Adjuvant Application of Shenmai Injection for Sepsis: A Systematic Review and Meta-Analysis

Objective To compare the efficacy and safety of Shenmai injection plus regular treatment versus regular treatment alone in the treatment of sepsis. Methods Review Manager 5.3 was used. Databases including PubMed, Cochrane Library, Wan-FANG, VIP, CNKI, and CBM were searched for the relative control and randomized studies till June 2022. Primary outcomes were 28-day mortality; case fatality rate; APACHE II score; and levels of procalcitonin, tumor necrosis factor-α, interleukin-6, and C-reaction protein. Secondary outcomes were measurements related to inflammatory reactions, coagulation function, immune function, and organ function. Results In total 21 studies with 1469 patients were included. Shenmai injection plus regular treatment of sepsis significantly decreased 28-day mortality rate (odds ratio: 0.36 (0.20, 0.63); p=0.0004) and case fatality rate (odds ratio: 0.32 (0.19, 0.54), p < 0.0001) and significantly further reduced APACHE II score (standard mean deviation: −1.14 (−1.30, 0.99); p < 0.00001). Procalcitonin reduction was similar in the two groups (standard mean deviation: −0.59 (−1.64, 0.46); p=0.27). Tumor necrosis factor-α (standard mean deviation: −1.96 (−2.79, −1.13); p < 0.00001), interleukin-6 (standard mean deviation: −1.52 (−2.13, −0.91); p < 0.00001), and C-reaction protein (standard mean deviation: −2.37 (−4.26, −0.49); p=0.01) levels were significantly further decreased in the Shenmai injection group. Most of the measurements related to inflammatory reaction, coagulation function, immune function, and organ function were significantly regulated in the Shenmai injection group. Furthermore, there was no adverse event, and two study groups had similar adverse event rates. Conclusion Shenmai injection as adjuvant therapy shall be effective and safe in the treatment of sepsis. However, further investigation is still warranted especially regarding safety. Adjuvant application of Shenmai injection in sepsis is worth further investigation.


Introduction
Sepsis refers to life-threatening organ dysfunction induced by the host response to infection disorders [1]. It is one serious complication among intensive care unit (ICU) patients affected by infection, shock, burn, etc. Sepsis accounted for 10%-40% of ICU cases [2], and the global incidence of sepsis is increasing 2%-9.5% each year [3]. is trend was mainly attributed to old age, immune suppression, and multiple resistant bacteria [3]. Morality rates of sepsis patients ranged from 20% to 50% [3][4][5][6]. In Western countries, it is a leading cause of ICU death only secondary to coronary heart disease [2]. Current treatment options for sepsis include early fluid resuscitation, anti-infection, vasoactive agent usage, protective ventilation strategy, hemopurification, etc. However, the poor prognosis of sepsis patients makes sepsis treatment still a challenge. It is necessary to develop a novel treatment strategy or option. Traditional Chinese medicine (TCM) has accumulated certain understanding and treatment experience for sepsis. erefore, TCM treatment methods are receiving more and more attention from domestic and international researchers.
Shenmai injection was a Chinese patent drug widely used as adjuvant therapy for sepsis in Chinese hospitals in recent decades. It was prepared with ginseng and Ophiopogon. e formula of Shenmai injection was modified from Sheng Mai San recorded in Qian Jing Yao Fang of the Tang Dynasty [7]. In TCM, this formula has the effect of enhancing qi, nourishing fluid, replenishing vital energy, promoting blood circulation, dispersing phlegm, etc. e major components of Shenmai injection include ginsenoside saponins, ophiopogon saponin, ophiopogon flavonoids, ginseng polysaccharides, and ophiopogon polysaccharides. Previous studies demonstrated that ginsenoside had an immune regulation activity that could improve the suppression of cyto-immune function in burn sepsis [8]. Ginsenoside has the obvious anti-inflammation effect that it could downregulate the levels of tumor necrosis factor (TNF-α), interleukin-1 (IL-1), IL-6, IL-8, IL-10, and nitric oxide (NO) [9]. It was also revealed that Shenmai injection could inhibit over-releasing of inflammatory mediators, reduce inflammatory reaction, clear free radical, regulate immune function, enhance cytometabolism, prevent shock and cardiac arrhythmia, have cardiotonic action, improve hemorheology, and prevent aging [10].
ere has been a number of studies investigating Shenmai injection in the treatment of sepsis, most of which were published in Chinese journals. However, there is no related meta-analysis to provide pooled evidence regarding this issue. erefore, we performed the present metaanalysis in order to evaluate the efficacy and safety of Shenmai injection as adjuvant therapy in the treatment of sepsis, which might provide an access for the domestic and international researchers to know about the application of Shenmai injection in sepsis.

