Shengmai (a traditional Chinese herbal medicine) for heart failure
Abstract
Background
Heart failure is a major public health problem worldwide. Shengmai, a traditional Chinese herbal medicine, has long been used as a complementary treatment for heart failure in China. This is an update of a Cochrane Review published in 2012.
Objectives
To determine the effect (both benefits and harms) of Shengmai in treatment of people with heart failure.
Search methods
We searched CENTRAL on The Cochrane Library (Issue 5 of 12, April 2013); DARE on The Cochrane Library (Issue 2 of 4, April 2013); MEDLINE (1948 to June Week 1 2013); EMBASE (1980 to 2013 Week 23); AMED (1985 to August 2008); BIOSIS (1969 to 7 June 2013); CBM (1978 to June 2013); VIP (1989 to June 2013); and CNKI (1979 to June 2013). We also handsearched Chinese journals and did not apply any language restrictions.
Selection criteria
We included randomised controlled trials (RCTs) of Shengmai plus usual treatment for heart failure versus usual treatment alone, or Shengmai versus placebo, irrespective of blinding status. In this update we only included studies with a clear description of randomisation methods and classified as true RCTs.
Data collection and analysis
Two authors independently selected trials, assessed methodological quality and extracted data. We calculated dichotomous data as risk ratios (RRs) and continuous data as mean differences (MDs) or standardized mean differences (SMDs) with corresponding 95% confidence intervals (CIs). We used a fixed‐effect model to perform meta‐analysis for outcomes without heterogeneity; and a random‐effects model to perform meta‐analysis for outcomes with heterogeneity.
Main results
We included a total of 14 RCTs (858 patients) in this review update, four of which were new trials. Of these 14 RCTs, 11 trials compared Shengmai plus usual treatment with usual treatment alone, and three trials compared Shengmai with placebo. Improvement of NYHA functional classification was more common in patients taking Shengmai plus usual treatment than in those receiving usual treatment alone (RR 0.37; 95% CI 0.26 to 0.51; 10 trials, 672 participants; low quality evidence). Beneficial effects of Shengmai in treating heart failure were also observed in other outcomes, including exercise test, ejection fraction and cardiac output. The three RCTs (106 patients) comparing Shengmai with placebo reported improvement in NYHA functional classification and in stroke volume. Three of the 14 RCTs reported a total of six patients with mild adverse effects and two were withdrawn due to the adverse effects. The adverse events rate was 1.21%.
Authors' conclusions
Shengmai may exert a positive effect on heart failure, especially for improving NYHA functional classification when Shengmai plus usual treatment is used. The review results should be interpreted with caution due to the high risk of bias of the included studies (particularly regarding allocation concealment and blinding), the small sample size of these studies, and the significant heterogeneity in outcomes such as ejection function, cardiac output and stroke volume. There was no evidence available concerning the effect of Shengmai on mortality, and more high quality studies with long‐term follow‐up are warranted.
Plain language summary
Shengmai (a traditional Chinese herbal medicine) for heart failure
Heart failure is a major public health issue worldwide. Despite advances in treatment, many people are affected and hundreds of thousand of people die each year. Shengmai is a traditional Chinese herbal medicine and is a mixture extracted from three herbs: Panax ginseng, Ophiopogon japonicus and Schisandra chinensis. In China it has been used to treat heart failure and ischemiac heart disease.
In this Cochrane Review update, we included 14 trials which compared Shengmai plus usual treatment with usual treatment alone, or compared Shengmai with placebo. Shengmai may be beneficial in improving heart function (low quality evidence) and in most trials, people who took Shengmai had few unwanted side effects. However, many of the included trials were of poor quality and included a small number of people. More high quality, large trials of Shengmai of treatment heart failure are needed.
Authors' conclusions
Summary of findings
| Shengmai plus usual treatment compared to usual treatment alone for heart failure | ||||||
| Patient or population: patients with heart failure | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | No of participants | Quality of the evidence | Comments | |
| Assumed risk | Corresponding risk | |||||
| Usual treatment alone | Shengmai plus usual treatment | |||||
| Mortality | See comment | See comment | Not estimable | 858 (11 trials) | See comment | 0 participants died |
| Lack of improvement in heart failure (NYHA class improved < I class or worsening of heart failure) | Moderate | RR 0.37 | 672 | ⊕⊕⊝⊝ | This outcome was transformed to be dichotomous and NYHA class improved < I class or worsening of heart failure was the event | |
| 279 per 1000 | 103 per 1000 | |||||
| Adverse effects | See comment | See comment | Not estimable | 4 (2 trials) | See comment | Two mild asomnia (1 trial) and two dry mouth and fidgety (1 trial) |
| Mean exercise tolerance test (after treatment) (min) | 6 mins | 6 higher | 60 | ⊕⊕⊝⊝ | The longer the exercise time, the better the heart function. | |
| Mean change in quality of life | 20 points | 7 higher | 60 | ⊕⊕⊝⊝ | ||
| Mean change in ejection fraction (%) | 3 to 13.1 | 9.02 higher | 354 | ⊕⊝⊝⊝ | ||
| Mean change in cardiac output (L/min) | 0.5 to 0.83 | 0.94 higher | 360 | ⊕⊝⊝⊝ | ||
| Mean change in stroke volume (mL) | 8.02 to 16.9 | 8.43 higher | 240 | ⊕⊝⊝⊝ | ||
| Change in cardiac index (L/min*m²) | 0.4 to 0.44 | 0.79 higher | 160 | ⊕⊝⊝⊝ | ||
| Hospitalization/rehospitalization | See comment | See comment | Not estimable | 858 (12 trials) | See comment | 0 participants hospitalized or re‐hospitalized |
| *The basis for the assumed risk (e.g. the median control group risk across trials) is provided in the 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). | ||||||
| GRADE Working Group grades of evidence | ||||||
| 1 Downgraded by 1 level for lacking of blinding and allocation concealment | ||||||
| Shengmai compared to placebo for heart failure | ||||||
| Patient or population: patients with heart failure | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | Number of participants | Quality of the evidence | Comments | |
| Assumed risk | Corresponding risk | |||||
| Placebo | Shengmai | |||||
| Mortality | See comment | See comment | Not estimable | 106 (3 trials) | See comment | 0 participants died |
| Lack of improvement in heart failure (NYHA class improved < I class or worsening of heart failure) | Moderate | RR 0.22 | 40 | ⊕⊕⊝⊝ | This outcome was transformed to be dichotomous and NYHA class improved < I class or worsening of heart failure was the event | |
| 1000 per 1000 | 220 per 1000 | |||||
| Adverse effects | See comment | See comment | Not estimable | 2 (1 trial) | See comment | One stomach discomfort and one hypoglycaemia were reported in one trial. |
| Mean change in exercise test (mins) | 0.11 | 3.56 lower | 18 | ⊕⊕⊝⊝ | The longer the exercise time, the better the heart function. | |
| Mean change in ejection fraction (%) | 0.2 | 12.45 higher | 52 | ⊕⊕⊝⊝ | ||
| Mean change in cardiac output (L/min) | 0.02 | 0.53 lower | 52 | ⊕⊕⊝⊝ | ||
| Mean change in cardiac index (L/min*m²) | 0.03 | 0.32 higher | 80 | ⊕⊕⊝⊝ | ||
| Mean change in ECG: Q‐Z | 1.48 | 10.72 higher | 80 | ⊕⊕⊝⊝ | ||
| Hospitalization/ rehospitalization | See comment | See comment | Not estimable | 106 (3 trials) | See comment | 0 participants hospitalized or re‐hospitalized |
| *The basis for the assumed risk (e.g. the median control group risk across trials) is provided in the 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). | ||||||
| GRADE Working Group grades of evidence | ||||||
| 1 Downgraded by 1 level for lacking of blinding and concealment | ||||||
Background
Description of the condition
Heart failure is a major public health issue, affecting over 5.1 million people in the USA and over 23 million people worldwide (Go 2013, Bui 2012). In the USA, over 650,000 new heart failure cases are diagnosed annually (Yancy 2013). Prevalence of heart failure increases substantially with age. Approximately 1% to 2% of adults in developed countries have heart failure; and this increases to over 10% in people aged 70 years or older (McMurray 2012). Despite developments in heart failure therapy, absolute mortality rates have remained at approximately 50% in the last five years (Yancy 2013), directly leading to 300,000 deaths annually (Lloyd‐Jones 2010). It places a considerable burden on the healthcare system due to high rates of hospitalizations, readmissions, and outpatient visits, with costs of over $39 billion annually in the USA alone (Bui 2012).
