Auricular Acupressure As A Therapy For Chemotherapy-Induced Nausea And Vomiting-A Meta-Analysis

Objective:Based on the concept of evidence-based medicine with reference to the requirements of Cochrane for systematic reviews and the use of statistical methods of Meta-analysis, the randomized-controlled trials of the home and abroad databases for auricular acupressure as a therapy for chemotherapy-induced nausea and vomiting were searched and assessed. Their interventions, antiemetics and chemotherapy drugs were analyzed. According to the time and day of vomiting, the therapeutic efficacy of auricular acupressure on acute and delayed emesis was investigated. Its control on nausea and vomiting, the significance of cumulative Meta-analysis, the occurrence of adverse effects and its improvement on the quality of life were evaluated. These were to provide the evidence for therapeutic efficacy of auricular acupressure on chemotherapy-induced nausea and vomiting, and a basis for further research.Methods:According to the requirements of the Cochrane Handbook of Systematic Reviews of Interventions, electronic searches were made through the Chinese and English databases. Manual searches were made by going through the references of the clinical reports and literature reviews. The selection of studies was divided into two phases:the first phase of selection was carried out by going through the titles, abstracts, key words to eliminate those non-randomized controlled trials, duplicate reports or similar studies, or those without full texts. The second phase was conducted by reading the full texts to select studies for meta-analysis based on the inclusion and exclusion criteria. Inclusion criteria: ① Randomized controlled trials or quasi-randomized controlled trials, regardless of whether or not blinding used. ② Clinically diagnosed cancer patients with chemotherapy-induced nausea and vomiting without restriction on their age, gender, race and nationality.③ Auricular acupressure as the only invtervention or added intervention in the experimental group on the basis of antiemetic medications used in the control group.④ Primary outcome:clinical efficacy on nausea and vomiting; secondary outcomes:occurrence of adverse events or improvement on quality of life. ⑤ Studies with full texts (including effect measures) and accurate data needed for analysis. Exclusion criteria: ① Non-randomized controlled trials or studies of observed cases without control groups. ② Crossover controlled trials. ③ Animal experiments or cell/tissue researches. ④ Non-clinical trials like literature reviews, Meta-analysis, commentary etc. ⑤Case/controlled (disease group and non-disease group comparison) studies. ⑥ Duplicate publications or duplicate selected articles. ⑦ Non-Chinese or English articles. ⑧ Studies with trial data non-traceable. The assessment of the quality of studies was based on CONSORT Statements and Cochrane tools for risk of bias. The extracted data included basic characteristics, interventions, efficacy results etc. The selection of studies, the assessment of studies’quality and the extraction of data were done by two reviewers independently and cross-checked with each other. Any disagreement was resolved by discussion or through arbitration by a third person. Microsoft Excel 2010 was used to build the database for data entry and management.Meta-analysis was performed by RevMan 5.3 provided by the Cochrane Collaboration. Before carrying out a Meta-analysis, the heterogeneity of the included studies need to be assessed with reference to the Q statistic for qualifying inconsistency and the I2 statistic for quantifying inconsistency. This study used (P<0.05,I2> 50%) to indicate the existence of statistically significant heterogeneity. According to the results of heterogeneity test: combining outcomes with fixed-effect models if there was no statistically significant heterogeneity among studies; combining outcomes with random-effects models if there was statistically significant heterogeneity among studies. RR (Risk Ratio) with its 95%CI will be the effect measure. All Meta-analyses were performed with the Mantel-Haenszel methods through RevMan 5.3. The final results were shown by forest plots. While analyzing the therapeutic efficacy of auricular