Chinese Medicinal Herbs for Childhood Pneumonia: A Systematic Review of Effectiveness and Safety

Objective. To assess the efficacy and safety of Chinese medicinal herbs for Childhood Pneumonia. Methods. We included randomized controlled trials (RCTs). The searched electronic databases included PubMed, the Cochrane Central Register of Controlled Trials, EMBASE, CBM, CNKI, and VIP. All studies included were assessed for quality and risk bias. Review Manager 5.1.6 software was used for data analyses, and the GRADEprofiler software was applied to classify the systematic review results. Results. Fourteen studies were identified (n = 1.824). Chinese herbs may increase total effective rate (risk ratio (RR) 1.18; 95% confidence interval (CI), 1.11–1.26) and improve cough (total mean difference (MD), −2.18; 95% CI, (−2.66)–(−1.71)), fever (total MD, −1.85; 95% CI, (−2.29)–(−1.40)), rales (total MD, −1.53; 95% CI, (−1.84)–(−1.23)), and chest films (total MD, −3.10; 95% CI, (−4.11)–(−2.08)) in Childhood Pneumonia. Chinese herbs may shorten the length of hospital stay (total MD, −3.00; 95% CI, (−3.52)–(−2.48)), but no significant difference for adverse effects (RR, 0.39; 95% CI, 0.09–1.72) was identified. Conclusion. Chinese herbs may increase total effective rate and improve symptoms and signs. However, large, properly randomized, placebo-controlled, double-blind studies are required.


Introduction
Childhood Pneumonia is an acute virus, bacterial, or fungal respiratory infection that affects the lungs [1]. The symptoms and signs of Childhood Pneumonia include cough, fever, rapid or difficult breathing, loss of appetite, and lower chest wall indrawing [2]. Auscultation reveals rales. However, severe Childhood Pneumonia is defined as cough or difficult breathing combined with lower chest wall indrawing [3]. Childhood Pneumonia has been identified as a mixed viralbacterial infection in 23%-33% of cases [4]. Respiratory syncytial virus is the most common viral cause of children pneumonia, and Streptococcus pneumoniae is the most common cause of bacterial pneumonia in children [5]. Each year, pneumonia kills an estimated 1.4 million children <5 years of age, accounting for 18% of all deaths worldwide [6].
Pneumonia is the most common reason for hospitalization in children <2 years of age [7]. The cost of antibiotic treatment for all children with pneumonia in 42 of the poorest countries is estimated to be about US $600 million per year. The estimated incidence rates are 0.29 and 0.05 episodes per childyear in low-income and high-income countries, respectively. Approximately 156 million new episodes occur each year, the majority in India (43 million), China (21 million), Pakistan (10 million), Bangladesh, Indonesia, and Nigeria (6 million each) [8].
Administering appropriate antibiotics at the early stage of pneumonia improves outcomes, particularly when the causative agent is bacterial [9]. However, antibiotic treatment of pneumonia in children remains mostly empirical because determining the etiologic pathogen is difficult in this age group [10]. According to the results of a questionnaire, 2.3.1. Study Selection. Two review authors independently browsed the titles and abstracts of all articles identified by the literature search. The same review authors independently estimated whether the trials met the inclusion criteria. Disagreements were resolved by discussion or consultation with a third author. We assessed abstracts from the initial search independently to identify studies that met the inclusion criteria. We telephone-interviewed authors of Chinese language articles and emailed the original authors of English articles to identify the randomization procedure and other methodological questions to ensure that the included studies were RCTs. If the required information was not available or if the required information did not meet the inclusion criteria, the article was excluded.

Data Extraction and Management.
We extracted data including methodological details and data from publications using a data extraction form. We extracted data on study characteristics, including methods, participants, interventions, and outcomes. There were no disagreements among the authors. Oral administration Guo, 1999 [17] Modified Ma Xing Shi Gan Tang  We extracted the formulation contents of the included studies, and the names of the herbs are provided in three languages (e.g., Chinese, Latin, and English) in Table 1.

Assessment of Risk of Bias in Included Studies.
The following items were independently assessed by our authors using the risk of bias assessment tool. (1) Was there adequate sequence generation (selection bias)? (2) Was allocation adequately concealed (selection bias)? (3) Was knowledge of the allocated interventions adequately prevented during the study (e.g., participants and personnel, outcome assessors) (detection bias)? (4) Were incomplete outcome data adequately addressed (attrition bias)? (5) Are reports of the study free of suggesting selective outcome reporting (reporting bias)? (6) Was the study apparently free of other problems that could put it at risk for bias?

