Chinese herbal medicine injections (CHMIs) for chronic pulmonary heart disease: protocol for a Bayesian network meta-analysis

Yuping Lei, Meili Wang, Guiqiang Sun, Yong Liu, Yapei Yang, Hao Dong2020 · DOI

TL;DR

This protocol outlines a rigorous Bayesian network meta-analysis designed to rank the efficacy and safety of various Chinese herbal medicine injections (CHMIs) for treating chronic pulmonary heart disease (CPHD). By synthesizing data from both direct and indirect clinical trial comparisons, the researchers aim to provide a clear, evidence-based hierarchy for these treatments to guide clinical decision-making.

What Problem Does This Paper Address?

Chronic pulmonary heart disease (CPHD) is a complex, often refractory condition where the heart's right ventricle fails due to long-term respiratory issues. In China, Chinese herbal medicine injections (CHMIs) are frequently integrated into treatment regimens. However, clinicians face a "paradox of choice": there are numerous CHMI formulations available, but very few head-to-head clinical trials comparing them directly.

Most existing studies compare a specific CHMI against standard Western medicine or a placebo. This leaves a significant gap in knowledge regarding which specific injection is the most effective or the safest. This paper provides the blueprint for a study that uses advanced statistics to bridge these gaps and rank the treatments.

Methods

The authors propose a Bayesian network meta-analysis (NMA). While a standard meta-analysis pools results from studies comparing the same two things, an NMA allows for indirect comparisons. For example, if Drug A is compared to Placebo in one study, and Drug B is compared to Placebo in another, NMA can statistically estimate how Drug A performs against Drug B.

Key technical components include:

  • Statistical Framework: Using WinBUGS and the Markov chain Monte Carlo (MCMC) method to synthesize evidence.
  • Ranking Metrics: The researchers will use SUCRA (Surface Under the Cumulative Ranking curve) values. SUCRA provides a numeric probability (from 0 to 1) that a specific treatment is the best among all those being compared.
  • Quality Control: Studies will be screened using the Cochrane "Risk of bias" tool and the GRADE (Grading of Recommendations, Assessment, Development, and Evaluation) framework to ensure the final recommendations are based on high-quality evidence.

Datasets

As this is a protocol for a systematic review, the "dataset" consists of all relevant clinical literature published from database inception through December 2020. The authors will search:

  • International Databases: PubMed, Google Scholar, Medline, Cochrane Library, and Web of Science.
  • Chinese Databases: China National Knowledge Infrastructure (CNKI), Wanfang Database, China Scientific Journal Database, and Chinese Biomedical Literature Database.
  • Inclusion Criteria: Randomized controlled trials (RCTs), quasi-RCTs, and high-quality prospective cohort studies involving patients diagnosed with CPHD.

Key Insights and Ideas

  • The Power of Indirect Evidence: The core "aha" moment is the application of Bayesian logic to medical literature. By treating the network of clinical trials as a connected graph, the researchers can extract insights that no single trial could provide on its own.
  • Holistic Outcome Measures: Rather than just looking at "effective vs. non-effective" rates, the study will analyze a wide array of physiological markers, including:
  • Hemodynamics: Mean pulmonary artery pressure (mPAP) and Left ventricular ejection fraction (LVEF).
  • Biomarkers: B-type natriuretic peptide (BNP).
  • Hemorheology: Blood viscosity and fibrinogen levels.
  • Blood Gas Analysis: Oxygen (PaO2) and carbon dioxide (PaCO2) levels.
  • Addressing Heterogeneity: The protocol includes specific plans for subgroup analysis (age, region, treatment duration) to ensure that the results aren't skewed by differences in how studies were conducted.

Key Results

Note: As this is a protocol, there are no quantitative results yet.

The expected output of the completed study will be:

  1. A comprehensive ranking of CHMIs based on their total effective rate (TER).
  2. A safety profile for each injection based on reported adverse events.
  3. A "closed-loop" inconsistency check to ensure that direct and indirect evidence align logically.

Limitations and Future Work

  • Language Bias: The search is limited to English and Chinese publications. While these are the primary languages for this field, relevant data in other languages might be missed.
  • Data Quality Dependency: The final meta-analysis is only as strong as the underlying trials. If the original RCTs have high risk of bias or small sample sizes, the NMA's conclusions will be weakened.
  • Future Directions: Once the analysis is complete, the authors intend to publish the results in a peer-reviewed journal to provide a definitive guide for clinicians treating CPHD with integrated medicine.

Summary updated

Abstract

Review question / Objective: Which Chinese herbal medicine injection (CHMI) is more effective for the treatment of patients with chronic pulmonary heart disease (CPHD)?Rationale: Chinese herbal medicine injections (CHMIs) are frequently used for various refractory diseases including chronic pulmonary heart disease (CPHD).However, due to the diversity of CHMIs treatments, its relative effectiveness and safety remain unclear.In our study, Bayesian network meta-analysis will be INPLASY 1 International Platform of Registered Systematic Review and Meta-analysis Protocols INPLASY PROTOCOL Chinese herbal medicine injections (CHMIs) for chronic pulmonary heart disease: protocol for a Bayesian network meta-analysis
BibTeX
@misc{lei2020chinese,
  author = {Yuping Lei and Meili Wang and Guiqiang Sun and Yong Liu and Yapei Yang and Hao Dong},
  title = {{Chinese herbal medicine injections (CHMIs) for chronic pulmonary heart disease: protocol for a Bayesian network meta-analysis}},
  year = {2020},
  doi = {10.37766/inplasy2020.12.0004},
}
Figure