Inclusion and Exclusion Criteria.
e inclusion criteria were as follows: (1) types of studies: randomized controlled studies; (2) types of participants: patients with sepsis following the diagnosis criteria of sepsis defined by the International Sepsis Definition Conference of 2001 [11]; (3) intervention: the patients in the control group only received regular Western treatment for sepsis, and the patients in the experimental group received regular Western treatment combined with Shenmai injection; regular treatment of sepsis may include fluid resuscitation, anti-infection, serum sugar control, maintenance of homeostatic equilibrium, mechanical ventilation, and comprehensive therapy such as nutrition support and administration of glucocorticoids. Treatment length is at least 7 days; (4) outcome and measurements: primary outcome included APACHE II score, 28-day mortality rate, case fatality rate, procalcitonin (PCT) levels, TNF-α levels, IL-6 levels, and C-reaction protein (CRP) levels. e included study had at least one primary outcome.
Exclusion criteria were as follows: (1) retrospective or observational studies and animal studies were excluded; (2) for the experimental group, only Shenmai injection was added to the regular Western treatment, regardless of administration methods. No other treatment such as electroacupuncture, acupoint application, or other Western drugs or TCM was used; (3) if the results of one study were divided into multiple articles, only one article with the relatively complete result was included and the rest articles were excluded.

Article Retrieval.
e studies investigating Shenmai injection as adjuvant therapy in the treatment of sepsis till June 2022 were retrieved using English and Chinese databases.
e English databases included PubMed and Cochrane Library.
e Chinese databases included Wan-Fang, VIP, CNKI, and CBM. e retrieval terms were "sepsis", "Shenmai injection". e Chinese retrieval terms were corresponding to Chinese for the retrieval terms above. e references of important articles were also manually retrieved.

Literature Screening, Data Extraction, and Quality
Assessment. Literature screening and data extraction were performed by two independent reviewers respectively. During literature screening, duplication, title, abstract, and main text were examined. e situation that data of one study were separately published in multiple articles was also examined. Only the article containing the most complete data was included for further examination.
e results of literature screening and data extraction by the two reviewers were compared, discussed, and determined. When necessary, a third experienced reviewer was invited. e extracted data included author name, publishing year, sample size, patient age, intervention in experimental group and control group, treatment length, and multiple outcome measurements.
Cochrane risk of bias tool was used for quality assessment of the included studies. e assessment items were random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, and selective reporting. Each item was rated as low, unclear, and high risk.

Statistical Analysis.
e statistical analysis was performed using the statistical software Review Manager 5.3. Odds ratios (ORs) for categorical variables were calculated; weighted mean difference (WMD) for numerical data was calculated; all effect sizes were expressed with a 95% confidence interval (95% CI). If different units were used for continuous variables in included studies, standard mean difference (SMD) was used for comparison. e heterogeneity test was performed using a standard chi-squared (I 2 Q) test. If p > 0.1, I 2 < 50%, the fixed effect model was used; if p > 0.1, I 2 > 50%, the random effect model was used. When heterogeneity was significant, a sensitivity test was performed. In the sensitivity test, studies were ruled out one by one. If the heterogeneity result was not changed, the result was stable. If the result was changed, the study identified could greatly impact the result and it was closely studied to further determine the source of heterogeneity. For the analysis including no less than 10 studies, a funnel plot was used to estimate the publication bias.

