Wenxin granule for cardiac arrhythmia:an overview of systematic reviews

Ruijin Qiu,Changming Zhong,Huichan Yuan,Xiaoyi Tang,Ya Huang,Tianmai He,Songjie Han,Manke Guan,Min Li,Jiayuan Hu,Xiaoyu Zhang, Jing Chen*,Hongcai Shang,4*

1Key Laboratory of Chinese Internal Medicine of Ministry of Education and Beijing, Dongzhimen Hospital, Beijing University of Chinese Medicine, Beijing, China.2The First Hospital of Tsinghua University, Beijing, China. 3Baokang Affiliated Hospital of Tianjin University of Traditional Chinese Medicine, Tianjin, China. 4School of Medical Information Engineering,Guangzhou University of Chinese medicine.Guangzhou,China.

Background

Cardiac arrhythmia is defined as the normal heart rhythm disturbed and manifested as abnormal heart rate, which includes atrial arrhythmias, ventricular arrhythmias and junctional arrhythmia. Atrial fibrillation is considered as the most common type of cardiac arrhythmia, which has an increasing morbidity in aged population and results in a substantial risk of thromboembolism and stroke [1,2].Ventricular arrhythmias may manifest palpitation,presyncope or syncope,which is related to sudden cardiac death[3].

At present, pharmaceutical drugs, implanted rhythm-restoration devices and catheter ablation are three major types of therapeutic methods for cardiac arrhythmia [4]. But limitations exist in these therapies,which may trigger or exacerbate arrhythmia, moreover,contraindications and adverse events hinder the widespread use [3]. Therefore, antiarrhythmic drugs with less side effects and better efficacy are important for patients with arrhythmia.

Wenxin granule (WenxinKeli) was initiated since 1980s in Guang’anmen Hospital, China Academy of Chinese Medicine Sciences and approved by China Food and Drug Administration (CFDA) in 1995. Wenxin granule is one of the most popular Chinese patent medicines for treating cardiac arrhythmia, which sold more than 0.2 billion boxes and 0.72 billion dollars until 2014[5].

Wenxin granule was extracted from Chinese herbal medicine, which comprised Danshen (Salvia miltiorrhizaBge.), Huangjing (Polygonatumsibiricum),Gansong(Nardostachysjatamansi DC.),Hupo(Ambrum),Sanqi (Panaxnotoginseng (Burk.)F.H.Chen). In recent years,the efficacy of Wenxin granule has been proved by experimental studies and clinical trials.

Experiments showed that Wenxin granule had the effect of inhibiting sodium-channel [6,7],inhibiting transient outward potassium current (Ito) and L-type calcium [8,9], regulating the signal transduction pathway of calcium/calmodulin-dependent protein kinase II [10].A lot of randomized controlled trials (RCTs) showed that Wenxin granule had beneficial effect on treating frequent premature ventricular contractions [11], frequent premature atrial contractions and atrial fibrillation[12,13].

In 2017, “Heart Rhythm Society of the Chinese Society of Biomedical Engineering” and the “NaoXin Tong Zhi Committee of the Chinese Society of Integrative Medicine” organized domestic and foreign experts to draft expert consensus statement on Wenxin granule in treating cardiac arrhythmia, which indicated that Wenxin granule could be applied in atrial premature beats, ventricular premature beats, and atrial fibrillation[14]. This expert consensus is very important to physicians in understanding and applying Wenxin granule in clinical practice.However,the consensus was based on single-center or multicenter RCTs but not systematic reviews/meta-analyses, meanwhile, the quality of RCTs has not been evaluated in the consensus.

There are a number of systematic reviews/meta-analyses existing for Wenxin granule in treating cardiac arrhythmia, which may provide evidence to physicians. However, the methodological quality and evidence quality of them are unclear, if they are valuable for clinical practice keep uncertain. So we conduct the overview of systematic reviews/meta-analyses for Wenxin granule in treating arrhythmia to identify methodological quality and evidence quality of them. In addition, we examine some methodological issues of systematic reviews/meta-analyses and primary RCTs, to identify research priorities for Wenxin granule in future.

Methods

Inclusion criteria

The inclusion criteria as follows: (1) Systematic reviews/meta-analyses of RCTs or quasi-RCTs for Wenxin granule in the treatment of any type of cardiac arrhythmia. (2) The systematic reviews/meta-analyses should be conducted by two or more authors. (3) The interventions should be Wenxin granule alone or Wenxin granule plus conventional therapy/western medicine. (4)The comparisons are placebo, no treatment, conventional therapy or western medicine. (5) The language should be in English or Chinese. (6) The systematic reviews/meta-analyses should be public published

Exclusion criteria

The exclusion criteria is follows:(1) The comparisons included other Chinese herbal medicine or non-pharmaceutical therapy (such as exercise or acupuncture). (2) Protocols. (3) Full texts cannot be retrieved. (4) Inconsistent/contradictory results and conclusions exist in the systematic review/meta-analyses(5) The investigators conducted meta-analyses inappropriately

Search strategy

We retrieved PubMed,Cochrane Library,Web of Science,Embase, China National Knowledge Infrastructure(CNKI), Wanfang Database, Chinese VIP Information(VIP), SinoMed from inception to October 31, 2018.The search strategy for English electronic databases is in Table 1. We used eligible terminology in Chinese databases retrieval.