In 2000, chronic heart failure prevalence was 0.9% for the general population in China (Gu 2003). Around four million patients, aged between 35 to 74 years, have heart failure in China. Prevalence of chronic heart failure among people aged 35 to 44 years was 0.4%, and in people aged over 55 years was 1.3%. With a large aging population, the number of people with heart failure in China is expected to increase (Jiang 2009). Of people hospitalised with cardiac diseases, 17.9% in 1980, 16.3% in 1990 and 16.9% in 2000 had heart failure. The proportion of cardiac disease mortality attributable to heart failure remains unchanged (39.9% in 1980, 37.7% in 1990 and 41.1% in 2000) (CMA 2002). There is little or no updated information available about heart failure in China for the last 10 years (Jiang 2009).
Conventional treatment for heart failure
As the failing heart attempts to maintain adequate function, the heart may adopt several compensatory mechanisms. These include: maintaining cardiac output through increased left ventricular end diastolic volume; increasing wall thickness through ventricular remodeling; and maintaining tissue perfusion with augmented mean arterial pressure through activation of neurohormonal systems. All of these compensatory mechanisms eventually lead to a vicious cycle of worsening heart failure. Treatment strategies have been developed based upon the understanding of these compensatory mechanisms (Kemp 2012). According to the latest guidelines for heart failure, drugs for routine use include diuretics, angiotensin converting enzyme inhibitors (ACEIs), angiotensin receptor blockers (ARBs), beta blockers and aldosterone antagonists (McMurray 2012; Yancy 2013). In addition, digoxin may be given to selected people to reduce morbidity or mortality (Gheorghiade 1991; Hood 2004; Yancy 2013).
Diuretics inhibit reabsorption of sodium or chloride at specific sites in the renal tubules and increase urine volume (Faris 2012). The main aim of using diuretics is to relieve symptoms of congestion or decrease physical signs of fluid retention in patients with heart failure (Yancy 2013). ACEIs and ARBs both act on the renin‐angiotensin‐aldosterone system. ACEIs are now the best studied class of agents in heart failure and ARBs are considered a reasonable alternative for people intolerant to ACEIs (Yancy 2013). Beta blockers inhibit the adverse effects of the sympathetic nervous system in people with heart failure. Long‐term treatment with beta blockers can lessen the symptoms of heart failure and enhance the patient's overall sense of well‐being (Yancy 2013). Aldosterone receptor antagonists compete with aldosterone to bind at the mineralocorticoid receptor and improve left ventricle structural remodeling and performance by increasing left ventricle ejection fraction as well as decreasing left ventricle end‐diastolic and end‐systolic volumes (Maron 2010; Jean 2011). Digoxin binds to the sodium‐potassium adenosine triphosphatase (Na+‐K+ ATPase) pump on the myocardial cell membrane and inhibit its function, resulting in an increase in calcium concentration inside the heart cell and improvement of ventricular performance through a sustained moderate positive inotropic effect (Bers 2010).
How the intervention might work
Chinese herbs and medicine are widely used in Chinese hospitals (Tian 2008; Song 2011), and closely relates to the cultural belief that traditional Chinese medicine is more natural, effective, has fewer adverse effects, and is preferred by patients to Western drugs (Tan 2006; Song 2011).
According to traditional Chinese medicine theory, heart failure results from heart 'Qi' deficiency, which develops into a deficiency of both Yin and Qi (Li 2005). Therefore Chinese medicine treatment aims to nourish Yin and invigorate Qi. Both nourishing Yin and invigorating Qi are terminologies of traditional Chinese medical science. According to traditional Chinese medicine theory, a medicine nourishing Yin mainly works on material aspects of the body, while invigorating Qi mainly works on functional aspects of the body. Shengmai is usually used in China as a complementary treatment to Western treatments recommended for acute heart failure, and also as a single treatment for chronic heart failure. It is a mixture extracted from three herbs: Panax ginseng (invigorates Qi and dispels stagnation), Ophiopogon japonicus (nourishes Yin and grows succus) and Schisandra chinensis (invigorates Qi) (Ma 2003). It is usually mass‐produced as a patented drug, based on a standardized formula in different forms (including capsule, powder, oral liquid and injection). Treatment duration varies as there is neither a restriction on the duration of drug administration nor a therapeutic window according to the manufacturer's instructions. In clinical practice, Shengmai is used to facilitate improvement of symptoms and is often discontinued after the relief of symptoms or disease remission. Adverse effects are rare and mild, and include rash, anaphylactic shock, a sense of gastric distention, lumbar and back pain, hypotension and tachycardia (Li 2006b). The adverse event rates in most prospective and retrospective studies are under 1% and allergic reaction is most common（Li 2009, Li 2013, Deng 2012).
Shengmai shows a special pharmacological function to heart failure, which may be divided into the following five aspects:
-
Positive inotropic effect: Animal experiments show that Shengmai has an intricate adjustment function to the myocardial intracellular calcium concentration and the effect is related to the dosage: low‐dose enhances intracellular calcium, while high‐dose inhibits it (Wang 2002b). At appropriate doses, Shengmai has a positive inotropic effect and enhances heart pump function, similar to digitalis (Mao 2003).
-
Dual‐directional regulation on blood pressure: Shengmai was shown to exert a dual‐directional regulation on blood pressure: the effect varied with patients' conditions. Animal experiments show that Shengmai reduces blood pressures of normal anaesthetized dogs (Chen 2001) but made arterial pressure of rabbits with hemorrhagic shock during recovery period rise (Xia 1999). Another human trial showed that Shengmai reduced initial high blood pressure in some patients and increased other patients' initial low blood pressure (Dong 1984; Li 2003).
-
Improving haemodynamic parameters: Shengmai could dilate coronary vessels, increase coronary flow and reduce the peripheral resistance to different degrees. Shengmai regulated blood pressure without increasing heart rate. The cardiac output was stable or increased and the left ventricular filling pressure was constant or reduced. Shengmai increased cardiac stroke volume with absence of myocardial oxygen consumption increase (Chen 2002).
-
Delaying the cardiac remodelling: Shengmai delayed cardiac remodelling in a number of animal experiments (Deng 2008; Xu 2012). The mechanism may entail indirect reduction of heart wall tension and the renin activity (Zhang 2007b).
-
Antioxidative effect: The antioxidative effect of Shengmai on myocardial injury and its ability to improve cardiac microcirculation by eradicating oxygen free radicals has been observed (Yu 2000; Wang 2002a). Fructus schisandrae has been reported to derive antioxidant activity in Shengmai (Li 1996). However, the components of Shengmai are complex and not every chemical constituent has been isolated and determined yet, so it is unclear what the main active ingredients and their targets are.
Why it is important to do this review
The use of Shengmai plus usual treatment may benefit heart failure patients more than usual treatment alone. This was observed in the previous versions of this review. However, the poor quality of these included trials compromised the reliability of the evidence. Since then, a large number of new studies on Shengmai have emerged, and a review update is required to obtain better evidence for the clinical use of Shengmai.
Objectives
To determine the effects (both benefits and harms) of Shengmai in treatment of people with heart failure.
Methods
Criteria for considering studies for this review
Types of studies
RCTs and randomised cross‐over trials of Shengmai plus usual treatment versus usual treatment alone, or Shengmai versus placebo for heart failure, irrespective of blinding status. We only included trials which (a) were classified as true RCTs or true randomised cross‐over trials, and (b) clearly stated wash‐out period if it was a cross‐over trial.
Types of participants
People of any gender, any age or any ethnic group with clinically defined heart failure regardless of its primary causes, such as rheumatic, valvular and congenital heart disease, as well as other relevant characteristics including diastolic and systolic heart failure, acute and chronic heart failure.
Types of interventions
We defined Shengmai as the products extracted from ginseng (Radix codonopsis pilosulae or Panax ginseng), Radix ophiopogonis (Liriope sticata or Ophiopogon japonicus), and Schisandra chinensis (or Fructus schisandrae). We deemed any form of administration of Shengmai acceptable for study inclusion, although it was administered mainly via intravenous infusion. We included trials regardless of treatment duration.
