Effectiveness of acupuncture for pain control after Cesarean section is associated with the length of experience of the acupuncturist: observational data from a randomized clinical trial

Background

In a recent randomized controlled investigation, acupuncture with indwelling intradermal needles reduced pain and accelerated mobilization in patients undergoing elective Cesarean section (CS). ¹ In this investigation, acupuncture administered bilaterally at four traditional auricular acupuncture (AA) point locations and six traditional acupuncture point locations of the body (n = 60) was compared with placebo acupuncture (n = 60) at the same sites using non-penetrating placebo devices. ² Both study interventions (acupuncture and the placebo procedure) were provided by three physicians— C.K. (resident radiologist with 1 year of clinical experience in radiology), T.H. and T.U. (both consultant anaesthesiologists with more than 20 years of experience in their clinical specialty) with various durations of expertise in acupuncture— according to previously described protocols.^1,2 One hundred and twenty patients were included between January 2015 and June 2017. Initially it was intended that the acupuncturist with the longest duration of experience (TU, >25 years) would perform the acupuncture for all study participants with assistance from the second acupuncturist (TH, <2 years’ experience in acupuncture). However, due to academic leave, TU was absent for several months in 2016 and, consequently, CK and TH had to perform the study interventions. In total, TU and TH (Team 1) performed study interventions in 90 patients, while CK and TH (Team 2) performed study interventions in 30 patients. In this brief report, we present a re-calculation of the main results of the original randomized investigation, stratified by the acupuncturist team who applied the treatment.

Findings

For the primary endpoint of the study, the pain intensity on movement on the first postoperative day, measured using the 11-point verbal rating scale (VRS-11), the mean difference between acupuncture and placebo groups where these interventions were provided by Team 1 reached −1.5 points (95% confidence interval (CI) −2.23 to −0.65; p < 0.001; effect size (ES) calculated as Cohen’s d 0.8; Table 1). Patients who received study interventions from Team 2 reported a mean difference between acupuncture and placebo of −0.76 points (95% CI: −2.16 to 0.64; p = 0.3; ES = 0.4).

Table 1.

Results according to acupuncturist team performing the study intervention.

	Acupuncturist team
	Team 1 (TU + TH)	Team 2 (CK + TH)
Total number of patients receiving acupuncture/placebo	45/45	15/15
Pain intensity on movement (first postoperative day)
Mean difference acupuncture vs. placebo (SEM)	1.5 (0.41)	0.76 (0.72)
Cohen’s d (95% CI)	−0.8 (−1.2 to −0.3)	−0.4 (−1.2 to 0.3)
P value (Students’ t-test)	<0.001	0.3
Mobilization on day of CS, number of patients receiving acupuncture/placebo
None	11/29	8/9
Sitting up	19/12	6/4
Standing up	12/0	1/0
Ambulation	3/2	0
P value (chi-square test)	<0.001	0.5
Mobilization on postoperative day 1, number of patients receiving acupuncture/placebo
Sitting up	0	0
Standing up	1/8	0/1
Ambulation	44/35	15/12
P value (chi-square test)	0.011	0.27
Foley out, number of patients receiving acupuncture/placebo
Day of CS	2/0	0
Postoperative day 1	42/31	13/10
Postoperative day 2	1/12	2/3
P value (chi-square test)	0.003	0.5

CS: Cesarean section; CI: confidence interval; SEM: standard error of the mean.

Seventy-six percent of patients who received acupuncture from Team 1 mobilized on the day of CS versus 32% of patients who received placebo (p < 0.001; ES calculated as relative risk 2.5; Table 1; Figure 1). In comparison, 47% of patients who received acupuncture from Team 2 mobilized on the day of CS versus 34% of patients who received placebo (p = 0.46; ES = 1.7; Table 1). On the first postoperative day, patients who received acupuncture from Team 1 mobilized more quickly and had their Foley catheter removed sooner than the placebo group, whereas there were no statistically significant differences between acupuncture and placebo among patients who received these interventions from Team 2.

Figure 1.

Results of patient mobilization on the day of Cesarean section (CS). Numbers within the bars represent the total number of patients per group.

The smaller group size (n = 30) receiving interventions from Team 2 (less experienced in acupuncture) might be the reason for the lack of statistical significance in the difference between acupuncture and placebo. However, observing the differences between acupuncture and placebo in patients treated by Team 1 versus Team 2, we note that Team 1, where acupuncture was provided by an experienced acupuncturist, achieved a clinically relevant reduction of postoperative pain, but this was not the case in the group treated by the less-experienced team. Clearly no firm conclusions can be drawn from this post hoc subgroup analysis of our data; however, this does raise an interesting hypothesis about the relevance of practitioner experience that runs counter to the findings of Vickers et al., ³ who performed an individual patient data meta-analysis of acupuncture for chronic pain that included 39 randomized clinical trials (RCTs) with 20,827 patients. They found that the experience of the practitioner did not correlate with outcomes in their univariate metaregression analysis. The differences in our results compared with those of Vickers et al. may be explained by the population, setting and/or indication (that is, acute postoperative pain versus chronic pain). For example, it could be that the technique required for acute surgical pain relief is more demanding, or the confidence of the experienced practitioner has a more powerful context effect in this setting.