Table 1 Full search strategies of English electronic databases

Study identification and data extraction

Two authors (Qiu RJ and Tang XY) independently identified inclusion and exclusion criteria for all of searched systematic reviews/meta-analyses via reading titles and abstracts. Potential ones would be retrieved full-texts for further identification. Any disagreement or controversy was solved by discussion or consulting the third author(Shang HC).

Two authors (Qiu RJ and Tang XY) extracted data from included systematic reviews/meta-analyses independently. The data was extracted for further assessment. The information of extracting included title,the first author’s name,number of authors,published time,search strategy,number of included trials and participants,interventions and comparisons, quality of primary trials,quality assessment tool, outcomes, results, conclusions.Any discrepancy was resolved by discussion or consulting the third author(Shang HC).

Methodological quality assessment of included systematic reviews/meta-analyses

Two authors (Qiu RJ and Yuan HC) assessed the methodological quality of all the included systematic reviews/meta-analyses by using the Assessment of Multiple Systematic Reviews(AMSTAR)instrument[15].It contains 11 items, for each item, the answer is ‘yes’when the article explicitly meets the item’s criterion, the answer is ‘no’when the article does not meet the item’s criterion explicitly,the answer is ‘can’t answer’when the item is related but not described completely or not reported at all, the answer is ‘not applicable’ when the item is not relevant. We judged as ‘yes’for item 3 if the reviewers of systematic reviews searched any two English databases(such as the Cochrane Library,PubMed,Embase or other databases) and any two Chinese databases(such as SinoMed,CNKI,Wanfang database or other database). If an item judged as ‘yes’, then the item got a score of 1, other kinds of answer would get a score of 0. The results were cross-checked, any inconformity would be resolved by discussion.

Evidence quality assessment of included systematic reviews/meta-analyses

If the methodological quality of systematic reviews/meta-analyses is poor,it is impossible for them to guide clinical practice, because of the possible exaggerated or inaccurate results. So if a systematic review/meta-analyze got score of 7 or more according to the assessment of methodological quality with AMSTAR instrument, then it would be evaluated evidence quality by using Grading of Recommendations Assessment,Development and Evaluation(GRADE)[16].Five factors such as limitation, inconsistency, indirectness,imprecision and publication bias would be evaluated and achieved consensus by two authors (Qiu RJ and Yuan HC). Then we conducted the assessment with GRADE profiler.

Results

Study identification

We searched 203 articles according to the search strategy.141 duplicated articles were dismissed. After screened titles and abstracts, 31 articles were excluded because they were uncorrelated to this overview. We retrieved 31 full-texts for further investigation, then 12 of them were excluded with reasons. In the end, 19 systematic reviews with 52,905 participants included in the overview. The flowchart of the overview of systematic reviews is in Figure 1.

Figure 1 The flowchart of the overview of systematic reviews

Characteristics of included systematic reviews/meta-analyses

The characteristics of the included 19 systematic reviews/meta-analyses are in Table 2. Among these 19 systematic reviews/meta-analyses, 4 (4/19, 21.05%) of them were published in English [17-20], the rest of them were published in Chinese. There was no Cochrane review included. The included systematic reviews/meta-analyses had a mean of 6 authors (range from 2 to 15). The publication time ranged from 2011 to 2018.

For diagnosis, the included systematic reviews/meta-analyses covered atrial fibrillation (6/19,31.58%), ventricular premature beat (5/19, 26.32%),atrial premature beat (1/19, 5.26%), a complex condition diagnosed as ventricular arrhythmia (1/19, 5.26%), a complex condition diagnosed as arrhythmias (6/19,31.58%).The systematic reviews contained a median of 16 clinical trials (ranged from 5 to 31). There were a variety of comparisons included in these systematic reviews, such as Wenxin granule compared with western medicine or conventional treatment,Wenxin granule plus western medicine compared with western medicine,Wenxin granule plus conventional treatment compared with conventional treatment, Wenxin granule compared with placebo.

For the quality of original studies, 12 of systematic reviews were assessed by Cochrane handbook, which showed that the majority of original studies were in moderate to high risk of bias. 7 of systematic reviews were assessed by Jadad scale or revised Jadadscale, the results showed that the majority of original studies were scored 1-4.

The mean number of outcomes in systematic reviews is 4 (ranged from 1 to 9). The majority of included systematic reviews assessed clinical efficacy(14/19, 73.68%) and adverse effects (15/19, 78.95%). No systematic reviews reported endpoints and economic evaluations. The outcome matrix showed that there are obvious heterogeneity for the outcome reporting in included systematic reviews which had the same disease(Table 3-5. For atrial fibrillation, there were 14 different outcomes extracted from systematic reviews,however,10(10/14,71.43%)of them were reported only once in there systematic reviews, which is as same as the ventricular premature beats.

