Published online Jan 6, 2021. doi: 10.12998/wjcc.v9.i1.102
Peer-review started: July 16, 2020
First decision: September 24, 2020
Revised: October 6, 2020
Accepted: November 12, 2020
Article in press: November 12, 2020
Published online: January 6, 2021
Nonalcoholic fatty liver disease (NAFLD) is a common chronic hepatic disease in clinical practice, affecting approximate one-third of the population in Asia. Conventional pharmacotherapies contain insulin sensitizers, antioxidants, cytoprotective drugs, and lipid lowering agents. However, the uncertain clinical efficacy and risk of adverse events still trouble clinical application. Hence, many patients introduce traditional Chinese medicine (TCM) formulas to the management of NAFLD. TCM formulas for NAFLD treatment have always been a popular research topic. Many relevant systematic reviews have been published in recent decades, but the results have not been consistent. Based on the different conclusions, it is not realistic to establish a standardized clinical pathway of TCM formulas for NAFLD. Therefore, this overview was conducted to critically assess the quality of available systematic reviews, summarize the results, and determine future research directions.
Various clinical and basic studies have reported the effectiveness of TCM formulas for NAFLD. Correspondingly, a number of systematic reviews have been published. It is well known that systematic reviews are high-level clinical evidence and can fundamentally affect the recommendation of an intervention. However, systematic reviews with different conclusions cause doctors to feel confused about their clinical decisions. By objectively evaluating the available systematic reviews, it is possible to sort out the high-quality evidence regarding TCM formulas and recognize future research issues in this area.
The aim of this overview was to critically appraise the available systematic reviews using well-acknowledged tools, and summarize high-quality evidence regarding TCM formulas for treating NAFLD. In addition, based on extracted outcomes, clinicians and researchers can determine directions for further research and prevent unnecessary duplications.
Seven English and Chinese databases were comprehensively searched including PubMed, EMBASE, Cochrane Library, Chinese Biomedical Literature Database (SinoMed), China National Knowledge Infrastructure (CNKI), Chinese Scientific Journal Database (VIP) and Wanfang. The search terms included “nonalcoholic fatty liver disease,” “Chinese medicines,” “systematic review,” and their synonyms. The eligibility of systematic reviews was determined by agreement of two authors or discussion with a third individual. The following information was extracted: authors, titles, year of publication, study size, details of methodological information, details of interventions, data analysis methods, outcomes, adverse effects, and funding information. The methodological quality and risk of bias of the included systematic reviews were assessed by A Measure Tool to Assess Systematic Reviews 2 (AMSTAR 2) and Risk of Bias in Systematic Review (ROBIS), respectively. The Grading of Recommendations Assessment, Development and Evaluation (GRADE) system was employed to appraise the evidence quality of the outcomes of systematic reviews.
Seven systematic reviews were included in total. All systematic reviews were published in the recent decade and conducted based on randomized controlled trials (RCTs). The methodological quality assessment using AMSTAR 2 showed that only one study was in moderate confidence. The remaining systematic reviews were judged as low or critically low confidence. The main reasons for downgrading were protocol registration, adequacy of literature search, justification of excluded studies, and impact of risk of bias on results. The risk of bias assessment employing ROBIS indicated that all systematic reviews were in the high-risk category in overall appraisal (Phase 3). The outcomes of included systematic reviews were assessed by the GRADE system. Most outcomes were downgraded to different levels. The main reasons contained inconsistency due to heterogeneity, imprecision due to inadequate sample size and publication bias due to excessive inclusion of small sample studies. Only two high-quality outcomes were found, namely TCM formulas based on the HuoXueHuaYu principle induced a better ultrasonic improvement rate than conventional medications, and TCM formulas had better efficacy than antioxidants in alanine aminotransferase (ALT) normalization. However, the quality of evidence should be downgraded when applying to clinical practice due to indirectness.
This overview critically evaluated available systematic reviews regarding TCM formulas for treating NAFLD by using well-established tools such as AMSTAR 2, ROBIS, and the GRADE system. The quality of included systematic reviews was not satisfactory. No systematic review was judged as a high confidence level or a low risk of bias. Protocol registration, literature adequacy, and risk of bias analysis were the main shortages. Only two high-quality outcomes were recognized. The corresponding outcomes were TCM formulas based on the HuoXueHuaYu principle showed better ultrasonic improvement than conventional medications, and TCM formulas were superior to antioxidant in ALT normalization. However, when making clinical recommendations, the quality of these outcomes should be further downgraded at least one level due to indirectness. According to the results, although various systematic reviews were published, it is still not sufficient to support the application of TCM formulas to NAFLD in clinical practice. The major reason is the unsatisfactory quality of primary clinical trials. Future emphasis should be paid to designing rigorous RCTs rather than repeatedly conducting systematic reviews.
TCM formulas for NAFLD have always been a popular topic in the gastrointestinal area. Based on available evidence, it is still not possible to establish a recommendation regarding TCM formulas in NAFLD management. Researchers should understand the essence of evidence-based medicine and avoid conducting unnecessary systematic reviews. It would be more valuable to design high-quality RCTs to lay a solid foundation.