acupressure as an intervention to improve the condition of chemotherapy-induced nausea and vomiting and the quality of life, if the short lines and the diamond symbol of effect estimates appeared on the right-hand side of the vertical line in a forest plot, it indicated that the experimental intervention was superior to the control intervention; if they appeared on the left-hand side of the vertical line in the forest plot, it implied the experimental intervention was not superior to the control intervention. While analyzing the therapeutic efficacy of auricular acupressure as an intervention for the occurrence of adverse events, if the short lines and the diamond symbol of effect estimates appeared on the left-hand side of the vertical line in a forest plot, it indicated that the experimental intervention was superior to the control intervention; if they appeared on the right-hand side of the vertical line in the forest plot, it implied the experimental intervention was not superior to the control intervention. Sensitivity analyses were deployed to examine the stability and reliability of the results of Meta-analyses. Funnel plots was used to detect the existence of publication bias. Subgroup analyses and cumulative Meta-analyses were performed for the included studies. Discussion and conclusion were based on the results of analyses.Results:4,368 articles were retrieved from Chinese databases:4,275 articles from China National Knowledge Infrastructure (CNKI),43 articles from Chinese Biomedical Journal Citation Database (CMCI) and 50 articles from WanFang Data (WF). After browsing titles, abstracts, key words and reading full-texts,24 articles were finally included.441 articles were identified from English databases:87 articles from Cochrane Library,31 articles from Ovid MEDLINE, 170 articles from PubMed,15 articles from CINAHL Plus,8 articles from Academic Search Premier,2 articles from EBSCOHost,2 articles from ProQuest, 80 articles from BioMed Central,45 articles from ScienceDirect,1 article from Web of Science. After the first phase of selection, only 1 English article was retrieved. But because it was a crossover controlled trial, it was finally excluded in the second phase of the selection process. So the 24 included studies were all conducted in mainland China. A total of 1,911 participants were identified in the 24 included studies which covered 755 males (39.5%) and 1,156 females (60.5%) with age ranging from 20 to 82 years old.The methodological quality of the included studies was assessed: ① Randomization method:2 studies reported the use of random number table,1 study was according to the admission sequence for treatment, and all the other studies did not mention the details of randomization method. All studies did not mention the use of allocation concealment. ② Blinding:All studies did not mention whether they conducted blinding. ③ Dropout:All studies did not have dropouts. ④ Follow-up:All studies did not mention the details about follow-up. ⑤ Baseline similarity:the experimental group and the control group were comparable for 19 studies which reported no statistically significant difference (P> 0.05) in the baseline characteristics between the two groups. According to the detailed items of CONSORT Statement, the assessment of the quality of reports for randomized controlled trials of auricular acupressure as a therapy for chemotherapy-induced nausea and vomiting revealed that the quality of the reports of the trials had problems. Based on the Cochrane tools of risk of bias provided by the Cochrane Handbook for Systematic Reviews of Interventions 5.1.0, the quality assessment of the design and implementation of the randomized controlled trials revealed that the clinical trials had risk of bias. Though the majority of studies described the use of randomization, inclusion/exclusion criteria and statistical methods, they had serious reporting flaws in random sequence generation, allocation concealment and blinding. The funnel plot also suggested that the presence of publication bias in the studies. In conclusion, all included studies had unclear biases and were low quality literatures, so the credibility of the Meta-analyses was reduced.The Meta-analysis of the 24 studies with 1,021 patients in the experimental