Sixteen (16/19, 84.21%) of systematic reviews reported adverse effects/events, the results showed that Wenxin granule alone or combined with western medicine had less adverse effects/events than control group, or there was no significant difference between Wenxin granule group and control group. However,there is limited evidence showed that the causality between Wenxin granule and adverse effects/events.The details of adverse effects/events are in Table 6.

For the conclusions, the majority included systematic reviews showed that Wenxin granule alone or combined with western medicine can significantly improve the efficacy and lower adverse effects, though the quality of design and methodology are poor.

Table 2 Characteristics of included systematic reviews/meta-analyses

Study ID Conditions No. of authors No.of Trials/participants Quality of original trials AMSTAR score Comparisons Outcomes and included trials/participants Conclusions Li M 2018[36]AF 15 24/2246 Unclear(Cochrane handbook)6 WG+CT vs CT/placebo;WG+WM vs WM 1. Mean ventricular rate(10/870)2.Frequency of AF(2 trials)3.Duration of AF(1 trial)4.SR conversion time(5/407)5.SR conversion rate(6/499)6.Recurrence rate of AF(5/346)7. Heart functional outcome (7 trials)8.ECG(6 trials)9.Adverse effect(10/1012)Compared with WM, WG combined with WM could effectively control rapid ventricular rate, reduce the frequency and duration of AF,shorten SR conversion time, increase SR conversion rate, reduce AF recurrence rate,improve cardiac structural and electrophysiological dysfunction with less adverse reactions. However, due to the low quality of the included trials, the evidence is weak.Du 2014[37]VPB 2 20/1985 1(Jadad scale,0-5)5 WG+CT vs CT 1.ECG(8/714)2.Clinical efficacy(20/1985)3.Adverse effect(11/1117)WG can significantly improve the clinical efficacy and ECG, reduce the adverse reaction ,but which needs to be further confirmed by more large-sample, high-quality RCTs.He 2016[18]VPB 5 10/2251 Moderate or high risk of bias(Cochrane handbook)10 WG vs placebo/ CT, WG +WM vs WM 1.VPB’s numbers(10/2168)2.Symptoms(9/1976)3.Adverse events(8/1998)In spite of poor methodology quality of most included trials, the results show that WG, alone or combined with WM, might be beneficial to reduce VPBs’ numbers and relieve symptoms for non-fatal VPBs patients. WG may be associated with an increased risk of gastrointestinal adverse events and a reduced risk of proarrhythmic reactions.Guo 2016[38]VPB 6 20/1568 Range 1-3(Revised Jadad scale,0-7)7 WG+metoprolol+CT vsmetoprolol+CT 1.Smptoms(10/703)2.Efficacy of Bradyarrhythmia(19/1476)3.Adverse events(13/952)WG + metoprolol can effectively improve the symptoms of patients, and significantly diminish the ventricular premature, while there is no significant difference in the incidence of

Study ID Conditions No. of authors No.of Trials/participants Quality of original trials AMSTAR score Comparisons Outcomes and included trials/participants Conclusions total adverse drug reactions and bradycardia between the two groups. Also the difference in the incidence of atrioventricular block is with marginal significance. So the safety between the two groups may need further study after more samples be tested.Shen J 2016[39]VPB 9 10/938 Range 1-3(Revised Jadad scale,0-7)6 WG+metoprololvsmetoprolol 1.Clinical efficacy(4/503)2.Symptoms(7/658)3.Dynamic ECG(6/435)4.Adverse effect(8/696)WG + metoprololtakes more curative effect than metoprolol alone, also can relieve symptoms and improve ECG, and has no significant difference in adverse reactions. The quantity and quality of included trials are limited. RCTs with higher quality for further proving the conclusion.Li M 2017[19]VPB with HF 8 8/595 High risk of bias(Cochrane handbook)8 based on CT,WG alone or WG+WM vs blank control or WM 1.The frequency of VPBs within 24 h(4/332)2.LVEF(6/475)3.BNP(4 RCTs)4.VPBs(6/511)5. LVED ( 12 weeks: 2/147, 8 weeks:3/220)6.6-min walking test(2/147)7.Adverse effects(8/595)WG may be effective and safe for treating VPBs and HF. However, these problems (blind uncertainty, single country) were main limitations and deficiencies, resulting in the lack of high quality evidence.Lu J 2018[40]ventricular arrhythmia 4 31/3075 Range 2-3(Revised Jadad scale,0-7)5 WG+metoprolol+CT vsmetoprolol+CT 1.Clinical efficacy(31/3075)2.24 h dynamic ECG(11/1124)3.ST segment depression(6/671)4. Duration of ST segment depression(7/751)5.Number of VPBs(7/751)6.Short-term ventricular tachycardia(5/478)Compared with metoprolol alone, WG +metoprolol can significantly improve the total clinical efficacy, reduce the ST segment depression, shorten the ST segment depression duration, and reduce the number of VPBs and s Short-term ventricular tachycardia. In addition,they can reduce adverse reactions rate.However, the disease status, drug dose, and