Types of outcome measures
Primary outcomes
-
Mortality
-
New York Heart Association (NYHA) function classification
-
Adverse effects
Secondary outcomes
-
Exercise test or six‐minute walk test performance
-
Quality of life as measured by various instruments
-
Change in haemodynamics
-
Change in myocardial contractility
-
Hospitalization and rehospitalization
-
Costs
Search methods for identification of studies
We searched the following electronic databases up to 13 June 2013: CENTRAL on The Cochrane Library (Issue 5 of 12, April 2013); DARE on The Cochrane Library (Issue 2 of 4, April 2013); MEDLINE on Ovid (1948 to June Week 1 2013); EMBASE on Ovid (1980 to 2013 Week 23); AMED (1985 to August 2008) (Note: we did not search AMED for this update as it is no longer available to the person conducting the searches); BIOSIS Citation Index (1969 to 7 June 2013); CBM (1978 to June 2013); VIP (1969 to June 2013); and CNKI (1979 to June 2013). We have included the search strategies from the original review (Appendix 1), the second updated strategies (Appendix 2), the third updated strategies (Appendix 3) and the search strategy for this update (Appendix 4). We did not apply any language restrictions.
Data collection and analysis
Selection of studies
Two authors (ZQ, ZJ) screened the studies identified from the search strategy. They included studies according to the prespecified selection criteria and using a selection form. We resolved any disagreements by discussion. Where necessary, we consulted a third author (CJ) to resolve disagreements.
Data extraction and management
Two authors (ZQ, ZJ) extracted data independently using a self‐developed data extraction form. We resolved any disagreements through discussion, or if necessary, in consultation with CJ. We sought data that were unavailable in the trial reports by contacting the principal investigators of the studies. We noted the diagnosis criteria in various studies and we recorded presence of clinical criteria including dyspnoea, orthopnoea, rales, S3 gallop, or peripheral edema, and cardiomegaly on chest X‐ray. When available, we recorded objective measurements of ejection fraction from echocardiography, radio nuclide ventriculography, or ventriculography performed at cardiac catheterization. Presence of an ejection fraction of ≥ 0.45 was the basis for identifying a subset of patients who had heart failure with preserved ejection fraction. We also noted the duration of heart failure. The length of the wash‐out period should have been at least five times longer than the half‐life period of the main ingredients of Shengmai (9.5 days) or proved there was no carry‐over effect (Li 2011).
Assessment of risk of bias in included studies
We based the assessment of risk of bias on the Cochrane Collaboration's tool for assessing risk of bias (Higgins 2011).Two authors (ZQ, ZJ) independently appraised the risk of bias in included trials across seven domains, including "sequence generation", "allocation concealment", "blinding of participants", "personnel and outcome assessors", "incomplete outcome data", "selective outcome reporting", and "other potential threats to validity". Each domain had three possible judgements: "low risk", "high risk" or "unclear risk". If necessary, we consulted a third author (CJ) to resolve disagreements.
Unit of analysis issues
We included both RCTs and randomised cross‐over trials. When incorporating cross‐over trials into a meta‐analysis, we followed the approach suggested by Elbourne 2002. For example, when we considered neither carry‐over nor period effects a problem, and the mean and standard deviation values of the participant‐specific difference between experimental and control group measurements were available, we took all measurements from experimental and control periods for analysis.
Data synthesis
We calculated dichotomous outcomes as risk ratios (RRs), and continuous data as mean differences (MDs) or standardized mean differences (SMDs). We analysed statistical heterogeneity using the Chi2 test (P value > 0.10 for statistical significance) and quantified using the I2 statistic (Higgins 2003). If heterogeneity was not significant (P＞0.05, I2 < 50%), we performed a quantitative data synthesis (meta‐analysis) using a fixed‐effect model. Where heterogeneity was significant (P ＜0.05), we analysed the cause to determine whether qualitative differences existed. If the clinical heterogeneity was not due to qualitative differences, we applied a random‐effects model. Otherwise, we conducted a qualitative description rather than quantitative data synthesis. We used funnel plot analysis (Egger 1997) to assess for publication bias.
Results
Description of studies
For the original review, our systematic searches (up to August 2005) retrieved 800 references. After reading titles and abstracts, 35 articles were retrieved in full text for further assessment. Of these, 15 articles were excluded based on the inclusion criteria (Characteristics of excluded studies), and we included a total of 20 articles. In addition, 23 articles were classed as 'awaiting classification'.
For the first review update, we re‐run the electronic search strategies up to September 2008, and 3621 new references were found. After reading titles and abstracts, 51 articles and 23 'awaiting classification' articles were retrieved in full text for further assessment. A total of 69 of these 74 articles were excluded (Characteristics of excluded studies). Three articles were included (Fang 1987; Liu 2007a; Zhao 2006), and two were categorized as 'studies awaiting classification' (Fan 2004; Jiang 1988). Meanwhile, we reassessed the 20 previously‐included articles using the updated inclusion criteria and 17 articles were excluded (Characteristics of excluded studies), and three remained as included studies (He 2004; Mao 2003a; Zhang 2002). Therefore, a total of six articles were included in the first update.
For the second review update, search strategies were re‐run to April 2011, and 1563 new articles were found. After reading titles and abstracts, 32 articles were retrieved in full text for further assessment. Full text of the two studies previously categorized as 'studies awaiting classification' were also retrieved (Fan 2004; Jiang 1988). A total of 31 of these 34 full‐text articles were excluded (Characteristics of excluded studies), and three new articles were included for this second review update (Jiang 1988; Wang 2010a; Zhai 2009).
For the current review update, we performed searches up to 13 June 2013. We identified a total of 1612 new articles (Figure 1). After screening titles and abstracts, we retrieved the full text articles of 37 studies for further assessment. Of these, we excluded 33 articles, for the reasons listed in the Characteristics of excluded studies. We included four new articles in this update (Ding 2012; Su 2012; Wan 2012; Liu 2013) and nine articles from the previous reviews, to give a total of 13 articles.
Study flow diagram of search results for update of 2014 Cochrane Review.
We have listed the characteristics of the included trials in the Characteristics of included studies section. One article contained two independent RCTs (Mao 2003a). One trial compared Shengmai plus usual treatment with usual treatment alone in 40 patients (Mao 2003a, part A); the other RCT compared Shengmai with placebo in the other 40 participants (Mao 2003a, part B). We included both RCTs but the trial did not originally report any information on the 80 participants. We obtained information of Mao 2003a, part A from its additional paper (Mao 2003b). However, we failed to find participants' information in the latter RCT (Mao 2003a, part B).
Therefore, we included a total of fourteen trials in this review update. Twelve trials were RCTs (Zhang 2002; Mao 2003a, Part A; Mao 2003a, Part B; He 2004; Zhao 2006; Liu 2007a; Zhai 2009; Wang 2010a; Ding 2012; Su 2012; Wan 2012; Liu 2013), and two were cross‐over trials (Fang 1987; Jiang 1988). We included 858 participants, of which 54.7% were male, and a mean age ranging from 41 to 70.4 years. The leading causes of heart failure were coronary heart disease, dilated cardiomyopathy and hypertensive heart disease (accounting for 87.5%) (Table 1; Table 2).