group and 890 patients in the control group was carried out with random-effects models. The combined RR=1.20,95% CI=[1.13,1.28], Z=5.83, P< 0.00001< 0.05 showed that there was statistically significant difference in the therapeutic efficacy of the two groups. The results supported that auricular acupressure could effectively improve the symptoms of chemotherapy-induced nausea and vomiting.As a single cancer-type research in this study, breast cancer cases accounted for the most (26.9% of the total cases), followed by leukemia (8.3%) and lung cancer (6.6%). Patients with a variety of cancers found in 12 studies: with lung cancer mainly (13.0%), with gastrointestinal cancer mainly (12.5%), others (21.4%). Patients without cancer-type identification in 4 other studies accounted for the rest (11.3%). Platinum based chemotherapy drugs were the most commonly used (60.8% of the total cases), followed by non-platinum based chemotherapy drugs (32.1%) and the rest with chemotherapy drugs unknown (7.1%). Ondansetron was the most commonly used antiemetic (35.3% of total cases), followed by granisetron (33.8%), tropisetron (10.3%), ondansetron or metoclopramide (9.1%), ondansetron or granisetron (6.6%), granisetron plus dexamethasone (1.7%), azasetron plus dexamethasone (1.7%), metoclopramide (1.5%).Shenmen and stomach were the most commonly selected auricular main acupoints (each 20.8% of the total 4,909 usage frequency of main acupoints), followed by sympathetic (15.4%), subcortex and spleen (each 6.9%), cardia (4.6%), diaphragm (4.’1%), liver (3.6%), endocrine and large intestine (each 2.9%), and the other 9 acupoints (esophagus, lung, heart, brain point, constipation point, rectum, sanjiao, small intestine, tumor corresponding part) accounted for 11.1%. Liver was the most commonly selected auricular adjunct acupoint (37.1% of the total 1,263 usage frequency of adjunct acupoints), followed by spleen (33.6%), gallbladder (10.4%), subcortex (5.5%), lung (3.4%), large intestine (3.2%), and the other four acupoints (kidney, heart, pancreas, adrenal gland) accounted for 6.8%.For subgroup analysis based on the type of interventions:Comparing simple auricular acupressure with antiemetics, combined RR= 1.11,95% CI= [1.00, 1.24], Z= 1.88, P=0.06> 0.05, the difference in efficacy was not statistically significant. Comparing auricular acupressure plus antiemetics with antiemetics, combined RR= 1.22,95% CI= [1.14,1.31], Z= 5.57, P< 0.00001, the difference in efficacy was statistically significant. It meant that the therapeutic efficacy of auricular acupressure was not superior to that of antiemetics. But on the basis of antiemetics, with the use of auricular acupressure may strengthen the antiemetic effect.For the comparison between "simple auricular acupressure and antiemetics", including 4 studies with experimental group of 187 patients and control group of 190 patients:Shenmen, stomach and sympathetic were the most commonly selected auricular main acupoints (each 22.0% of the total 852 usage frequency of main acupoints), followed by subcortex and spleen (each 7.0%), diaphragm and cardia (each 6.7%), esophagus (6.6%). Liver was the most commonly selected auricular adjunct acupoint (40.9% of the total 311 usage frequency of adjunct acupoints), followed by spleen (18.3%), subcortex (22.5%), gallbladder (18.3%).For the comparison between "auricular acupressure plus antiemetics and antiemetics", including 22 studies with experimental group of 834 patients and control group of 700 patients:Shenmen and stomach were the most commonly selected auricular main acupoints (each 20.6% of the total 4,057 usage frequency of main acupoints), followed by sympathetic (14.0%), subcortex and spleen (each 6.9%), liver (4.4%), cardia (4.2%), endocrine, large intestine and diaphragm (each 3.5%), and the other 9 acupoints (esophagus, lung, heart, brain point, constipation point, rectum, sanjiao, small intestine, tumor corresponding part) accounted for 11.9%. Spleen was the most commonly selected auricular adjunct acupoint (38.7% of the total 952 usage frequency of adjunct acupoints), followed by liver (35.8%), gallbladder (7.8%), lung (4.5%), large intestine (4.2%), and the other four acupoints (kidney, heart, pancreas, adrenal gland) accounted for 9.0%.For