Study ID Conditions No. of authors No.of Trials/participants Quality of original trials AMSTAR score Comparisons Outcomes and included trials/participants Conclusions 7.Adverse effect(16/1857)follow-up time of included studies were not completely unified, and the quality of the included studies was low, which affected the reliability of the results.Miao 2016[41]APB 4 16/1441 High or unclear(Cochrane handbook)6 WG + WM/CT vsWM/CT;WG vs WM/CT 1.Clinical efficacy(13 trials)2.Symptoms(2 trials)3.ECG(3 trials)WG alone or combined with WM can be more effective in improving the ECG and clinical symptoms in patients with atrial premature contractions. However, the low quality of the included studies caused a bad effect on the conclusion which still need high quality studies to demonstrate.Li 2011[23]Arrhythmias 6 24/25640 Moderate or high risk(Cochrane handbook)7 WG vsmexiletine/propafenone/amiodarone 1. Clinical efficacy (WG vsmexiletine:4/472,WG vspropafenone: 15/1677, WG vsamiodarone:5/415)2. ECG (WG vsmexiletine:2/205, WG vspropafenone:10/1281, WG vsamiodarone:2/171)3.Gastrointestinal adverse reaction (WG vspropafenone:10/915, WG vsamiodarone:4/383)4. Adverse effect rate (WG vspropafenone:7/673 WG vsamiodarone:4/383)The effect of WG in anti-arrhythmia is not worse than the WM, and the incidence of adverse reactions is lower. The study was limited by the quality of included studies.Shi 2012[42]Arrhythmias 2 10/987 Moderate or high risk(Cochrane handbook)3 WG vs WM 1.Clinical efficacy(10/987)The comprehensive analysis shows that WG is a safe and effective drug for arrhythmia.

AF: atrial fibrillation;APB: Atrial premature beat; CT: conventional therapy; NVAF: Non-valvular atrial fibrillation; SR: sinus rhythm;TCM: traditional Chinese medicineVPB: Ventricular premature beat;WG: Wenxin granule; WM: western medicine.

Table 3 The outcome matrix of included reviews for atrial fibrillation

Table 4 The outcome matrix of included reviews for ventricular premature beats

Table 5 The outcome matrix of included reviews for other types of arrhythmia

Table 6 Reported adverse effects/events in included reviews

Study ID Adverse effects/events Relationship between AEs and WG bradycardia(n=19),gastrointestinal reaction(n=12),fatigue(n=4),hypotension(n=4),atrioventricular block(n=8).The WG group had less rate of AEs([RR=0.70,95%CI:0.50-0.98,P＝0.04)than control group.Shen J 2016[39]8(8/10,80.00%)RCTs mentioned AEs.There was no statistical difference between WG group and control group in the rate of AEs(RR=0.77,95%CI:0.42-1.18,P=0.18).Not reported Li M 2017[19]All of the RCTs(8/8,100%)reported adverse drug reactions.In the treatment group,AEs included bradycardia(n=8),dizziness(n=6)and nausea(n=5).In the control group,AEs included,including bradycardia,low cardiac function,thyroid dysfunction,gastrointestinal reaction,dizziness,etc(n=16).Not reported Lu J 2018[40]16(16/31,51.61%)RCTs reported AEs.WG group had less AEs than control group.(OR=0.56,95%CI:0.40-0.77,P=0.0005).Not reported Li 2011[23]14(14/24,58.33%)RCTs reported gastrointestinal reaction.WG group can lower gastrointestinal reaction than propafenone(OR=0.24,95% CI: 0.12 -0.45, P <0.01). There was no statistical difference between WG andamiodarone (OR=0.58, 95% CI: 0.28-1.22, P>0.05).The total incidence of gastrointestinal reactions in WG group was significantly lower than control group (OR = 0.34, 95% CI:0.21- 0.54, .P < 0.01). 11 (11/24, 45.83%) RCTs reported AE of arrhythmia. WG group can lower arrhythmia than propafenone(OR=0.15,95%CI: 0.05-0.47,P<0.01)and amiodarone(OR=0.06,95%CI: 0.01-0.24,P<0.01).The total incidence of gastrointestinal reactions in WG group was significantly lower than control group(OR=0.10,95%CI:0.04-0.23,P<0.01).Yes He 2014[43]6(6/25,24.00%)RCTs reported gastrointestinal reaction.WG group can lower gastrointestinal reaction than amiodarone(OR=0.54,95%CI:0.29-0.98,P＝0.04).6(6/25,24.00%)RCTs reported AE of arrhythmia.WG group can lower arrhythmia than amiodarone(OR=0.05,95%CI:0.02-0.16,P<0.00001).Yes Shi 2017[22]All of RCTs mentioned AEs(7/7,100%)AEs,which included dizziness,nausea,loss of appetite,thirst,bradycardia,hypotension,gastrointestinal reaction.WG group had less AEs than control group(RR=0.41,95%CI:0.22-0.77,p<0.01)Not reported Zhang J 2017[44]16(16/23,69.57%)RCTs mentioned AEs.WG+amiodarone had less AEs than amiodarone(OR=0.68,95%CI:0.49.0.94,P=0.02).Not reported Liang Y 2017[21]4(4/5,80%)RCTs mentioned AEs.In treatment group,6(6/186,3.23%)patients had AEs,including gastrointestinal reaction(n=4),sinus bradycardia(n=1),abnormal liver function(n=1).In control group,22(22/186,11.83%)patients had AEs,including gastrointestinal reaction(n=15),sinus bradycardia(n=5),abnormal liver function(n=2).Not reported