| Trial | Number | Sex (male/female) | Average age/ (Age range) | Duration of heart failure | Original disease | Heart function |
| 40 | 35/5 | 58/(39 to 80) | CHD (n = 40) | NYHA II (n = 37), NYHA III (n = 3) | ||
| 60 | 33/27 | 41 | CHD (n = 21), RHD (n = 12), HBP (n = 15), DCM (n = 12) | NYHA II (n = 8), NYHA III (n = 34), NYHA IV (n = 18) | ||
| 120 | 64/56 | /(38 to 75) | 3 to 12 years | CHD (n = 52), HBP (n = 42), DCM (n = 11), RHD (n = 15) | NYHA II n = 31), NYHA III n = 52), NYHA IV n = 37) | |
| Mao 2003a, Part A | 40 | 17/23 | 59/(40 to 70) | 6.70 ± 6.50 years | CHD (n = 33), RHD (n = 2), HBP (n = 2), PHD (n = 2) | NYHA II (n = 13), NYHA III (n = 27) |
| Mao 2003a, Part B | 40 | NYHA II (n = 14), NYHA III (n = 26) | ||||
| 100 | 59/41 | /(34 to 78) | 0.5 to 8 years | DCM (n = 100) | NYHA II (n = 20), NYHA III (n = 71), NYHA IV (n = 9) | |
| 40 | 14/26 | 61(41 to 70) | 0.5 to 30 years | CHD (n = 35), RHD (n = 3), PHD (n = 2) | NYHA II (n = 18), NYHA III (n = 22) | |
| 60 | 34/26 | /(50 to 75) | 2 to 14 years | CHD (n = 39), RHD (n = 7), PHD (n = 2), DCM (n = 4), HBP (n = 6), congenital heart disease (n = 2) | NYHA III (n = 35), NYHA IV (n = 25) | |
| 74 | 43/31 | /(60 to 85) | CHD (n = 41), HBP (n = 29), DCM (n = 4) | NYHA II (n = 7), NYHA III (n = 45), NYHA IV (n = 22) | ||
| 26 | 23/3 | 43.8/(23 to 60) | DCM (n = 26) | NYHA II (n = 23), NYHA III (n = 3) | ||
| 66 | 32/34 | /(44 to 85) | CHD (n = 38), RHD (n = 5), HBP (n = 11), PHD (n = 12) | NYHA II (n = 14), NYHA III (n =39), NYHA IV (n = 13) | ||
| 50 | 33/17 | /(60 to 75) | NYHA III (n = 33), NYHA IV (n = 17) | |||
| 62 | 28/34 | 62.4/(48 to 70) | 4 to 11 year/6.9 ± 2.9 years | HBP (n = 26), CHD (n = 29), diabetes mellitus (n = 32) | NYHA II (n = 22), NYHA III (n = 30), NYHA IV (n = 10) | |
| 80 | 33/47 | 70.74/72.56 | 8,07± 6.30 years/8.92 ± 4.68 years | Old myocardial infarction (n = 25), hypertensive heart disease (n = 56), diabetes mellitus (n = 33), hyperlipidaemia (n = 26) | NYHA III (n = 61), NYHA IV (n = 19) |
CHD = coronary heart disease; RHD = rheumatic heart disease; PHD = pulmonary heart disease; DCM = dilated cardiomyopathy; HBP =hypertensive heart disease; HCM = hypertrophic cardiomyopathy; NYHA = New York Heart Association.
| Trial | Agent | Dose | Form | Duration of treatment (days) |
| Shanghai | 10 mL po bid | oral liquid | 20 | |
| Huaxi | 20 to 40 mL + GS /iv gtt qd | injection | 14 | |
| Shanxi | 20 mL/40 mL/60 mL 5% GS 250 mL/iv gtt qd | injection | 15 | |
| Mao 2003a, Part A | Yibin | 20, 40, 60 mL + 5% 200 mL polarized solution /iv gtt qd | injection | 14 |
| Mao 2003a, Part B | Yibin | 20, 40, 60 mL + 5% 200 mL polarized solution /iv gtt qd | injection | 14 |
| Huaxi | 60 mL + 5% GS 250ml/iv gtt qd | injection | 14 | |
| Huaxi | 20 mL/40 mL/60 mL + 5% GS 100 mL + 0.25 g KCL + R‐I 1u | injection | 7 | |
| Shanghai | 10 mL po bid | oral liquid | 20 | |
| not stated | 40 mL + 5% GS 250 mL/iv gtt qd | injection | 15 | |
| Jiangsu | 60 mL + GS 250 mL/iv gtt qd | injection | 15 | |
| not stated | 100 to 180 mL + 5% GS 150 mL/iv gtt qd | injection | 10 | |
| not stated | 100 mL po bid | powder | 14 | |
| Suzhong | 40 mL + 5% GS 100 mL/iv gtt qd | injection | 14 | |
| Shanxi | 40 mL + 5% GS 250 mL or 0.9% NS 250ml/iv gtt qd | injection | 7 |
po = by mouth; iv gtt = intravenously guttae; GS = glucose and saline.
The wash‐out period in both Fang 1987 and Jiang 1988 was 10 days, which met our inclusion criteria (9.5 days). Fang 1987 proved they were free from carry‐over effect in two ways. Firstly, there were no significant difference (P ＜ 0.05) in serial indexes of the heart function between two groups at the beginning of both phases. Secondly, the overall level of heart function of the participants at the end of phase 1 was not significantly different from that at the end of phase 2.
We assessed the quality of the evidence using the GRADE approach and almost all of the evidence was low quality (summary of findings Table for the main comparison; summary of findings Table 2).
Methodological quality
All of the included trials were RCTs. Many studies provided a description of "patients were randomly allocated to..." without further detailed information. We contacted the authors for more information of methodology by telephone and eliminated 1608 studies.
We could not obtain the protocols of included trials. Few trials described the blinding method. Two trials mentioned the use of double‐blinding (Fang 1987; Jiang 1988), and one assessed one of the reported outcomes blindly (He 2004). None of the included trials mentioned allocation concealment. Only one trial provided detailed information of baseline comparisons other than a description "baseline characteristics are similar between Shengmai group and control group (P > 0.05)" (Zhao 2006).
Participants
All of the 858 participants were Chinese (Table 1), 52.2% (448/858) were male and 43.1% (370/858) were female (Note: as Mao 2003a, Part B did not report any information on its participants, the statistical data did not include these 40 participants). The participants' ages varied from 23 to 85 years, and the mean age ranged from 41 to 70.4 years.
Seven trials reported the duration of heart failure (Zhang 2002; Mao 2003a; Zhao 2006; Liu 2007a; Zhai 2009; Ding 2012; Wan 2012), which varied from 0.5 to 30 years.
The spectrum of primary causes for heart failure differed among the 12 trials that reported this data (Fang 1987; Jiang 1988; Zhang 2002; Mao 2003a, Part A; He 2004; Zhao 2006; Liu 2007a; Zhai 2009; Wang 2010a; Ding 2012; Su 2012; Wan 2012). Of these, 328 patients were coronary heart disease patients from nine trials (Fang 1987; Mao 2003a, Part A; He 2004; Liu 2007a; Zhao 2006; Zhai 2009; Wang 2010a; Su 2012; Wan 2012), 157 patients were dilated cardiomyopathy patients from six trials (Jiang 1988; Zhang 2002; He 2004; Liu 2007a; Zhai 2009; Wang 2010a), 44 patients were rheumatic heart disease patients from six trials (Mao 2003a, Part A; He 2004; Zhao 2006; Liu 2007a; Zhai 2009; Su 2012), 187 patients were hypertensive heart disease patients from eight trials (Mao 2003a, Part A; He 2004; Liu 2007a; Zhai 2009; Wang 2010a; Ding 2012; Su 2012; Wan 2012); 18 patients were pulmonary heart disease from four trials (Mao 2003a, Part A; Zhao 2006; Zhai 2009; Su 2012), and two patients were congenital heart disease patients from Zhai 2009. Twenty‐five patients had old myocardial infarction in Ding 2012, 65 patients in two trials had diabetes mellitus (Ding 2012; Wan 2012), and 26 patients in one study had hyperlipidaemia (Ding 2012). Two trials did not report the primary cause of heart failure (Mao 2003a, Part B; Liu 2013).
All included trials provided baseline data on NYHA function classifications (NYHA FC). There were 207 participants with NYHA class II, 481 with NYHA class III and 170 with NYHA class IV.
Diagnosis
Eleven trials stated the diagnostic criteria of heart failure: three based on criteria from the World Health Organization (WHO) (Fang 1987; Jiang 1988; Zhang 2002), three based on Framingham and Boston (Mao 2003a, Part A; Mao 2003a, Part B; Zhao 2006; Ding 2012); one from Guidelines for diagnosis and treatment of chronic heart failure 2007 (Wang 2010a); one based on Guidelines for treatment of systolic heart failure (Zhai 2009); one based on Reference standards for the treatment of left ventricle diastolic heart failure (Wan 2012); one based on Guidelines for clinical research on the treatment of new Chinese traditional medicine in congestive heart failure (Liu 2013). The remaining three articles did not mention the diagnostic criteria, but provided the NYHA class (He 2004; Liu 2007a; Su 2012). All included trials adopted NYHA class for clinical assessment of patients, and one also used the specific activity scale (SAS) (Zhao 2006).
