Results Characteristics of eligible studies Forty-four studies involving 5022 participants were included. The characteristics of the included studies are summarized in Supplementary Material Table 1. The earliest study was conducted in 1990, while the latest one was in 2016. The interventions included auricular acupressure and bloodletting in the auricle. The duration of the trials ranged from 20 min to 12 weeks. One trial28 made a comparison between auricular acupressure and sham acupoint, nine trials29–37 between auricular acupressure and ADs, five trials36,38–41 between auricular acupressure and no intervention, 25 trials42–66 between auricular acupressure plus antihypertensive drugs (AAPADs) and ADs alone, one trial70 between bloodletting in the auricle and no intervention, two trials68,71 between bloodletting in the auricle and ADs, three trials67–69 between bloodletting in the auricle plus ADs and ADs alone. In general, all of the trial patients had an average age of 40 years or more. Methodological quality All included trials were reported as parallel-group randomized trials, while only 17 trials (38.63%, 17/44) reported the method of sequence generation and, unfortunately, no trial described the process of allocation concealment. Only three trials mentioned the blinding process. For selective reporting, we made judgements by comparing the outcome measures mentioned in the methods section with the results from the original references: all trials were evaluated as low risk. Only four trials mentioned the drop-out of patients. No trial provided a pre-trial sample size estimation. The methodological quality is shown in Supplementary Material Figure 1. Effects of the interventions The graded quality of evidence in terms of the outcomes of SBP change, DBP change, and efficacy rate were identified for 25 RCTs (Table 1). Although a total of 44 RCTs were included in this review, of these only 25 RCTs were combined by meta-analysis due to their consistent comparison and high homogeneity, while others were analyzed by qualitative description which could not be given a graded quality of evidence. Table 1. Grading of Recommendations Assessment, Development, and Evaluation (GRADE) summary of 25 randomized controlled trials (RCTs) comparing auricular acupressure plus antihypertensive drugs (AAPADs) to antihypertensive drugs (ADs) alone in hypertension patients. Patient or population: patients with hypertensionSettings: inpatients/outpatientsIntervention: AAPADComparison: AD Outcomes Illustrative comparative risks a (95% CI) Relative effect(95% CI) No. of participants(studies) Quality of the evidence(GRADE) Comments Assumed risk Corresponding risk AD AAPAD SBP change magnitude between baseline and post-intervention The mean SBP change magnitude between baseline and post-intervention in the control groups was 135.73 The mean SBP change magnitude between baseline and post-intervention in the intervention groups was 5.06 lower (6.76 to 3.36 lower) 929(10 studies) ⊕⊕⊕⊝Moderateb DBP change magnitude between baseline and post-intervention The mean DBP change magnitude between baseline and post-intervention in the control groups was 84.75 The mean DBP change magnitude between baseline and post-intervention in the intervention groups was 5.3 lower (6.27 to 4.33 lower) 929(10 studies) ⊕⊕⊕⊝Moderateb Efficacy rate Study population RR 1.22 (1.17–1.26) 2017 (21 studies) ⊕⊕⊝⊝Lowb,c 731 per 1000 892 per 1000 (855–921) Moderate 740 per 1000 903 per 1000 (866–932) CI: Confidence interval; DBP: diastolic blood pressure; RR: risk ratio; SBP: systolic blood pressure; ROB: risk of bias. GRADE Working Group grades of evidence: High quality: further research is very unlikely to change our confidence in the estimate of effect; Moderate quality: further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate; Low quality: further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate; Very low quality: we are very uncertain about the estimate. a The basis for the assumed risk (e.g. the median control group risk across studies) is provided in these footnotes. The corresponding risk (and its 95% CI) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). b According to the ROB graph (Supplementary Material Figure 1). c Funnel plot was asymmetrical. Effect of auricular acupressure As for the comparison of auricular acupressure and sham acupoint, only one trial28 reported the efficacy rate in terms of this comparison, and found a statistical difference between auricular acupressure and sham acupoint, in favor of the experimental group (n=55 out of 60 patients; RR, 0.28; 95% CI, 0.10–0.47, p=0.003). Moreover, two trials reported the magnitude of BP change between baseline and post-intervention comparing auricular acupressure and ADs. One trial34 comparing auricular acupressure to nifedipine gastrointestinal therapeutic system (GITS) found auricular acupressure was superior to ADs in reducing SBP (n=118 patients; MD, −0.92; 95% CI, –1.09– −0.75, p<0.00001), but there was no significant difference between two groups on decreasing DBP (n=118 patients; MD, −0.07; 95% CI, –0.23– −0.09, p=0.40). However, the other trial36 comparing auricular acupressure to a kind of AD (the specific medicine is unclear) showed the AD had better effect on reducing SBP (n=203 patients; MD, 2.70; 95% CI, 2.37–3.03, p<0.00001) as well as on decreasing DBP (n=203 patients; MD, 2.50; 95% CI, 1.92–3.08, p<0.00001). In addition, five trials reported the outcome by