Quality assessment of included systematic reviews

Methodological quality assessment of included systematic reviews

The details of methodological quality assessment for included systematic reviews are listed in Table 7. The results are summarized in Table 8. The AMSTAR score ranged from 3 to 10 with a mean score of 6.5, the median score was 6. The systematic reviews had a high quality of methodology published in English. The relationship of publication time with AMSTAR score is in Table 9.

There were 9 (9/19, 47.37%) systematic reviews had the score of 7 or more.Among all of the 19 systematic reviews, only 1 (1/19, 5.26%) registered a protocol.

84.21%) systematic reviews combined the findings of studies with appropriate methods. 17 (17/19, 89.47%) systematic reviews assessed publication bias. Only 5(5/19, 26.32%) of the 13 systematic reviews stated the conflict of interest or provided the information of fund assistance.

Quality of evidence for included systematic reviewsFor the 9 systematic reviews which had a score of 7 or more, 3 of them were excluded for assessing quality of evidence because the diagnosis are complex, which included different types of arrhythmia that was difficult to provide evidence for clinical practice [21-23]. The results of GRADE are summarized in Table 10.The results showed that the overall quality of evidence was limited (range“very low” to “low”), no “moderate” or “high” quality of evidence existed. The quality of evidence was downgraded because of the following limitations: (1)Almost all of included original trials had limitations in study design or conduction,so the risk of bias was serious or very serious; (2) systematic reviews existed serious or very serious inconsistency; (3) systematic reviews existed serious imprecision; (4) systematic reviews may exist publication bias.

Table 7 The methodological quality assessment of included systematic reviews with AMSTAR

Score 7 5 8 Item 11 No No No Item 10 Yes Yes Yes Yes Yes Yes Item 9 Item 8 Yes No Yes Item 7 Yes Yes Yes Item 6 Yes No Yes Item 5 No No No Can’t Item 4 answer No Yes Item 3 Yes Yes Yes Item 2 Yes Yes Yes Item 1 No No No Stud ID Shi 2017[22]Zhang J 2017[44]Liang Y 2017[21]

Table 8 Methodological quality assessment of included systematic reviews(n=19)

Table 9 The relationship of publication time with AMSTAR score

Table 10 The quality of evidence for included reviews

Study ID Conditions No.of No.of patients Effect Quality 10 WG+amiodarone group Amiodarone group RR 0.51(0.32 to 0.8)<53/1000(<21to<73)Very low 26/480(5.4%)51/475(10.7%)He M,2016[18]VPB WG or WG+WM Placebo/CT/WM VPB's numbers 10 940/1134(82.9%)646/1034(62.5%)RR 1.16(1 to 1.35)>100/1000(>0 to>219)Very low Symptoms 9 20/1044(1.9%)31/954(3.2%)RR 0.59(0.35 to 1.01)<13/1000(<21 to>0)Very low Adverse effects 8 20/1044(1.9%)31/954(3.2%)RR 0.59(0.35 to 1.01)<13/1000(<21 to>0)Very low Guo 2016[38]VPB WG+metoprolol metoprolol Smptoms 10 317/356(89%)246/347(70.9%)RR 1.26(1.17 to 1.36)>184/1000(>121 to>255)Low Efficacy of Bradyarrhythmia 19 660/741(89.1%)536/735(72.9%)RR 1.22(1.16 to 1.29)>160/1000(>117 to>211)Low Adverse events 13 49/478(10.3%)69/474(14.6%)RR 0.7(0.5 to 0.98)<44/1000(<3 to<73)Very low Li M 2017[19]VPB with HF The frequency of VPBs 2 WG+WM Blank control or WM-<337.92(366.94 to<308.91)Very low 106 102 2 WG Blank control-<1001.14(1257.65 to<744.63) Very low 62 62 4 WG or WG+βblockers β blockers or blank control-<427.08(526.73 to<327.43)Very low 168 164

Study ID Conditions No.of No.of patients Effect Quality WG or WG+β blocker β blockers or blank control LVEF 6 239 236->4.12(2.97 to>5.27)Very low The effect of VPBs 6 37/257(14.4%)75/254(29.5%)RR 0.48(0.34 to 0.69)<154/1000(<92 to<195)Very low The variation of LVED 5 183 184-<0.61(<1.41 to>0.18)Very low

Discussion

This is the first overview of systematic reviews/meta-analyses on Wenxin granule in treating arrhythmias.We summarized the main finding of included 19 systematic reviews and assessed the methodological and evidence quality of them.The results showed that the majority of included systematic reviews are in low methodological quality.Though some of the methodological quality can be accepted, the evidence quality of them is from “very low” to “low”, which is difficult to guide clinical practice.