Intervention
Eleven trials administered Shengmai by intravenous infusion (Zhang 2002; Mao 2003a,Part A; Mao 2003a, Part B; He 2004; Zhao 2006, Liu 2007a; Zhai 2009; Wang 2010a; Ding 2012; Su 2012; Wan 2012), and three trials gave it as an oral liquid (Fang 1987; Jiang 1988, Liu 2013). All trials but three (Zhai 2009; Su 2012; Liu 2013) reported the manufacturers of Shengmai and included: Huaxi (Zhang 2002; He 2004; Zhao 2006), Shanghai (Fang 1987; Jiang 1988), Shanxi (Liu 2007a; Ding 2012), Yibin (Mao 2003a), Jiangsu (Wang 2010a) and Suzhong (Wan 2012).
Shengmai dosage varied: the injection ranged from 20 mL to 180 mL per day and according to Pharmacopoeia of the People's Republic of China, 10 mL Shengmai injection is extracted based on a standardized formula: Panax ginseng (1.00 g), Ophiopogon japonicu (3.12 g) and Schisandra chinensis (1.56 g); the volume of oral liquid was 20 mL per day (Jiang 1988) (Panax ginseng,Ophiopogon japonicu and Schisandra chinensis were 11 g in total) or 200 mL per day (Liu 2013) (including: Panax ginseng (24.00 g), Ophiopogon japonicu (24.00 g) and Schisandra chinensis (12.00 g)).
Treatment duration varied from seven days to 20 days, and the mean duration was 13.79 days (SD 3.81).
No study reported the duration of follow up.
Eleven trials compared Shengmai plus usual treatment with usual treatment alone. Of the eleven trials, four provided detailed information about usual treatment, including drug names and administration of medicines (Zhang 2002; Mao 2003a, Part A; Zhao 2006; Su 2012). There were some differences amongst them regarding usual treatment. Seven trials (He 2004; Liu 2007a; Zhai 2009; Wang 2010a; Ding 2012; Wan 2012; Liu 2013) described usual treatment as cardiotonic, diuretic and vasodilator instead of giving detailed information. Three trials compared Shengmai with placebo (Fang 1987; Jiang 1988; Mao 2003a, Part B). We have listed the intervention details in Table 2.
Outcomes
Regarding primary outcomes, only one trial reported on mortality. Twelve trials reported the NYHA classification of clinical status (Zhang 2002; Mao 2003a,Part A; Mao 2003a, Part B; He 2004; Zhao 2006; Liu 2007a; Zhai 2009; Wang 2010a; Ding 2012; Su 2012; Wan 2012; Liu 2013). Three trials reported mild adverse effects in six patients: in the high‐dose (60 mL/d) Shengmai group, two had mild asomnia (2/30, 6.67%) (Liu 2007a), two had dry mouth and were fidgety (2/10, 20%) (Zhao 2006), two patients in the Shengmai group (20 mL/d) had stomach discomfort and hypoglycaemia (2/26 7.69%) and were withdrawn from Jiang 1988. Four trials reported no adverse effects at the end of treatment (He 2004; Zhai 2009; Wang 2010a; Ding 2012). Seven trials did not report on this outcome (Fang 1987; Zhang 2002; Mao 2003a,Part A; Mao 2003a, Part B; Su 2012; Wan 2012; Liu 2013).
For secondary outcomes, none of the included trials reported hospitalisation, rehospitalization or costs. Two trials reported exercise test (Jiang 1988; He 2004); one trial reported quality of life ("Minnesota Living with Heart Failure Questionnaire" adopted) (Zhai 2009); eight trials reported the change in haemodynamics (Fang 1987; Jiang 1988; Zhang 2002; Liu 2007a; Zhai 2009; Wang 2010a; Ding 2012, Wan 2012); four trials reported the change in ejection fraction (Zhang 2002; Liu 2007a; Zhai 2009; Ding 2012) and only one trial reported the change of quality of life (Zhai 2009).
No trial reported changes in medication due to worsening of symptoms or heart failure. All fourteen trials reported the outcomes measured at the end of treatment.
Risk of bias in included studies
We presented the risk of bias of included trials in the Risk of bias tables in the 'Characteristics of included studies' section. We have summarized the risk of bias in each domain for each included trial in Figure 2 and as percentages across all included trials in Figure 3.
Risk of bias summary: review authors' judgements about each risk of bias item for each included trial.
Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included trials.
Although all included trials "randomly allocated" participants, only seven trials reported on sequence generation (Fang 1987; Liu 2007a; Zhai 2009; Wang 2010a; Ding 2012; Wan 2012; Liu 2013). We contacted the authors of the remaining seven trials by telephone for the detailed randomisation method used in their trials. The methods of randomisation sequence generation included computer random number generator (Zhang 2002; Ding 2012), throwing dice (Fang 1987; Jiang 1988; He 2004), and random number table (Mao 2003a, Part A; Mao 2003a, Part B; Zhao 2006; Liu 2007a; Zhai 2009; Wang 2010a; Su 2012; Wan 2012; Liu 2013). None of the included trials reported allocation concealment. Although thirteen of fourteen trials stated there were no significant differences at baseline between treatment and control groups (Fang 1987; Jiang 1988; Zhang 2002; Mao 2003a, Part A; Mao 2003a, Part B; Zhao 2006; Liu 2007a; Zhai 2009; Wang 2010a; Ding 2012; Su 2012; Wan 2012; Liu 2013), only Zhao 2006 detailed baseline characteristics data and He 2004 did not report on baseline comparisons. It was difficult to assess risk of bias in a domain such as sequence generation as baseline data were not available for 13 of 14 trials, so selection bias was still unclear and likely to exist.
Two cross‐over trials compared Shengmai to placebo used double‐blinding (Fang 1987; Jiang 1988). He 2004 assessed blindly one of the reported outcomes. Other trials did not provide information about blinding and there was high risk of performance bias and detection bias in these trials.
Apart from two participants who were withdrawn due to adverse effects in the Shengmai group in Jiang 1988, trial investigators reported all the other participants' outcomes. As we could not obtain any original trial protocols, the risk of bias of incomplete outcome assessment was unclear. Also, the risk of selective reporting bias was unclear. We generated a funnel plot for NYHA class (Figure 4), and found no evidence of reporting bias. However, there were too few included trials for other outcomes to generate a funnel plot.
Funnel plot of comparison: 1. Shengmai plus usual treatment versus usual treatment alone, outcome: 1.1 Lack of improvement in heart failure (NYHA class improved < I class or worsening of heart failure).
Zhao 2006 was funded by State Chinese Medicine Administration Bureau, Zhang 2002 was under the support of the National Science and Technology Department's Ninth Five‐year Plan, and Mao 2003a was funded by the Education Commission of Tianjin Province. The remaining trials were not funded by any institution and all funding was from public authorities. Based on sources of funding, the included trials were unlikely to have had any conflict of interests.
We had insufficient information to determine bias from other sources.
Effects of interventions
See: Summary of findings for the main comparison Shengmai plus usual treatment compared to usual treatment alone for heart failure; Summary of findings 2 Shengmai compared to placebo for heart failure
Shengmai plus usual treatment versus usual treatment alone
Mortality
There were no data available on this outcome.
NYHA class
Ten trials (Zhang 2002; Mao 2003a, Part A; He 2004; Zhao 2006; Liu 2007a; Zhai 2009; Wang 2010a; Su 2012; Wan 2012; Liu 2013), including 672 participants, provided data on NYHA class. A meta‐analysis of data on the change of NYHA classification showed that Shengmai plus usual treatment was more beneficial compared with usual treatment alone. The summary relative risk of NYHA FC improved < I class, or worsening of heart failure was 0.37 (95% CI 0.26 to 0.51; 10 trials, 672 participants; Analysis 1.1).
Exercise test or six‐minute walk test performance
He 2004 showed that Shengmai plus usual treatment was better than usual treatment alone for improving exercise tolerance (MD 6.00, 95% CI 4.48 to 7.52; one trial, 60 participants; Analysis 1.2).
Quality of life
Zhai 2009 showed that Shengmai plus usual treatment was not significantly different from usual treatment alone (MD 7.00, 95% CI ‐1.98 to 15.98; one trial, 60 participants; Analysis 1.3).
Ejection fraction
Four trials with 354 participants presented data on ejection fraction (Zhang 2002; Liu 2007a; Zhai 2009; Wang 2010a). We detected considerable heterogeneity (P < 0.00001, I² = 91%) and we conducted meta‐analysis using the random‐effects model. Shengmai plus usual treatment was better than usual treatment alone in improving ejection fraction (MD 9.02, 95% CI 5.95 to 12.08; four trials, 354 participants; Analysis 1.4). The intervention effect on this outcome had statistical significance in each of the four trials.