Literature Screening Result and Characteristics of the Included Studies.
In total, 38 articles were retrieved from the databases mentioned above, 30 studies were retrieved from CNKI, 26 from CBM, 28 from VIP, 22 from Wan-Fang, and 0 from PubMed and Cochrane library. After examination of duplication and situation of multiple articles from one study, 8 articles were excluded. After examination of the title and abstract, 5 articles were excluded. After the main text examination, 4 articles were excluded, and 21 studies were finally included in the present meta-analysis. e study selection process is shown as a flowchart in Figure 1.
e characteristics of included studies are summarized in Table 1  . In total, 21 studies with 1469 participants were included in this study. All studies were randomized controlled studies, and the baseline characteristics of patients were consistent without statistical difference between the experimental group and the control group. e time span was 2010-2021. e dosage of Shenmai injection ranged from 40 mg/d to 240 mg/d. e length of treatment ranged from 7 to 10 days.

Quality Assessment Result.
Among the included studies, 6 studies reporting random number table methods [12,14,18,20,21,26,29] and 3 studies reporting simple random sampling methods [15,16,32] were rated as low risk regarding selection bias of sequence generation. e other studies only mentioned randomization and were rated as unclear risk. None of the studies reported follow-up as they only evaluated short-term efficacy. Five studies had provided description or data of adverse events [25][26][27][29][30][31]. e blindness of allocation, participants, and assessment was not reported by all studies, so they were rated as unclear. All studies provided complete data regarding the variables mentioned in the study methods, so they were all rated as low risk. e quality assessment results are summarized in Figure 2 and Figure 3. e quality of included studies was not high.

Rate of 28-Day Morality.
In total, 6 studies were included in this study ( Figure 4). e heterogeneity test indicated that the difference between included studies was not significant (p � 0.81, I 2 � 0%) and a fixed effect model was used. e analysis result indicated that the 28-day mortality rate of the Shenmai injection group was significantly lower than that of the control group (OR: 0.36 (0.20, 0.63); p � 0.0004).

Case Fatality Rate.
In total, 8 studies with 586 participants were included in the present analysis ( Figure 5). e heterogeneity test result indicated that there was no significant heterogeneity (p � 0.97, I 2 � 0%) and a fixed effect model was used. e analysis result indicated that the case fatality rate of the Shenmai injection group was significantly lower than that of the control group (OR: 0.32 (0.19, 0.54); p < 0.0001).

APACHE II Score Reduction from Baseline.
In total, 11 studies were included in the analysis of APACHE II score reduction, which included 810 participants. According to the heterogeneity test result (p < 0.00001, I 2 � 79%), there was a significant difference between the included studies (Figure 6). A random effect model was used. e analysis result showed that the APACHE II score reduction of the Shenmai injection group was significantly higher than that of the control group (SMD: −1.14 (−1.30, −0.99); p < 0.00001). e sensitivity test was performed, and the significant heterogeneity was stable upon removing each study.

PCT Level Reduction.
In total, 5 studies including 428 participants were included in the present analysis ( Figure 7). e heterogeneity test result indicated that there was a significant difference between the included studies (p < 0.00001, I 2 � 96%). A random effect model was used. e PCT level of the Shenmai injection group was slightly further decreased compared with that of the control group but showed no significant difference (SMD: −0.59 (−1.64, 0.46); p � 0.27). e sensitivity test indicated that the heterogeneity was stable.

TNF-α.
In total, 8 studies with 594 participants were included in the present analysis ( Figure 8). e heterogeneity test result demonstrated that the difference between the included studies was significant (p < 0.00001, I 2 � 94%). Random effect model was used. e analysis result indicated that the level of TNF-α in the Shenmai injection group was significantly further decreased compared with the control group (SMD: −1.96 (−2.79, −1.13); p < 0.00001). e sensitivity test demonstrated that the heterogeneity was stable.
3.3.6. IL-6. In total, 7 studies including 544 participants were included in the present analysis ( Figure 9). e heterogeneity test result indicated that the included studies were significantly different (p < 0.00001, I 2 � 90%). Random Evidence-Based Complementary and Alternative Medicine effect model was used. e analysis result indicated that the IL-6 level in the Shenmai injection group was significantly further decreased compared with the control group (SMD: −1.52 (−2.13, −0.91); p < 0.00001). e sensitivity test suggested that the heterogeneity was stable.