However,we could learn much from these low quality of systematic reviews,which may help researchers identify research priorities for further research, so that the

quality of studies will be improved and investment waste will be decreased.

Improve the quality of original trials

The overview of systematic reviews showed that the majority of original trials were at high or unclear risk of bias, or had low Jadad score. No reporting of randomization method, allocation concealment and blinding method existed in many original trials, which resulted in the low methodological quality of original trials.For Chinese herbal medicine, it is difficult to produce placebo which has the similar odour and color. Maybe it is one of the problems for investigators to conduct blinding method. In addition, the publication bias may exist, which signified that some negative results did not report.

In the future, the usage of CONSORT (consolidated standards of reporting trials)in clinical trials of TCM[24]will improve the quality of reporting.Providing study protocol registration before conducting RCTs may reduce bias of reports. If the quality of original trials is improved,it is much better for researchers to produce the best evidence in systematic reviews.

Identify the indications of Wenxin granuleWenxin granule has been approved for treating arrhythmias for more than 20 years,in this overview of systematic reviews, there were 3 conditions included, such as atrial fibrillation,ventricular premature beat,and atrial premature beat.However,it still keeps uncertain that on which kind of cardiac arrhythmia that Wenxin granule takes the best beneficial effect.

In majority of included systematic reviews, Wenxin granule took the role of adjuvant therapy,which combined with western medicine or conventional treatment.The placebo controlled trials were lack in included systematic reviews, the comparison of Wenxin granule and western medicine were few. These problems made it difficult for doctors and patients to decide that if Wenxin granule can substitute western medicine in treating arrhythmias in clinical practice. So it needs much more placebo controlled trials to ascertain the efficacy and safety of Wenxin granule.

Improve the quality of systematic reviews

This overview showed that the majority of systematic reviews had methodological flaws, such as lack of protocol, grey literature search, or sufficient description of included studies’ characteristics. Some systematic reviews existed inappropriate data synthesis,inappropriate conclusions, or even potential publication bias. These flaws resulted in the likelihood of selection bias, reporting bias, or mistaken results, which made it impossible to provide high quality of evidence to clinical practice.

In the future, the AMSTAR [15] and PRISMA standards[25]should be recommended to researchers of systematic reviews to improve the reporting quality and methodological quality.

Improve reporting of harms in RCTs and systematic reviews

Though the majority of systematic reviews reported the adverse effects/events, we can see that not all of original RCTs mentioned them (Table 5). The information of adverse effects/events were limited, no causality assessment between Wenxin granule and the harms, so it is difficult to identify the safety of Wenxin granule.

In the future,the researchers should report harms in RCTs[26] and systematic reviews of TCM [27], which will help researchers better understand the safety of TCM.

Develop core outcome sets of arrhythmia

This overview show that the reporting outcomes were various in different systematic reviews of specific conditions. Researchers preferred to report complex outcome, such as clinical efficacy. ECG efficacy,symptoms in some systematic reviews.For all of included systematic review, there were no endpoints reporting,which is very important to patients and reflected the long-term effect of drugs.

Furthermore, there were diverse outcomes reporting in original trials, for example, there were no relevant outcomes, the treatment time did not meet the inclusion criteria, no treatment time reporting, or the data was incomplete or cannot be extracted, so that some studies had to be excluded from systematic review.The excluded studies which reported inappropriate outcome measures showed in Table 11.

Table 11 Excluded studies because of inappropriate outcome measures

Though the RCTs were included in systematic reviews, not all of them can be merged in meta-analysis because not all of included studies reported predefined outcomes(Table 2).

To solve these problems, developing core outcome set (COS), agreed outcomes that should be measured and reported,as a minimum set in all clinical trials in specific areas of health or health care, may be a good approach[28]. If researchers of clinical trials and systematic reviews report outcomes in according to COS, the original trials may could be synthesized and contributed results in systematic reviews,which may provide the best evidence to clinical practice. On the other hand, the missing outcome reporting will be identified easily,which signifies the potential bias of reporting.

From now on, there is an ongoing COS for clinical trials of TCM in non-valvular atrial fibrillation [29]. In the future, we hope that researchers develop much more COSs for clinical trials in treating other kind of arrhythmia,to improve the quality and value of trials.

Reducing research waste

Many RCTs cannot be included in systematic reviews or cannot be conducted meta-analysis in systematic reviews showed that there were research waste.On the other hand,only a small number of RCTs were included in similar systematic reviews.For example,there were 68 RCTs for ventricular premature beats from all of included systematic reviews, only 7 of them were included in 2 different reviews, while for the 89 RCTs of atrial fibrillation, 9 of them were included in 2 different systematic reviews,4 of them were included in 3 different systematic reviews.

To reduce research waste and increase value of research, researchers could do many things, such as improving research design, conduct and analysis,conducting a study within a trial [30,31], as well as improving research regulation and management, such as working with regulators, policy makers, patients and public to increase participants recruitment and retention,data monitoring and sharing[32].