Cardiac output
Four trials including 360 participants presented data on this outcome (Zhang 2002; Liu 2007a; Zhai 2009; Ding 2012). There was considerable heterogeneity (P < 0.00001, I² = 92%) and we conducted random‐effects meta‐analysis. Shengmai plus usual treatment was better than usual treatment alone for improving cardiac output (MD 0.94, 95% CI 0.26 to 1.63; four trials, 360 participants; Analysis 1.5) and the intervention effect had statistical significance in each of the four trials.
Stroke volume
Three trials which recruited 240 participants reported on stroke volume (Zhang 2002; Zhai 2009; Ding 2012). Substantial heterogeneity was present (P < 0.00001, I² = 97%) and we conducted random‐effects meta‐analysis. Shengmai plus usual treatment was not significantly different from usual treatment alone in increasing stroke volume (MD 8.53, 95% CI 0.08 to 16.98; three trials, 240 participants; Analysis 1.6).The intervention effect on this outcome had statistical significance in each of the three trials.
Cardiac index
Two trials with 160 participants reported on this outcome (Zhang 2002; Zhai 2009). There was considerable heterogeneity (P = 0.04, I² = 77%) and we undertook random‐effects meta‐analysis. Shengmai plus usual treatment was better than usual treatment alone at changing cardiac index (MD 0.79, 95% CI 0.59 to 0.99; two trials, 160 participants; Analysis 1.7).The intervention effect on this outcome had statistical significance in each of the two trials.
Left ventricular end‐diastolic volume
Wang 2010a studied this outcome in 74 participants and showed that there was no significant difference between Shengmai plus usual treatment and usual treatment alone (MD 7.20, 95% CI ‐9.80 to 24.20; one trial, 74 participants; Analysis 1.8).
Left ventricular end‐systolic volume
Wang 2010a, with 74 participants, reported on left ventricular end‐systolic volume and showed that Shengmai plus usual treatment was better than usual treatment alone for improving this outcome (MD 14.60, 95% CI 0.50 to 28.70; one trial, 74 participants; Analysis 1.9).
Hospitalization/rehospitalization
No data were available on this outcome.
Costs
No data were available on this outcome.
Shengmai versus placebo
Mortality
No data were available on this outcome.
NYHA class
Mao 2003a, Part B, which included 40 participants, showed that Shengmai treatment was more beneficial than placebo treatment at improving NYHA classification of clinical status. The summary relative risk of NYHA FC improved < I class, or worsening of heart failure was 0.22 (95% CI 0.11 to 0.44; one trial, 40 participants; Analysis 2.1).
Exercise test (time change)
Jiang 1988 demonstrated that Shengmai treatment was not significantly different from placebo treatment changing exercise test (MD ‐3.56, 95% CI ‐7.19 to 0.07; one trial, 18 participants; Analysis 2.2).
Ejection fraction
Jiang 1988 showed that Shengmai treatment was better than placebo treatment improving ejection fraction (MD 12.45, 95% CI 8.96 to 15.94; one trial, 52 participants; Analysis 2.3).
Cardiac output
Jiang 1988 showed that Shengmai treatment was not significantly different from placebo treatment at increasing cardiac output (MD ‐0.53, 95% CI ‐1.08 to 0.02; one trial, 52 participants; Analysis 2.4).
Stroke volume
Fang 1987 demonstrated that Shengmai treatment was better than placebo treatment at increasing stroke volume (MD 5.80, 95% CI 0.16 to 11.44; one trial, 80 participants; Analysis 2.5).
Stroke volume index
Fang 1987 showed that Shengmai treatment was not significantly different from placebo treatment changing stroke volume index (MD 3.95, 95% CI ‐0.10 to 8.00; one trial, 80 participants; Analysis 2.6).
Cardiac index
Fang 1987 demonstrated that Shengmai treatment was better than placebo treatment for changing cardiac index (MD 0.32, 95% CI 0.04 to 0.60; one trial, 80 participants; Analysis 2.7).
Myocardial contractility
Fang 1987 showed that Shengmai treatment was better than placebo treatment for increasing Heather Index (HI) (MD 1.88, 95% CI 0.38 to 3.38; one trial, 80 participants; Analysis 2.8).
Fang 1987 provided data for ECG change and demonstrated that Shengmai treatment was better than placebo treatment for changing Q‐Z (MD 10.72, 95% CI 5.10 to 16.34; one trial, 80 participants; Analysis 2.9) and Adz/dtMax (MD 0.10, 95% CI 0.01 to 0.19; one trial, 80 participants; Analysis 2.10), but was not significantly different from placebo treatment at changing Q‐B/B‐X (MD 0.03, 95% CI ‐0.00 to 0.06; one trial, 80 participants; Analysis 2.11).
Hospitalization/rehospitalization
No data were available on this outcome.
Costs
No data were available on this outcome.
Adverse effects
Three trials reported mild adverse effects in six patients: mild asomnia occurred in 2/90 patients (2.22%) within the Shengmai group (Liu 2007a), and both were in the high‐dose (60 mL/d) Shengmai group (2/30, 6.67%); dry mouth and fidgety responses occurred in 2/30 patients (6.67%) in the Shengmai group (Zhao 2006), and both occurred in the high‐dose (60 mL/d) Shengmai group (2/10, 20%); stomach discomfort and hypoglycaemia occurred in 2/26 patients (7.69%) within the Shengmai group and they were withdrawn from the trial (Jiang 1988). Four trials reported no adverse effects at the end of treatment (He 2004; Zhai 2009; Wang 2010a; Ding 2012). Seven trials did not report on this outcome (Fang 1987; Zhang 2002; Mao 2003a Part A; Mao 2003a Part B; Su 2012; Wan 2012; Liu 2013)
Sensitivity analyses and subgroup analyses
We did not perform any sensitivity analyses because the included trials were of similar low quality. We were unable to perform any subgroup analyses because of insufficient sample size and incomplete data of heart failure types.
We made a thorough investigation of heterogeneity, but there was insufficient information to explain the cause of the heterogeneity, so we still applied the random‐effects model for these indexes.
Discussion
This review suggests that Shengmai (injection, oral liquid and powder forms) may have some positive effects on heart failure, but the results should be interpreted with great caution due to the poor quality of the included trials.
For the comparison of Shengmai plus usual treatment versus usual treatment, potential benefits of Shengmai were observed for several outcomes, including upgrading NYHA FC and better performances in exercise test, ejection fraction, cardiac output, cardiac index and left ventricular end‐systolic volume. Shengmai's positive effects were also reported in ejection fraction, stroke volume, cardiac index and myocardial contractility (Heather index, Q‐Z, Adz/dtMax) when Shengmai treatment was compared to placebo treatment in patients with suspected heart failure. In this updated review, no of the included trials reported on hospitalisation. Hospitalization may be a useful outcome for evaluation in Western countries. However, it is difficult to measure accurately in China due to the imbalance in the allocation of medical resources, difficulty in accessing medical treatment in secluded regions, and relatively sparse medical resources. These make hospitalisation unreliable or even misleading. In addition, only one trial ( Zhai 2009) reported on the quality of life and no trials reported on economic outcomes. All of the included trials had relatively short treatment periods without a long‐term follow‐up, and we could not obtain data on the primary outcome, mortality.
All of the participants in the included trials were Chinese. Traditional Chinese medicine is widely believed in China to be effective for most diseases, with few adverse effects (Tan 2006; Song 2011). Therefore, Chinese herbs and medicine are used more extensively in hospitals in China than in Western countries. All of the participants were from major hospitals and were more likely under more severe medical conditions. These factors could undermined the representativeness of the sample.
The lack of detailed description of randomisation sequence generation and data of baseline characteristics limited our confidence in the random allocation and baseline balance of the included trials. Along with the absence of allocation concealment in all the trials, which could cause selective enrolment of participants on the basis of prognostic factors or other subjective factors, we expected a possible risk of selective bias towards greater effect of the intervention. Almost all of these included trials were conducted without blinding, which caused a high risk of performance bias and detection bias. Lack of blinding in randomised trials is associated with exaggerated results of intervention effect (Pildal 2007). In particular, one of the primary outcomes NYHA FC was a subjective endpoint assessed by physicians and this was very likely to exaggerate the intervention effect. For the three included trials which compared Shengmai with placebo, the lack of blinding of participants also gave rise to the high risk of placebo effect. Although we conducted comprehensive searches and tried to avoid language and location bias, we only found trials published in Chinese. Most trials had positive Shengmai findings and thus we cannot exclude potential publication bias. The overall methodological quality of the trials was poor and care must be taken when interpreting the outcomes.
