in to post a comment.