subgroup analysis based on the type of antiemetics, the antiemetic effect of the auricular acupressure group (or plus antiemetics) was superior to that of metoclopramide group (combined RR= 1.44,95% CI= [1.15,1.80], Z= 3.17, P= 0.002), ondansetron group (combined RR= 1.13,95% CI= [1.03, 1.24], Z= 2.63, P= 0.008), granisetron group (combined RR= 1.29,95% CI = [1.13,1.46], Z= 3.89, P= 0.0001) and ondansetron/granisetron group (combined RR= 1.20,95% CI= [1.01,1.44], Z= 2.02,P=0.04), all P< 0.05, there were statistically significant differences in efficacy.For antiemetics like dopamine receptor antagonists (eg, metoclopramide), there were 2 studies with experimental group of 51 patients and control group of 66 patients. Shenmen, stomach and sympathetic were the most commonly selected auricular main acupoints (each 21.9% of the total 233 usage frequency of main acupoints), followed by subcortex, liver and spleen (each 6.9%), cardia and tumor corresponding part (each 6.8%). Liver and subcortex were the most commonly selected auricular adjunct acupoints (each 50.0% of the total 70 usage frequency of adjunct acupoints).For antiemetics like serotonin receptor antagonists (eg. ondansetron, granisetron, tropisetron, azasetron), there were 21 studies with experimental group of 938 patients and control group of 792 patients. Shenmen and stomach were the most commonly selected auricular main acupoints (each 21.2% of the total 4,420 usage frequency of main acupoints), followed by sympathetic (15.3%), subcortex (7.3%), spleen (6.6%), diaphragm (4.5%), cardia (4.0%), liver, endocrine and large intestine (each 2.9%), esophagus (2.6%), and the other six acupoints (lung, heart, brain point, constipation point, rectum, sanjiao) accounted for 8.6%. Liver was the most commonly selected auricular adjunct acupoint (38.4% of the total 1,129 usage frequency of adjunct acupoints), followed by spleen (37.6%), gallbladder (10.2%), large intestine (3.5%), subcortex (3.1%), lung, kidney and heart (each 2.4%).For hybrid antiemetics like serotonin receptor antagonists with glucocorticoids (eg dexamethasone), there were 2 studies with experimental group and control group each of 32 patients. Shenmen, stomach, sympathetic, liver, spleen and cardia were the most commonly selected auricular main acupoints (each 12.5% of the total 256 usage frequency of main acupoints), followed by endocrine, large intestine, small intestine and esophagus (each 6.25%). Lung, gallbladder, pancreas and adrenal gland were the most commonly selected auricular adjunct acupoints (each 25.0% of the total 64 usage frequency of adjunct acupoints).For subgroup analysis based on the type of chemotherapeutic drugs, the antiemetic effect of the auricular acupressure group (or plus antiemetics) was superior to that of antiemetic group for all types of chemotherapeutic drugs:platinum based (combined RR= 1.14,95% CI= [1.08,1.21], Z=4.70, P<0.00001); non-platinum based (combined RR= 1.24,95% CI= [1.08,1.42], Z= 3.09, P= 0.002); unknown (combined RR= 1.73,95%CI= [1.32,2.26], Z= 3.99, P< 0.0001), all P< 0.05, there were statistically significant differences in efficacy.For platinum based chemotherapeutic drugs, there were 15 studies with experimental group of 632 patients and control group of 531 patients. Shenmen and stomach were the most commonly selected auricular main acupoints (each 21.9% of the total 2,889 usage frequency of main acupoints), followed by sympathetic (15.9%), subcortex (8.1%), diaphragm (6.9%), spleen (5.5%), cardia (5.1%), large intestine (5.0%), esophagus (4.5%), lung (2.9%), liver (1.1%), endocrine and small intestine (each 0.6%). Liver was the most commonly selected auricular adjunct acupoint (39.8% of the total 875 usage frequency of adjunct acupoints), followed by spleen (31.8%), gallbladder (15.0%), subcortex (8.0%), lung, pancreas and adrenal gland (each 1.8%).For non-platinum based chemotherapeutic drugs, there were 6 studies with experimental group of 316 patients and control group of 296 patients. Shenmen and stomach were the most commonly selected auricular main acupoints (each 18.6% of the total 1,694 usage frequency of main acupoints), followed by sympathetic (14.9%), spleen (9.6%), endocrine (7.6%), liver (6.0%), and cardia, heart, brain point, constipation