Conclusions

At the present,the majority of systematic reviews showed that Wenxin granule may have potential beneficial effect on improving clinical efficacy and lowering adverse effects/events for patients with arrhythmia. However,with the limitation of methodological and evidence quality of included systematic reviews, further investigations are needed. In the future, high-quality evidence from well designed and performed clinical trials and systematic reviews which focused on the endpoints are necessary. In addition, developing COS for arrhythmia are necessary to reduce heterogeneity of outcome reporting, as well as many others things should be done to reduce research waste.

1. Y. Miyasaka, M. E. Barnes, B. J. Gersh, et al.Secular trends in incidence of atrial fibrillation in Olmsted County, Minnesota, 1980 to 2000, and implications on the projections for future prevalence.Circ 2006,114:119-125

2. L. Friberg, M. Rosenqvist, G. Y. Lip. Evaluation of risk stratification schemes for ischaemic stroke and bleeding in 182 678 patients with atrial fibrillation:the Swedish Atrial Fibrillation cohort study.EurHeart J 2012,33:1500-1510.

3. S. G. Priori, C. Blomstrom-Lundqvist, A. Mazzanti,et al. 2015 ESC Guidelines for the Management of Patients With Ventricular Arrhythmias and the Prevention of Sudden Cardiac Death.Rev Esp Cardiol(Engl Ed)2016,69:176

4. L. Macle, S. Nattel. Arrhythmias in 2015: Advances in drug, ablation, and device therapy for cardiac arrhythmias.Nat Rev Cardiol 2016,13:179

5. Liu S, Chen J. Marketing strategy of wenxingranuel of Buchang Pharmaceuticals. Bus Econ 2014, 14:2014.

6. A. Burashnikov, A. Petroski, D. Hu, et al.Atrial-selective inhibition of sodium-channel current by WenxinKeli is effective in suppressing atrial fibrillation.Heart Rhythm 2012,9:125-131.

7. D. Hu, H. Barajas-Martinez, A. Burashnikov, et al.Antzelevitch, "Mechanisms underlying atrial-selective block of sodium channels by WenxinKeli: Experimental and theoretical analysis.Int J Cardiol2016,207:326-334.

8. Y. Minoura, B. K. Panama, V. V. Nesterenko, et al.Effect of WenxinKeli and quinidine to suppress arrhythmogenesis in an experimental model of Brugada syndrome. Heart Rhythm 2013, 10:1054-1062.

9. Tang QZ, Shi XT, Wang T, et al. Effects of Nardostachyschinensisbatal extract on sodium and calcium channels in rabbit ventricular myocytes.Chin J Cardiol 2004,32:267-270.

10. Xing Y,Gao Y,Chen J,et al,Wenxin-Keli Regulates the Calcium/Calmodulin-Dependent Protein Kinase II Signal Transduction Pathway and Inhibits Cardiac Arrhythmia in Rats with Myocardial Infarction.Evid Based Complement Alternat Med 20113, 2013:464508.

11. Hua W, Gao RL, Zhao BC, et al. The Efficacy and Safety of WenxinKeli in Patients with Frequent Premature Ventricular Contractions: A Randomized,Double-blind, Placebo-controlled, Parallel-group,Multicenter Trial.Chin Med J (Engl) 2015, 128:2557-2564.

12. Wang HP, Zhu HJ, Yan CF. The clinical efficady of WenxinKeli combined with low dose of propafenone in treating frequent atrial premature beats.ModJ IntegrTradi Chin Western Med 2012, 21:3489-3490,.

13. Meng Z, Tan J, He Q, et al. WenxinKeli versus Sotalol for Paroxysmal Atrial Fibrillation Caused by Hyperthyroidism: A Prospective, Open Label, and Randomized Study. Evid Based Complement Alternat Med 2015,2015:101904.

14. Heart Rhythm Society of the Chinese Society of Biomedical Engineering; Nao Xin Tong Zhi Committee of the Chinese Association of Integrative Medicine. Expert Consensus on Wenxin Granule for Treatment of Cardiac Arrhythmias.Chin Med J(Engl)2017,130:203-210.

15. B. J. Shea, J. M. Grimshaw, G. A. Wells, et al。Development of AMSTAR: a measurement tool to assess the methodological quality of systematic reviews.BMC Med Res Methodol 2007,7:10.

16. D.Atkins,D.Best,P.A.Briss,et al.Grading quality of evidence and strength of recommendations.BMJ 2004,328:1490.

17. Chen Y, Nie S, Gao H, et al. The effects of wenxinkeli on p-wave dispersion and maintenance of sinus rhythm in patients with paroxysmal atrial fibrillation:a meta-analysis of randomized controlled trials.Evid Based Complement Alternat Med 2013,2013:245958.

18. He M, Lv Z, Yang ZW, et al. Efficacy and safety of Chinese herbal medicine WenxinKeli for ventricular premature be ats: A systematic review. Complement Ther Med 2016,29:181-189.

19. Li M, QiuR, TianG, et al. WenxinKeli for Ventricular premature complexes with Heart failure:A Systematic Review and Meta-Analysis of Randomized Clinical Trials. Complement Ther Med 2017,33:85-93.