Measures of Treatment
Effect. Data analyses were performed using the Cochrane Collaboration's RevMan software, version 5.1.6. Results are expressed as risk ratios (RR) and 95% confidence intervals (CIs) for dichotomous outcomes (e.g., mortality, effective rate, adverse effects, and relapse rate) and as mean differences (MD) with 95% CIs for continuous outcomes (such as time to clinical recovery and length of hospital stay).

Assessment of Heterogeneity.
Heterogeneity was analyzed using the chi-square test on − 1 degrees of freedom, and an alpha of 0.05 was used for statistical significance with the 2 test. 2 values of 25, 50, and 75% corresponded to low, medium, and high levels of heterogeneity, respectively.

Data Synthesis.
We used fixed-effects and randomeffects models for the pooled data analysis. We performed a pooled analysis for the 14 studies.

Subgroup Analyses.
Subgroup analyses were performed according to formula type and Chinese medicinal herb preparation type, using the same comparators (e.g., same types of antibiotics).

Sensitivity
Analyses. Sensitivity analyses were conducted by excluding low-quality studies (based on descriptions of randomization, allocation concealment, blinded assessment of outcomes, and description/analyses of withdrawals and dropouts) and a comparison of the merger analysis results for the fixed-and random-effects models.

Search
Results. An initial search identified 2,502 potentially relevant articles. Of these, 15 were in the English database. A total of 891 articles were initially included after duplicate publications were removed and any obviously irrelevant were excluded; 800 articles were later excluded, because they did not meet the inclusion criteria. Of the 91 potentially eligible reports, 77 were excluded for further assessment because telephone interviews with the original authors revealed that they were not RCTs. Therefore, 14 studies (1,824 participants) were included in this paper. All 14 studies were published in Chinese (Figure 1).

Inclusion
Criteria. The diagnostic criteria for childhood pneumonia in all studies included fever >37.5 ∘ C, chest recession, increased respiratory rate, cough, rales, or difficulty breathing combined with fast breathing and a change on chest films.

Intervention.
Chinese medicinal herbs interventions were given as oral decoctions or intravenous infusions. The longest therapy duration was 3 weeks, and the shortest was 5 days. Follow-up duration was not mentioned by any of the authors. All 14 studies compared Chinese medicinal herbs plus basic therapy versus basic therapies.

Risk of Bias in Included Studies
. The methodological quality of each study's randomization sequence, allocation concealment, blinding, incomplete outcome data, selective outcome reporting, and potential threats are summarized in Figures 2 and 3.

Randomization and Allocation
Concealment. Ten studies [15-20, 23, 25-27] reported using a random-number table, and four [21,22,24,28] reported using a computer-generated random-number table. None of the trials used allocation concealment. Therefore, all studies had a high risk of selection bias.

3.3.2.
Blinding. Ten studies [15-17, 20-24, 26, 27] used single blinding (outcome assessment was blinded). We interviewed the original authors by telephone to determine blinding because the blinding methods were not described. Thus, these studies had a low risk of performance bias and low detection bias. The other studies [     Evidence-Based Complementary and Alternative Medicine

Flow of Participants and Intention-to-Treat.
None of the studies reported withdrawal, dropout, and/or loss during followup. The method of handling missing data regarding intention-to-treat or per-protocol analysis was not addressed.

Other Potential Sources of Bias.
None of the 14 studies [15][16][17][18][19][20][21][22][23][24][25][26][27][28] did not describe patient compliance. The appropriateness of the statistical analyses used was assessed, and the methods of all studies were considered appropriate. Although we conducted comprehensive searches and tried to avoid bias, we could not exclude potential publication bias because all 14 studies were published in China. Subgroup one study [20] compared the modified Ma Xing Shi Gan Tang formula plus ceftazidime versus ceftazidime and showed no difference in total effective rate (Figure 4; analysis 1.1.2 of analysis 1.1; RR, 1.11; 95% CI, 0.98-1.27).