In the review process, two authors worked independently on trial collection, data extraction and risk of bias assessment. We reached consensus through discussion or in consultation with a third author. The potential biases in the review process were low, but may still exist for inevitable subjective factors when we assessed the risk of bias and the unexplainable significant heterogeneity in the random‐effects meta‐analysis for the outcomes, including ejection fraction, cardiac output, stroke volume, and cardiac index.
Six out of the 858 participants had mild adverse effects, and the rate of adverse effects was 0.70%. Four of the six participants with adverse effects were in the high‐dose Shengmai group. The risk ratio of high dose of Shengmai (> 60 mL per day) was 3.46. Our review suggests that high dose of Shengmai injection would more likely cause adverse effects. Another systematic review (Li 2009) conducted with a total of 28,305 patients using Shangmai reported that 215 patients suffered from adverse events, with a adverse effect rate of 0.76%. Li 2013 conducted a safety evaluation in 2013 on Shengmai in 515 patients. The rate of adverse events in this retrospective study was 0.78% (4/515). Another safety evaluation on Shengmai in 4079 patients (Deng 2012) reported the incidence of adverse events of Shengmai injection was 0.12% (95% CI 0.04% to 0.29%). However, a retrospective study in 1012 patients (Cheng 2011) reported that "main adverse events of Shengmai included 388 cases of fever and systemic damages (38.34%), 202 cases of skin and appendages damages (19.96% ), 113 cases of cardiovascular system damages (11.17%), and 70 cases of gastrointestinal system damages (6.92%)". The adverse rate of the last study was obviously higher than those in the former listed studies, including our review, but we could not find the reason. The inconsistent results of the adverse events should be monitored in the further use of Shengmai. Allergic reaction was most common adverse event, which may result from Individual constitution, age, overdose and unreasonable drug combination. In addition, clinicians should be aware of the rare possibility of anaphylactic shock when applying Shengmai.
According to traditional Chinese medicine theory, heart failure is a deficiency of Yin and Qi, but it is difficult to standardize and quantify as diagnosis is usually based on the clinician's subjective experience. Traditional Chinese medicine scholars have long been devoted to establishing a classification standard for syndromes of Chinese medicine to make it more quantitative, objective and acceptable, by which it is possible to measure deficiency of Yin and Qi of heart failure patients as NYHA classification standards measures heart function. Many problems still exist with these evaluation methods and no uniform standards has been formed (Wang 2008).
Future research needs to employ high quality methodology and more meaningful outcomes relevant to efficacy and safety for long‐term follow‐up. If efficacy is shown in well‐designed trials, isolating the active components of Shengmai and testing it in a different cultural backgrounds is necessary. If there is sufficient evidence in the future, we will attempt to conduct subgroup analysis to determine the effect of Shengmai on patients with different types of heart failure and of different ethnic backgrounds. Moreover, the method and dosage of administration should be analysed to explore the optimal therapeutic regimen of Shengmai for heart failure.
Study flow diagram of search results for update of 2014 Cochrane Review.
Risk of bias summary: review authors' judgements about each risk of bias item for each included trial.
Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included trials.
Funnel plot of comparison: 1. Shengmai plus usual treatment versus usual treatment alone, outcome: 1.1 Lack of improvement in heart failure (NYHA class improved < I class or worsening of heart failure).
Comparison 1 Shengmai plus usual treatment versus usual treatment alone, Outcome 1 Lack of improvement in heart failure (NYHA class improved < I class or worsening of heart failure).
Comparison 1 Shengmai plus usual treatment versus usual treatment alone, Outcome 2 Exercise test (after treatment).
Comparison 1 Shengmai plus usual treatment versus usual treatment alone, Outcome 3 Change in quality of life.
Comparison 1 Shengmai plus usual treatment versus usual treatment alone, Outcome 4 Change in ejection fraction (%).
Comparison 1 Shengmai plus usual treatment versus usual treatment alone, Outcome 5 Change in cardiac output (L/min).
Comparison 1 Shengmai plus usual treatment versus usual treatment alone, Outcome 6 Change in stroke volume (mL).
Comparison 1 Shengmai plus usual treatment versus usual treatment alone, Outcome 7 Change in cardiac index L/min*m².
Comparison 1 Shengmai plus usual treatment versus usual treatment alone, Outcome 8 Change in left ventricular end‐diastolic volume (mL).
Comparison 1 Shengmai plus usual treatment versus usual treatment alone, Outcome 9 Change in left ventricular end‐systolic volume (mL).
Comparison 2 Shengmai versus placebo, Outcome 1 Lack of improvement in heart failure (NYHA class improved < I class or worsening of heart failure).
Comparison 2 Shengmai versus placebo, Outcome 2 Change in exercise test (mins).
Comparison 2 Shengmai versus placebo, Outcome 3 Change in ejection fraction (%).
Comparison 2 Shengmai versus placebo, Outcome 4 Change in cardiac output (L/min).
Comparison 2 Shengmai versus placebo, Outcome 5 Change in stroke volume (mL).
Comparison 2 Shengmai versus placebo, Outcome 6 Change in stroke volume index.
Comparison 2 Shengmai versus placebo, Outcome 7 Change in cardiac index L/min*m².
Comparison 2 Shengmai versus placebo, Outcome 8 Change in Heather Index.
Comparison 2 Shengmai versus placebo, Outcome 9 Change in ECG: Q‐Z.
Comparison 2 Shengmai versus placebo, Outcome 10 Change in ECG: Adz/dtMax.
Comparison 2 Shengmai versus placebo, Outcome 11 Change in ECG: Q‐B/B‐X.