CRP.
In total, 10 studies including 786 participants were included in the present analysis ( Figure 10). e heterogeneity test indicated the included studies were significantly different (p < 0.00001, I 2 � 99%). Random effect model was used. Analysis result indicated that CRP level in the Shenmai injection group was significantly further decreased compared with the control group (SMD: −2.37 (−4.26, −0.49); p � 0.01). e sensitivity test suggested that the heterogeneity result was stable.

Publication Bias.
A funnel plot of the APACHE II score was adopted to estimate the publication bias (Figure 11). e asymmetry was not significant, indicating the publication bias is not obvious and its impact on analysis result can be ignored.

Measurements Related to Immune Function, Coagulation Function, Inflammation Reaction, and Organ
Function. Other measurements were also compared and the results are summarized in Table 2. e measurements associated with coagulation function included PT, TT, APTT, FIB, PLT, and D-D; the measurements associated with immune function included CD3, CD4, CD8, CD4/ CD8, and NK; the measurements associated with inflammation factor included IL-1 and IL-8, besides TNF-α and IL-6 above; the measurements associated with organ function included ALT, AST, BUN, Cr, LDH, CK, CK-MB, BNP, and TnI; other measurements included WBC and Hb.
e analysis results indicated that the addition of Shenmai injection could significantly further decrease the levels of PT, TT, APTT, D-D, IL-8, LDH, CK, CK-MB, BNP, and TnI compared with regular treatment alone; the addition Records excluded (n = 5) with reasons: children sepsis patients were included in 1 study; not only Shenmai injection but also other drugs were added in experimental group by 2 studies; patients with septic shock were included in 2 studies Full-text articles assessed for eligibility (n = 25) Full-text articles excluded, with reasons (n = 4): treatment length were 3 days less than 7 days in 3 studies; none of the required measurements was included in 1 study Studies included in qualitative synthesis (n = 21) Studies included in quantitative synthesis (meta-analysis) (n = 21)

Discussion
Shenmai injection is widely used as adjuvant therapy in the treatment of sepsis in China. ere are a number of studies investigating Shenmai injection. erefore, we performed the present meta-analysis to evaluate the efficacy and safety of Shenmai injection as adjuvant therapy in the treatment of sepsis.