20. Wang C, Wu W, Wu H, et al. The clinical efficacy and adverse reaction of wenxin granule and amiodarone for patient with atrial fibrillation: A meta-analysis. Int J Clin Exp Med 2017, 10:14256-14265.

21. Liang Y, Yao X. Wenxin granule combined with amiodarone in the treatment of heart failure with arrhythmia: a meta-analysis. Chin J Integr Med Cardio Cerebrovasc Dis 2017,16:2006-2009.

22. ShiS, GengY,DuB, et al. Treatment of CHD-induced arrhythmias by WenxinKeli combined with bisoprolol: a systematic review and meta-analysis.Beijing J Tradi Chin Med 2017, 1:6-11.

23. LiJ, TangH, LiJ, ZhouZ, et al.A meta-analysis on effect of WenxinKeli in treatment of arrhythmia.Chin J EvidBasedCardiovasc Med 2011,2:84-89,100.

24. Bian Z, Liu B, Moher D, et al. Consolidated standards of reporting trials (CONSORT) for traditional Chinese medicine: current situation and future development.Front Med 2011,5:171-177.

25. D. Moher, A. Liberati, J. Tetzlaff, D. G. Altman.Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. BMJ 2009,339:b2535.

26. J. P. Ioannidis, S. J. Evans, P. C. Gotzsche, et al.Better reporting of harms in randomized trials: an extension of the CONSORT statement. Ann Intern Med 2004,141:781-788.

27. L. Zorzela, Y. K. Loke, J. P. Ioannidis, et al.PRISMA harms checklist:improving harms reporting in systematic reviews.BMJ 2016,352:i157.

28. M. Clarke and P. R. Williamson. Core outcome sets and systematic reviews.Syst Rev 2016,5:11.

29. Qiu R, Li M, Zhang X, et al. Development of a core outcome set (COS) and selecting outcome measurement instruments (OMIs) for non-valvular atrial fibrillation in traditional Chinese medicine clinical trials:study protocol.Trials 2018,19:541.

30. J. P. Ioannidis, S. Greenland, M. A. Hlatky, et al.Increasing value and reducing waste in research design, conduct, and analysis. Lancet 2014, 383:166-175.

31. S. Treweek, S. Bevan, P. Bower, et al.Trial Forge Guidance 1: what is a Study Within A Trial(SWAT)?Trials 2018,19:139.

32. S. R. Al-Shahi, E. Beller, J. Kagan, et al.Increasing value and reducing waste in biomedical research regulation and management. Lancet 2014, 383:176-185.

33. Du H,Dai X.Efficacy and safety of Wenxin Granule in treatment of atrial fibrillation: a Meta-analysis. J Anhui Univ Chin Med 2014,6:27-30.

34. Zhao M, Wang H, Xiong F, et al. Effect of Wenxin Granule combined metoprolol on atrial fibrillation: a meta-analysis.China Med 2014,4:458-463.

35. Wang H, Yang B, Zhou S. Efficacy and safety of Wenxin Granules combining amiodarone on atrial fibrillation: a Meta-analysis. Chin J Evid Based Cardiovascu Med 2015,2:161-164.

36. Li M, Qiu R, Sun Y, et al. The Clinical Study ofWenxinKeli in the Treatment of Atrial Fibrillation:a Systematic Review. World Sci Technol/Mod Tradi Chin Med Mater Med 2018,20:1761-1771.

37. Du H,Dai X.Efficacy and safety of Wenxin Granule in the treatment of premature ventricular beats: a Meta-analysis.J ClinElectrocardiol 2014,5:366-369.

38. Guo R, Jv N, Lin X,et al. Meta-analysis on efficacy and safety of Wenxin Granule combined with metoprolol for treatment of ventricular premature.Chin Tradi Herb Drugs 2016,6:1053-1060.

39. Shen J, Yuan P, Tang J, et al. Efficacy and safety of Wenxin Granules combining betaloc on ventricular premature beat：a Meta-analysis. Chin J EvidBased Cardiovascu Med 2016,7:786-791.

40. Lu J, Liu S, Li J.Wenxin Granules Combined with Metoprolol in the Treatment of Coronary Heart Disease with Ventricular Arrhythmias: A Meta-analysis.J Tradi Chin Med Univ Hunan 2018,38:670-676.

41. Miao P, Zhang M, Zhang L, et al. Effects of Wenxin granule on electrocardiogram and clinical symptoms in adults with atrial premature contractions: a Meta-analysis of randomized controlled trials. J Tianjin UnivTradi Chin Med 2016,3:160-166.

42. Shi H, Peng W. Meta-analysis of efficacy and safety of Wenxin Granules in the treatment of arrhythmia.China Med Herb 2012,5:65-66,68.

43. He Y, Liu Y, Zou A. Meta-analysis on Wenxin Granules in treatment of arrhythmia.Chin Tradi Herb Drugs 2014,15:2277-2282.

44. ZhangJ, ChenY, LuY, et al. Meta-analysis of therapeutic effect of Wenxin granules combined with amiodarone for arrhythmia in coronary heart disease.Tradi Chin .Drug Res Clin Pharmacol 2017, 2:.244-248