| Shengmai plus usual treatment compared to usual treatment alone for heart failure | ||||||
| Patient or population: patients with heart failure | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | No of participants | Quality of the evidence | Comments | |
| Assumed risk | Corresponding risk | |||||
| Usual treatment alone | Shengmai plus usual treatment | |||||
| Mortality | See comment | See comment | Not estimable | 858 (11 trials) | See comment | 0 participants died |
| Lack of improvement in heart failure (NYHA class improved < I class or worsening of heart failure) | Moderate | RR 0.37 | 672 | ⊕⊕⊝⊝ | This outcome was transformed to be dichotomous and NYHA class improved < I class or worsening of heart failure was the event | |
| 279 per 1000 | 103 per 1000 | |||||
| Adverse effects | See comment | See comment | Not estimable | 4 (2 trials) | See comment | Two mild asomnia (1 trial) and two dry mouth and fidgety (1 trial) |
| Mean exercise tolerance test (after treatment) (min) | 6 mins | 6 higher | 60 | ⊕⊕⊝⊝ | The longer the exercise time, the better the heart function. | |
| Mean change in quality of life | 20 points | 7 higher | 60 | ⊕⊕⊝⊝ | ||
| Mean change in ejection fraction (%) | 3 to 13.1 | 9.02 higher | 354 | ⊕⊝⊝⊝ | ||
| Mean change in cardiac output (L/min) | 0.5 to 0.83 | 0.94 higher | 360 | ⊕⊝⊝⊝ | ||
| Mean change in stroke volume (mL) | 8.02 to 16.9 | 8.43 higher | 240 | ⊕⊝⊝⊝ | ||
| Change in cardiac index (L/min*m²) | 0.4 to 0.44 | 0.79 higher | 160 | ⊕⊝⊝⊝ | ||
| Hospitalization/rehospitalization | See comment | See comment | Not estimable | 858 (12 trials) | See comment | 0 participants hospitalized or re‐hospitalized |
| *The basis for the assumed risk (e.g. the median control group risk across trials) is provided in the 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). | ||||||
| GRADE Working Group grades of evidence | ||||||
| 1 Downgraded by 1 level for lacking of blinding and allocation concealment | ||||||
| Shengmai compared to placebo for heart failure | ||||||
| Patient or population: patients with heart failure | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | Number of participants | Quality of the evidence | Comments | |
| Assumed risk | Corresponding risk | |||||
| Placebo | Shengmai | |||||
| Mortality | See comment | See comment | Not estimable | 106 (3 trials) | See comment | 0 participants died |
| Lack of improvement in heart failure (NYHA class improved < I class or worsening of heart failure) | Moderate | RR 0.22 | 40 | ⊕⊕⊝⊝ | This outcome was transformed to be dichotomous and NYHA class improved < I class or worsening of heart failure was the event | |
| 1000 per 1000 | 220 per 1000 | |||||
| Adverse effects | See comment | See comment | Not estimable | 2 (1 trial) | See comment | One stomach discomfort and one hypoglycaemia were reported in one trial. |
| Mean change in exercise test (mins) | 0.11 | 3.56 lower | 18 | ⊕⊕⊝⊝ | The longer the exercise time, the better the heart function. | |
| Mean change in ejection fraction (%) | 0.2 | 12.45 higher | 52 | ⊕⊕⊝⊝ | ||
| Mean change in cardiac output (L/min) | 0.02 | 0.53 lower | 52 | ⊕⊕⊝⊝ | ||
| Mean change in cardiac index (L/min*m²) | 0.03 | 0.32 higher | 80 | ⊕⊕⊝⊝ | ||
| Mean change in ECG: Q‐Z | 1.48 | 10.72 higher | 80 | ⊕⊕⊝⊝ | ||
| Hospitalization/ rehospitalization | See comment | See comment | Not estimable | 106 (3 trials) | See comment | 0 participants hospitalized or re‐hospitalized |
| *The basis for the assumed risk (e.g. the median control group risk across trials) is provided in the 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). | ||||||
| GRADE Working Group grades of evidence | ||||||
| 1 Downgraded by 1 level for lacking of blinding and concealment | ||||||
| Trial | Number | Sex (male/female) | Average age/ (Age range) | Duration of heart failure | Original disease | Heart function |
| 40 | 35/5 | 58/(39 to 80) | CHD (n = 40) | NYHA II (n = 37), NYHA III (n = 3) | ||
| 60 | 33/27 | 41 | CHD (n = 21), RHD (n = 12), HBP (n = 15), DCM (n = 12) | NYHA II (n = 8), NYHA III (n = 34), NYHA IV (n = 18) | ||
| 120 | 64/56 | /(38 to 75) | 3 to 12 years | CHD (n = 52), HBP (n = 42), DCM (n = 11), RHD (n = 15) | NYHA II n = 31), NYHA III n = 52), NYHA IV n = 37) | |
| Mao 2003a, Part A | 40 | 17/23 | 59/(40 to 70) | 6.70 ± 6.50 years | CHD (n = 33), RHD (n = 2), HBP (n = 2), PHD (n = 2) | NYHA II (n = 13), NYHA III (n = 27) |
| Mao 2003a, Part B | 40 | NYHA II (n = 14), NYHA III (n = 26) | ||||
| 100 | 59/41 | /(34 to 78) | 0.5 to 8 years | DCM (n = 100) | NYHA II (n = 20), NYHA III (n = 71), NYHA IV (n = 9) | |
| 40 | 14/26 | 61(41 to 70) | 0.5 to 30 years | CHD (n = 35), RHD (n = 3), PHD (n = 2) | NYHA II (n = 18), NYHA III (n = 22) | |
| 60 | 34/26 | /(50 to 75) | 2 to 14 years | CHD (n = 39), RHD (n = 7), PHD (n = 2), DCM (n = 4), HBP (n = 6), congenital heart disease (n = 2) | NYHA III (n = 35), NYHA IV (n = 25) | |
| 74 | 43/31 | /(60 to 85) | CHD (n = 41), HBP (n = 29), DCM (n = 4) | NYHA II (n = 7), NYHA III (n = 45), NYHA IV (n = 22) | ||
| 26 | 23/3 | 43.8/(23 to 60) | DCM (n = 26) | NYHA II (n = 23), NYHA III (n = 3) | ||
| 66 | 32/34 | /(44 to 85) | CHD (n = 38), RHD (n = 5), HBP (n = 11), PHD (n = 12) | NYHA II (n = 14), NYHA III (n =39), NYHA IV (n = 13) | ||
| 50 | 33/17 | /(60 to 75) | NYHA III (n = 33), NYHA IV (n = 17) | |||
| 62 | 28/34 | 62.4/(48 to 70) | 4 to 11 year/6.9 ± 2.9 years | HBP (n = 26), CHD (n = 29), diabetes mellitus (n = 32) | NYHA II (n = 22), NYHA III (n = 30), NYHA IV (n = 10) | |
| 80 | 33/47 | 70.74/72.56 | 8,07± 6.30 years/8.92 ± 4.68 years | Old myocardial infarction (n = 25), hypertensive heart disease (n = 56), diabetes mellitus (n = 33), hyperlipidaemia (n = 26) | NYHA III (n = 61), NYHA IV (n = 19) | |
| CHD = coronary heart disease; RHD = rheumatic heart disease; PHD = pulmonary heart disease; DCM = dilated cardiomyopathy; HBP =hypertensive heart disease; HCM = hypertrophic cardiomyopathy; NYHA = New York Heart Association. | ||||||
| Trial | Agent | Dose | Form | Duration of treatment (days) |
| Shanghai | 10 mL po bid | oral liquid | 20 | |
| Huaxi | 20 to 40 mL + GS /iv gtt qd | injection | 14 | |
| Shanxi | 20 mL/40 mL/60 mL 5% GS 250 mL/iv gtt qd | injection | 15 | |
| Mao 2003a, Part A | Yibin | 20, 40, 60 mL + 5% 200 mL polarized solution /iv gtt qd | injection | 14 |
| Mao 2003a, Part B | Yibin | 20, 40, 60 mL + 5% 200 mL polarized solution /iv gtt qd | injection | 14 |
| Huaxi | 60 mL + 5% GS 250ml/iv gtt qd | injection | 14 | |
| Huaxi | 20 mL/40 mL/60 mL + 5% GS 100 mL + 0.25 g KCL + R‐I 1u | injection | 7 | |
| Shanghai | 10 mL po bid | oral liquid | 20 | |
| not stated | 40 mL + 5% GS 250 mL/iv gtt qd | injection | 15 | |
| Jiangsu | 60 mL + GS 250 mL/iv gtt qd | injection | 15 | |
| not stated | 100 to 180 mL + 5% GS 150 mL/iv gtt qd | injection | 10 | |
| not stated | 100 mL po bid | powder | 14 | |
| Suzhong | 40 mL + 5% GS 100 mL/iv gtt qd | injection | 14 | |
| Shanxi | 40 mL + 5% GS 250 mL or 0.9% NS 250ml/iv gtt qd | injection | 7 | |
| po = by mouth; iv gtt = intravenously guttae; GS = glucose and saline. | ||||
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Lack of improvement in heart failure (NYHA class improved < I class or worsening of heart failure) Show forest plot | 10 | 672 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.37 [0.26, 0.51] |
| 2 Exercise test (after treatment) Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 3 Change in quality of life Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 4 Change in ejection fraction (%) Show forest plot | 4 | 354 | Mean Difference (IV, Random, 95% CI) | 9.02 [5.95, 12.08] |
| 5 Change in cardiac output (L/min) Show forest plot | 4 | 360 | Mean Difference (IV, Random, 95% CI) | 0.94 [0.26, 1.63] |
| 6 Change in stroke volume (mL) Show forest plot | 3 | 240 | Mean Difference (IV, Random, 95% CI) | 8.53 [0.08, 16.98] |
| 7 Change in cardiac index L/min*m² Show forest plot | 2 | 160 | Mean Difference (IV, Random, 95% CI) | 0.79 [0.59, 0.99] |
| 8 Change in left ventricular end‐diastolic volume (mL) Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 9 Change in left ventricular end‐systolic volume (mL) Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Lack of improvement in heart failure (NYHA class improved < I class or worsening of heart failure) Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 2 Change in exercise test (mins) Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 3 Change in ejection fraction (%) Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 4 Change in cardiac output (L/min) Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 5 Change in stroke volume (mL) Show forest plot | 1 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 6 Change in stroke volume index Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 7 Change in cardiac index L/min*m² Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 8 Change in Heather Index Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 9 Change in ECG: Q‐Z Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 10 Change in ECG: Adz/dtMax Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 11 Change in ECG: Q‐B/B‐X Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