Evidence-Based Complementary and Alternative Medicine
Our analysis demonstrated that the addition of Shenmai injection to regular treatment of sepsis significantly decreased the 28-day mortality rate and case fatality rate and significantly further reduced the APACHE II score. Furthermore, there was no adverse event, and two study groups had similar adverse event rates. erefore, Shenmai injection as adjuvant therapy shall be effective and safe in the treatment of sepsis.
In the early phase of sepsis, the immune system was over-activated, releasing of inflammatory mediators was out of control, and systemic inflammatory response syndrome was induced. TNF-α, IL-6, and CRP are the most important pro-inflammatory mediators or inflammatory markers in the process [33]. If the microorganism was not cleared efficiently, early inflammatory reaction in sepsis could activate coagulation system, initiate coagulation cascade reaction, inhibit fibrinolysis pathway, induce microcirculation disturbance, interfere blood and oxygen supply to tissues, and lead to organ dysfunction [34]. In summary, the pathogenesis of sepsis is an interaction result of immune reaction, inflammatory reaction, and coagulation reaction; organ dysfunction (MODS) and disseminated intravascular coagulation (DIC) are common complications developed in sepsis progression and are major causes of death by sepsis [35].
According to our study result, a combination of Shenmai injection further decreased PCT, TNF-α, IL-6, CRP, and IL-8 levels, which indicated Shenmai injection suppressed proinflammatory mediator releasing and decreased systematic inflammatory reactions. e levels of PT, TT, APTT, and D-D were further decreased and the levels of FIB and PLT were further increased in the Shenmai injection group, which indicated that Shenmai injection improved coagulation function. e levels of CD3, CD4, and CD4/CD8 were further increased in the Shenmai injection group, which indicated that Shenmai injection improved immune function. e levels of LDH, CK, CK-MB, BNP, and TnI were further decreased and the elevation trend of ALT, AST, BUN, and Cr was reversed in the Shenmai injection group, which indicated that Shenmai injection had multiple organ protection activity. e regulation activity of Shenmai injection on immune function [36], inflammation reaction [37][38][39], and hemorheology [40][41][42] and further protection on multiple organs [37,42] by Shenmai injection had been elucidated by a large number of animal or clinical studies. erefore, we may conclude that the beneficial impact of Shenmai injection on inflammatory reaction, immune function, coagulation function, and organ function might contribute to better treatment outcome in the Shenmai injection group. e result of adverse events indicated Shenmai injection was safe, however, which was concluded based on quite incomplete safety data. Previous studies had monitored adverse events related to Shenmai injection. According to the study performed by Wen et al. among 1216 cases, the rate of adverse events by Shenmai injection was 0.82% [43]. e study performed by Wang et al. with 421 medical cases using Shenmai injection reported an adverse event rate of 3.6% [44]. e study by Wen et al. was performed in a TCM hospital where the usage of Shenmai injection would be more standardized.
is might contribute to the lower adverse event rate. However, in our study, the adverse events were reported by only a small portion of included studies and the safety information was incomplete, so interpretation of safety of Shenmai injection should be cautious. e current treatment outcome for sepsis is not very satisfying. Accompanying more and more understanding of complicated mechanism of sepsis, researchers are seeking for novel treatment options, such as immune regulation therapy. Shenmai injection has a nature origin so that it could solve the problem of overuse of antibiotics in sepsis treatment. It has multiple effects including immune regulation activity. e treatment guidance strategy of Shenmai in sepsis was targeting microorganisms, removing toxicity, and promoting vital energy at the same time, which is similar to the current investigation trend and could be a good complementary. erefore, Shenmai injection is a good potential candidate for sepsis treatment and is worth further investigation. However, the complicated components and agent interaction of TCM shall be Evidence-Based Complementary and Alternative Medicine addressed, and suitable investigation methods need to be developed or applied.
is is the first meta-analysis evaluating the efficacy and safety of Shenmai injection as adjuvant therapy in sepsis treatment.
is study not only evaluated the effect of Shenmai injection on major treatment outcomes, but also evaluated the effect of Shenmai injection on measurements related to coagulation function, immune function, and organ functions, thus it provided a relatively comprehensive assessment of effect of Shenmai on sepsis. However, there are still some limitations in this study. (1) Some wellestablished indicators for systematic assessment of efficacy in sepsis treatment were not included such as ICU stay/hospital stay, mechanical ventilation rates, and rate of multiple organ               adverse events into evaluation. erefore, future relative clinical studies shall take these into consideration: inclusion of more well-established indicators; detailed information on study procedure; and a more complete and standard safety report.
Currently, the TCMs are mainly used within China and are known by very few physicians or researchers out of China. Moreover, some international medical journals' acceptance of the study of TCM is low due to the complicated components.
erefore, the researchers would intend to  publish the studies on Shenmai injection in Chinese Medical Journals. us, the study quality and article writing followed a relatively lower standard. However, TCM is receiving more and more attention from international researchers, and this trend in the publication of TCM in Chinese Medical Journals is changing. When it comes to the credibility of the results, many studies are usually of lower quality, but they still can elucidate the efficacy and safety of Shenmai injection to a large extent. Moreover, Shenmai injection is widely used in Chinese hospitals, and there are several publications on Shenmai injection in sepsis in recent years which demonstrated that the efficacy and safety of Shenmai injection are recognized by Chinese physicians. erefore, I think the relatively low quality of studies included in present analysis may lead to certain bias, but the analysis results and conclusion are still reliable.

Conclusion
Shenmai injection as adjuvant therapy shall be effective and safe in the treatment of sepsis. However, further investigation is still warranted, especially regarding the safety profile. Adjuvant application of Shenmai injection in sepsis is worth further investigation.

Data Availability
All data generated or analyzed during this study are included in this published article.

Consent
Written consent for publication has been obtained from all the authors.