measuring the SBP and DBP before and after the treatment, three found a significant difference between auricular acupressure and ADs in lowering SBP after treatment, in favor of the auricular acupressure group,29,31,33 and the others did not. The data of these three trials could not be pooled due to the different kinds of ADs used in the AD groups. Four of five trials found auricular acupressure significantly effective in reducing DBP after treatment,29,31,34,37 and the data of these trials could not be pooled either due to high heterogeneity. Eight trials reported the efficacy rate by counting the number of those who had achieved BP targets after treatment. The meta-analysis from these eight trials showed that there is no statistical difference between groups in efficacy rate (n=495 out of 598 patients; RR, 0.99; 95% CI, 0.95–1.03, p=0.62; I2=0%). Furthermore, three trials reported the result from the comparison of auricular acupressure and no intervention, and found a significant difference between auricular acupressure and no intervention group in decreasing SBP, in favor of auricular acupressure,38,39,41 nevertheless the data of these trials could not be pooled due to high heterogeneity. In terms of DBP, two of three trials found a significant difference between the two groups and the other did not. Five trials reported the outcome of BP target achievement, and meta-analysis from four trials38–41 found a statistical difference between groups in achieving BP targets in favor of the auricular acupressure group (n=247 out of 238 patients; RR, 1.27; 95% CI, 1.05–1.55, p=0.02; I2=0%). Effect of auricular acupressure combined with ADs Within the comparison between AAPADs and ADs alone, 14 trials42,44,46,47,49,52,53,57,59–62,65,66 reported the BP value before and after treatment. Even when we used a random-effect model, heterogeneity was too large (I2=82%), which might have been due to clinical heterogeneity or low methodological quality. Then, we tried to find the existence of this heterogeneity, and found that if we excluded four trials44,47,52,61 with outliers of larger BP value before treatment, the meta-analysis showed AAPADs had a better effect on reducing SBP than ADs alone (n=464 patients; MD, −5.06; 95% CI –6.76– −3.36, p<0.00001; I2=32%) as well as DBP (n=464 patients; MD, −5.30; 95% CI –6.27– −4.33, p<0.00001; I2=0%), which are shown in Figures 2 and 3. Twenty-one trials reported the efficacy rate, and the efficacy rate in the AAPAD group was significantly higher than that in the control group with ADs alone (RR, 1.22; 95% CI, 1.17–1.26; p<0.00001; I2=0%), which is shown in Figure 4. Figure 2. Forest plot of the comparison between auricular acupressure plus antihypertensive drugs (AAPADs) vs antihypertensive drugs (ADs) alone for the outcome systolic blood pressure (SBP) after treatment. CI: confidence interval; SD: standard deviation. Figure 3. Forest plot of the comparison between auricular acupressure plus antihypertensive drugs (AAPADs) vs antihypertensive drugs (ADs) alone for the outcome diastolic blood pressure (DBP) after treatment. CI: confidence interval; SD: standard deviation. Figure 4. Forest plot of the comparison between auricular acupressure plus antihypertensive drugs (AAPADs) vs antihypertensive drugs (ADs) alone for the outcome efficacy rate. CI: confidence interval; SD: standard deviation. Effects of bloodletting Two trials reported the efficacy rate comparing bloodletting in the auricle to ADs, and meta-analysis from these two trials showed that there is no statistical difference between the groups in efficacy rate (RR, 0.65; 95% CI, 0.28–1.51, p=0.32; I2=0%). One trial compared bloodletting in the auricle to no intervention, and found the efficacy rate in the experimental group significantly higher than that in the control group (n=59 out of 61 patients; RR, 5.1; 95% CI 1.05–24.71; p=0.04), and also found a significant difference between the two groups in lowering both SBP (n=61 patients; MD, −8.77; 95% CI –10.20– −7.34, p<0.00001) and DBP (n=61 patients; MD, −5.50; 95% CI –7.24– −3.76, p<0.00001). As for the comparison of bloodletting in the auricle plus ADs versus ADs alone, three trials reported the efficacy rate, and two of them found a statistical difference between groups, one69 compared bloodletting in the auricle plus ADs to felodipine extended-release tablet or irbesartan alone (n=156 out of 178 patients; RR, 1.37; 95% CI, 1.21–1.55, p<0.00001); one study67 used nifedipine in the control group (n=57 out of 60 patients; RR, 1.21; 95% CI, 1.05–1.40, p=0.009). One trial used amlodipine besylate tablets in the control group and did not find a statistical difference between groups.68 Safety A total of 42 trials reported on this outcome. In all, four incidents were reported in the bloodletting in the auricle group, while four incidents were reported in the AD group. Side reactions included: hematoma at the bloodletting auricle (n=2 patients, 25%), red and swollen at the bloodletting auricle (n=1 patient, 12.5%), headache (n=4, 50%), palpitation (n=1 patient, 12.5%) and ankle edema (n=1 patient, 12.5%). Extent of publication bias Conducting the funnel plot (Figure 5) for SBP and DBP after treatment between AAPADs and ADs alone suggested that no evidence of publication bias could be detected. Figure 5. Funnel plot of 14 trials comparing (a) systolic blood pressure (SBP) and (b) diastolic blood pressure (DBP) after treatment. MD: mean difference; SE: standard error.