point and rectum (each 3.8%), subcortex (3.5%), sanjiao (2.2%). Liver and spleen were the most commonly selected auricular adjunct acupoints (each 40.9% of the total 220 usage frequency of adjunct acupoints), followed by large intestine (18.2%).For chemotherapeutic drugs unknown, there were 3 studies with experimental group of 73 patients and control group of 63 patients. Shenmen and stomach were the most commonly selected auricular main acupoints (each 22.4% of the total 326 usage frequency of main acupoints), followed by sympathetic (14.1%), subcortex and liver (each 13.2%), spleen, cardia and tumor corresponding part (each 4.9%). Spleen was the most commonly selected auricular adjunct acupoint (33.9% of the total 168 usage frequency of adjunct acupoints), followed by liver (17.8%), lung, kidney and heart (each 16.1%).For subgroup analysis based on time of vomiting,4 studies involved 162 patients in the experimental group and 147 patients in the control group. In respect of acute and delayed emesis, compared the auricular acupressure plus antiemetics group with the antiemetic group, the difference was not statistically significant in therapeutic efficacy. For acute emesis, combined RR=1.13,95% CI=[0.93,1.38], Z= 1.24, P=0.21; for delayed emesis, combined RR= 1.38,95%CI= [0.99,1.91], Z= 1.92, P= 0.05. Comparing the value of such RRs, it could be seen that when auricular acupressure plus antiemetics compared with antemetics, its control of delayed emesis was superior to its control of acute emesis. For acute and delayed emesis, shenmen and stomach were the most commonly selected auricular main acupoints (each 24.5% of the total 661 usage frequency of main acupoints), followed by subcortex (18.0%), sympathetic (15.9%), spleen and large intestine (each 6.5%), liver (4.1%). Spleen was the most commonly selected auricular adjunct acupoint (38.4% of the total 232 usage frequency of adjunct acupoints), followed by liver (26.8%), lung, kidney and heart (each 11.6%).For subgroup analysis based on day of vomiting,2 studies involved experimental group and control group each of 85 patients. Day 1:combined RR = 1.00,95%CI= [0.90,1.11], Z= 0.00, P= 1.00; Day 2:combined RR= 1.24, 95%CI= [1.08,1.43], Z= 2.97, P= 0.003; Day 3:combined RR= 1.20,95%CI = [1.05,1.36], Z= 2.75, P= 0.006; Day 4:combined RR= 1.22,95%CI= [1.06, 1.41], Z= 2.76, P=0.006; Day 5:combined RR= 1.29,95%CI= [1.10,1.52], Z=3.07,P= 0.002; Day 6:RR= 1.34,95% CI= [1.11,1.63], Z= 2.97,P= 0.003; Day 7:RR= 1.31,95% CI= [1.08,1.57], Z= 2.80, P= 0.005. Compared the therapeutic efficacy of auricular acupressure plus antiemetics with that of antiemetics, except for the first day in which there was no statistically significant difference, P< 0.05 for all the other days implied that in respect of delayed emesis, there were statistically significant differences. The results of Meta-analysis supported that when auricular acupressure plus antiemetics compared with antiemetics, its control of delayed emesis was superior to its control of acute emesis. Based on the date of vomiting, shenmen, stomach and sympathetic were the most commonly selected auricular main acupoints (each 33.3% of the total 255 usage frequency of main acupoints). Liver and spleen were the most commonly selected auricular adjunct acupoints (each 50.0% of the total usage 170 frequency of adjunct acupoints).For Meta-analysis based on the nature of vomiting,1 study involved experimental group and control group each of 60 patients. In respect of nausea and vomiting, there was no statistically significant difference in the therapeutic efficacy of auricular acupressure compared with that of antiemetics. But comparing nausea RR= 1.23,95% CI= [0.94,1.60], Z= 1.51, P= 0.13 with vomiting RR= 0.94,95% CI= [0.76,1.15], Z= 0.64, P= 0.52, it could be seen that when auricular acupressure compared with antemetics (ondansetron), its control of nausea was superior to its control of vomiting. For nausea and vomiting, shenmen, stomach, sympathetic, subcortex and spleen were the most commonly selected auricular acupoints.According to the sequence of publication year of articles, cumulative Meta-analysis was carried for those studies existed in each time point. Using random-effects models for analysis, at a=0.05 level, starting from 