Effects of Interventions
Subgroup one study [21] compared the San Ao Tang formula plus azithromycin versus azithromycin and showed no difference in total effective rate ( Figure 4; analysis 1.1.3 of analysis 1.1; RR, 0.99; 95% CI, 0.91-1.08).
Subgroup one study [22] compared the modified Zhi Sou San formula plus erythrocin versus erythrocin and showed a significant increase in total effective rate (Figure 4; analysis 1.1.4 of analysis 1.1; RR, 1.41; 95% CI, 1.24-1.61). ⊕○○○ Very low 1,5 Important * The basis for the assumed risk (e.g., the median control group risk across studies) is provided in the footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; RR: risk ratio. Grade: working group grades of evidence. High quality: further research is very unlikely to change our confidence in the estimate of effect. Handbook description: randomized controlled trial. Moderate quality: further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate. Cochrane Handbook description: relegation randomized controlled trial. Low quality: further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate. Cochrane Handbook description: two or more degradation factors of randomized controlled trials. Very low quality: we are very uncertain about the estimate. Cochrane Handbook description: more than three degradation factors of randomized controlled trials.
Reduce the evidence quality factors: methodology defect, included in the research results of the inconsistency, indirect evidence, inexactness, and publication bias. Increase the level of evidence factor: large effect quantity, confounding factors cannot change effect quantity, or the existing concentration-response relationship. 1 There is a high risk of selection bias, performance bias, and detection bias. 2 Some studies showed a significant difference, but some studies showed no significant difference. 3 Few studies included. 4 The protocol of the published studies could not be compared. 5 Only one study included. Lei, 2010 [24] He at el, 2011 [20] He, 2011 [19] Guo, 1999 [17] Duan and Feng, 2011 [27] Zhao    Test for overall effect: Test for overall effect: Test for overall effect: Test for overall effect:               Evidence-Based Complementary and Alternative Medicine Subgroup one study [23] compared a self-developed TCM prescription plus erythrocin versus erythrocin and showed no difference in total effective rate ( Figure 4; analysis 1.1.5 of analysis 1.1; RR, 1.13; 95% CI, 0.91-1.39).
Subgroup one study [22] compared the modified Zhi Sou San formula plus erythrocin versus erythrocin and showed a significant decrease in adverse effects ( Subgroup two studies [16,18] compared the modified Ma Xing Shi Gan Tang formula plus azithromycin versus azithromycin and showed a significant difference for cough ( Figure 6 Subgroup one study [21] compared the San Ao Tang formula plus azithromycin versus azithromycin and showed a significant difference in cough ( Figure 6; analysis 1

Other
Outcomes. Mortality, relapse rate, TCM outcomes (e.g., tongue coat and pulse condition), and economic index were not reported in any of the 14 studies.

GRADE Quality of Evidence
The "GRADEprofiler" of the Cochrane Collaboration Network was used to classify the systematic review results. The quality of evidence was low to very low (Table 3).
First, randomization was mentioned in all 14 studies. However, one study [25] described the randomization method in detail, whereas 13 did not [15][16][17][18][19][20][21][22][23][24][26][27][28]. We interviewed the authors by telephone and determined that a random number table or a computer-generated randomnumber table was used to generate the allocation sequence. Second, none of the studies mentioned a blinding method, but we interviewed the authors by telephone and found that 10 studies [15-17, 20-24, 26, 27] used single blinding (i.e., the outcome assessment was blinded), and four [18,19,25,28] did not. The lack of blinding participants, healthcare providers, and assessors can introduce performance and detection bias. Third, none of the studies addressed the incomplete outcome data, such as missing data due to attrition or exclusion. The inadequate handling of missing data can compromise statistical analyses. Fourth, the majority of experimental Chinese herbal medicine interventions were prepared by the investigators without detailed information describing their underlying rationale for the formulation, dosage, or the manufacturing process, and the quality control processes for their tested interventions are unknown. Thus, independent validation of these findings is necessary. This paper has several methodological limitations. First, although all data were collected from reports or from direct contact with the authors, many items on the "risk of bias" assessment tool could only be classified as "unclear. " Second, different Chinese herbal medicine interventions were grouped together for analysis in some cases. The results might have been compromised by the heterogeneity within each Chinese herbal medicine intervention and the study design. Third, the concept of TCM syndrome was not considered when analyzing the data, as all studies only considered Childhood Pneumonia, not TCM symptoms. Therefore, the actual therapeutic effect might not have been fully captured.
Based on these reasons, the TCM RCTs should be conducted in accordance with the Consolidated Standards of Reporting Trials for Traditional Chinese Medicine [29] detailed report.

Conclusions
Chinese herbal medicines may increase total effective rates, improve clinical symptoms and signs, and shorten the length of hospital stay for children with pneumonia. In a word, Chinese herbal medicines are effective for Childhood Pneumonia. However, there is insufficient evidence to confirm whether Chinese herbal medicines decrease adverse effects, mortality, and TCM outcomes (such as tongue coat and pulse condition). All results were supported by poor methodological quality studies. Thus, larger, multicenter, high methodological quality studies of Chinese herbal medicines for Childhood Pneumonia are needed. These studies should include patients with Childhood Pneumonia and interventions with Chinese herbal medicines. More data, particularly concerning adverse events, are necessary. Meanwhile, future studies should determine the most appropriate drug and 24 Evidence-Based Complementary and Alternative Medicine dosage for Childhood Pneumonia. So, further trials would help to clarify the validity of the findings of this paper and could determine more clearly the role of Chinese herbal medicines in Childhood Pneumonia in comparison with other therapies. Consequently, the purpose is to guide our application in clinical.