The criteria for selection were as follows: (a) SRs/MAs of only enrolled randomized controlled trials (RCTs), (b) subjects were diagnosed with FD according to the Rome diagnostic criteria, (c) interventions applied to the experimental group included acupuncture alone or in combination with conventional medicine (CM), while interventions applied to the control included sham acupuncture or CM, (d) reporting at least one prespecified outcome: symptom score (SS), nepean dyspepsia index (NDI), main frequency of electrogastrogram (MF-EGG), gastric half-emptying time (GHET), plasma motilin level (PML), and effective rate (ER). The exclusion criteria were as follows: (a) repeated publications, (b) graduate dissertation, (c) lack of further data, and (d) conference abstracts.

Systematic reviews/Meta-analysis of acupuncture on FD were systematically searched in the following databases from their inception till August 2021: the Cochrane Library, PubMed, Web of Science, Embase, Chinese Biomedical Database, Chinese VIP information, and Wanfang Data. Functional dyspepsia, acupuncture, and systematic review were used as key search terms. A strategy for using PubMed search is shown in Table 1.

Literature selection and data extraction were conducted by two independent reviewers. For literature selection, titles and abstracts were screened first, and the full texts of potential literature were read for further evaluation. From the articles included, the following data were extracted: first author, publication year, sample size, interventions, outcomes, and main results. Disagreements were settled via discussion and consensus.

Methodological quality, reporting quality, and evidence quality of the enrolled studies were evaluated by two independent reviewers using the AMSTAR-2 tool, the PRISMA statement, and the GRADE system, respectively. The AMSTAR-2 consisted of 16 items, and the methodological quality was ranked as high, moderate, low, or critically low (8). The PRISMA statement consisted of 27 items, with each of the 16 items given a rating of yes, no, or partly yes (9). Evidence quality with GRADE was considered from five aspects (risk of bias, inconsistency, indirectness, imprecision, and publication bias) and given a rating of high to critically low (10). Disagreements were settled via discussion and consensus.

The literature search yielded 193 citations, of which 55 duplicates were removed. After screening the title and abstract, 121 citations were excluded and the remaining 17 were further evaluated through full-text reading. Finally, 10 SRs/MAs (11–20) met the inclusion criteria (Figure 1).

Then, SRs/MAs published from 2014 to 2021 were included. The interventions applied to the experimental groups were acupuncture alone or in combination with CM, while those administered to the control groups were CM or sham acupuncture. More details are presented in Table 2.

According to the AMSTAR-2, only one Cochrane SR/MA presented high level of methodological quality, while the other SRs/MAs presented critically low levels of methodological quality. Lack of protocol registration, no list of excluded trials, or lack of a comprehensive search strategy were the main limitations of the included SRs/MAs (Figures 2, 3).

According to the PRISMA statement, only one Cochrane SR/MA fully reported information on the 27 items, while the other SRs/MAs missed serious information in item Q5 (protocol and registration), Q8 (search), and Q27 (funding). More details are shown in Table 3.

With the GRADE system, the evidence quality of all concerned outcome indicators ranged from “critically low” to “moderate” due to the risk of bias, inconsistency, imprecision, and publication bias. Furthermore, upgrading factors such as large effect and evidence of dose-response gradient were not applicable to the included outcome indicators. More details are shown in Table 4.

First, 10 SRs/MAs published between 2014 and 2021 were included, nine of which were published over the past 5 years, indicating an increasing interest in acupuncture for FD. Second, based on the AMSTAR-2 tool, the PRISMA statement, and the GRADE system, the methodological quality, reporting quality, and evidence quality, respectively, of the included studies did not meet the requirements of the Cochrane Handbook, which means that the conclusions of the included studies may be different from the real situation and cannot provide reliable evidence for the investigators. Third, there is considerable room for improving the methodological quality and reporting the quality of SRs/MAs. This study highlights common areas for improvement such as protocols designed and registered in advance, comprehensive literature search strategies, providing a transparent list of excluded trials, and declaring funding and conflicts of interest. Fourth, there is considerable room for improving the methodological quality of RCTs. This study highlights the common areas for improvement in RCTs, such as a strict implementation of randomization and allocation concealment and adequate reporting of blinding methods. In addition, while almost all outcome measures lead to positive conclusions, it is worth noting that most authors were reluctant to draw definitive conclusions on the effects of acupuncture for FD due to the small size or low quality of the RCTs.

This study highlights several challenges for producers of SRs/MAs that should be addressed. First, by using the AMSTAR-2 tool and the PRISMA statement, it was observed that the registration of protocols was far from being appreciated by authors of SRs/MAs. Only one Cochrane SR/MA provided a protocol registration, whereas none of the other studies provided that. It is well-known that a preregistration protocol helps to improve transparency, minimize the potential risk of bias arising from the research process, reduce duplicate work between different researchers, and keep the study up-to-date (22). Authors are advocated to register protocols in a freely open database (PROSPERO, http://www.crd.york.ac.uk/prospero) to avoid study bias. Secondly, the risk of bias in identification and literature selection was far from being paid attention to by authors of SRs/MAs. That is, authors of SRs/MAs should provide a comprehensive search strategy and be aware that an appropriate range of databases or electronic sources of published literature should be included. In addition to the database search, other methods were considered for identifying relevant literature, such as meeting reports and trial registration platforms. Furthermore, a list of excluded trials and account for exclusion help to guarantee transparency; it can be presented by direct reference or as a supplementary file. Finally, funding sources should be fully reported as results from commercially funded research might be biased toward funders (23).

Based on the GRADE system, the risk of bias of the enrolled RCTs was the most common downgrading factor, indicating that the root cause of the reduced quality of evidence lies in the quality of the original RCTs. Only well-designed and rigorously conducted RCTs may reduce or avoid bias (24). Specific methods of randomization should be clearly described to reflect whether randomization has been successfully achieved. In addition, the information on allocation concealment and blind method should be fully reported. Due to the peculiarity of acupuncture, although blinding physicians is challenging, attempts should be made to blind patients, other caregivers, and outcome assessors to minimize the risk of bias. With the development of evidence-based acupuncture, it is hoped that researchers will continue to promote standardization and precision in the future by improving the technical operating procedures of acupuncture. Moreover, it should be realized that a large sample size and high-quality RCTs should be fundamental to high-quality evidence sources.

This study highlights areas of improvement in the SR/MA process that may help guide future high-quality research. However, the widely used AMSTAR-2 tool, PRISMA statement, and GRADE system are all subjective evaluation tools, and the accuracy of assessor's assessments cannot be guaranteed.