2006, the overall effect measures of all time points were with statistical significance. The added number of cases made RR estimates stabilize and its confidence interval become narrower. The results of cumulative Meta-analysis supported that auricular acupressure could effectively improve chemotherapy-induced nausea and vomiting.Meta-analysis suggested that auricular acupressure could improve the adverse effects caused by antiemetics in various degrees, such as constipation (combined RR= 0.35,95% CI= [0.18,0.65], Z= 3.30, P= 0.001), fatigue (combined RR= 0.39,95% CI= [0.26,0.58], Z= 4.64, P< 0.00001), hiccup (RR= 0.40,95% CI= [0.20,0.77], Z= 2.74, P= 0.006), diarrhea (RR= 0.60, 95% CI= [0.37,0.97], Z=2.09, P=0.04) and poor appetite (RR=0.27,95% CI=[0.17,0.42], Z=5.85, P<0.00001) etc. But for bloating (combined RR= 0.62,95% CI= [0.23,1.71], Z= 0.92, P= 0.36), lack of energy (combined RR= 0.73,95% CI= [0.28,1.86], Z= 0.67, P= 0.51), abdominal pain (RR= 0.40,95% CI=[0.11,1.40], Z=1.44, P=0.15), headache (RR=0.20,95% CI= [0.01,3.95], Z=1.06, P=0.29), mild fever (RR= 0.60,95%CI= [0.15, 2.40], Z=0.72, P=0.47), mild liver damage (RR= 0.83,95% CI= [0.27,2.58], Z= 0.32, P= 0.75), etc., there were no statistical significances in efficacy.For constipation,4 studies compared auricular acupressure (or plus antiemetics) with antiemetics involving experimental group of 252 patients and control group of 206 patients. Shenmen, stomach and spleen were the most commonly selected auricular acupoints (each 14.4% of the total 1,750 usage frequency of acupoints), followed by subcortex (8.2%), large intestine (7.3%), sympathetic (5.8%), lung and diaphragm (each 4.8%), endocrine, liver, cardia, rectum, constipation point, heart and brain point (each 3.7%).For fatigue,2 studies compared auricular acupressure (or plus antiemetics) with antiemetics involving experimental group of 103 patients and control group of 102 patients. Shenmen, stomach, spleen and sympathetic were the most commonly selected auricular acupoints (each 20.0%of the total 515 usage frequency of acupoints), followed by subcortex (11.7%), large intestine (8.3%).For hiccups, diarrhea and poor appetite,1 study compared auricular acupressure plus antiemetics with antiemetics involving experimental group of 85 patients and control group of 42 patients. Shenmen, stomach, spleen, subcortex, large intestine, lung and diaphragm were selected auricular acupoints.1 study involving experimental group of 85 patients and control group of 42 patients suggested that auricular acupressure plus antiemetics compared with antiemetics could be more effective in the improvement on the quality of life of patients, RR=1.39,95% CI= [1.12,1.74], Z= 2.94, P= 0.003 < 0.05. For the improvement on the quality of life, Shenmen, stomach, spleen, subcortex, large intestine, lung and diaphragm were selected auricular acupoints.Conclusion:According to the principles of evidence-based medicine with reference to the requirements of Cochrane for systematic reviews and through the use of Meta-analysis, it showed that auricular acupressure as a therapy for chemotherapy-induced nausea and vomiting had certain therapeutic efficacy and was worthy of further clinical studies. Comparing simple auricular acupressure with antiemetics, there was no statistically significant difference in their therapeutic efficacy. But the use of auricular acupressure plus antiemetics can enhance the antiemetic effect to a certain extent. Meta-analysis revealed that in the prevention and treatment of delayed emesis, auricular acupressure plus antiemetics may be superior to simply using antiemetics. And in the prevention and treatment of nausea, auricular acupressure might be superior to simply using antiemetics. Meta-analysis suggested that auricular pressure might have various degrees of improvement on some adverse effects caused by antiemetics and might be able to improve the quality of life of patients.As the quality of the included studies was not high with selection, performance and detection bias, this might lower the credibility of the results of this study. Further large-scale multi-center clinical studies should be carried out to establish the data for evidence-based medicine and provide reliable grounds for clinical application.