TOPIC DEVELOPMENT

This topic was developed in response to a nomination from Juli Olson, DC, DACM, National Lead for Acupuncture, Integrative Health Coordinating Center. The scope was further developed with input from the topic nominator, the ESP Coordinating Center, and the review team. The scope of this report includes the following:

1. Evidence maps that provide a visual overview of the distribution of evidence for acupuncture.
2. An accompanying narrative that helps stakeholders interpret the state of the evidence to inform policy and clinical decision-making.

DATA SOURCES AND SEARCHES

The literature searches used for these maps are based on the searches used for the original Evidence Map of acupuncture completed in 2012 and early 2013. Four databases were included in the search, which covered March 2013 to April 2021: PubMed, Allied and Complementary Medicine Database (AMED), Cochrane Database of Systematic Reviews (CDSR), and DARE (Database of Abstracts of Reviews of Effects, ending search in 2014 when DARE ceased production). See Appendix A for full search strategies.

STUDY SELECTION

Each title was screened independently by 2 authors for relevance; any article chosen by either reviewer was included in the abstract screen. Abstracts were then reviewed in duplicate with any discrepancies resolved by group discussion. In order to be included, abstracts or titles needed to be about efficacy or effectiveness of acupuncture for an adult health condition and be a systematic review. A systematic review was defined as a review that had a documented systematic method for identifying and critically appraising evidence. At this stage, we also selected titles and abstracts of systematic reviews about treatments and conditions for which acupuncture might be included; for example, we included titles such as “Interventions for the reduction of prescribed opioid use in chronic non-cancer pain” or “Non-pharmacologic treatments for symptoms of diabetic peripheral neuropathy: A systematic review.” Systematic reviews were still eligible if they covered other interventions and results for acupuncture were reported separately. We did so because reviews with mixed acupuncture modalities included mostly manual acupuncture studies. Interventions such as laser acupuncture, moxibustion alone, needling, and traditional Chinese medicine (TCM) without mention of acupuncture and fire acupuncture were excluded.

We abstracted condition type when reviewing abstracts and presented a list of conditions for which we found reviews to the Operational Partner to determine which conditions were of interest to the VA. Any conditions not selected by the Operational Partner were then excluded from further review.

From this large collection of systematic reviews that included acupuncture as a treatment, we next restricted eligibility to reviews that used formal methods to assess the certainty (or strength or quality) of the evidence for conclusions. In general, this meant using Grading of Recommendations, Assessment, Development and Evaluations (GRADE).⁷ However, other formal methods were also included, such as the approach utilized by the US Agency for Healthcare Research & Quality Evidence-based Practice Center program.⁸ To remain eligible, an included review had to both 1) state or cite the method used and 2) report the certainty (or strength or quality) of evidence for each conclusion (see ^{footnote 1}).

After applying this restriction, many health conditions had only 1 systematic review meeting eligibility criteria, and we used this review for the map.

For some conditions, we identified more than 1 review meeting the eligibility criteria. For these conditions, we first assessed whether the reviews differed in some other feature used to classify reviews on our map – for example, a systematic review on condition X included only studies comparing acupuncture to sham, while another systematic review on condition X only included studies comparing acupuncture to other active therapies. In such cases, we included both reviews on the map, with the appropriate designations (such as “versus sham” and “versus active therapy”). If there were multiple reviews on the same condition, and they did not differ in some other feature, then we selected the 1 systematic review that we judged as being most informative for readers. In general, this was the most recent review or the review with the greatest number of included studies. Systematic reviews otherwise meeting eligibility criteria that were not included in the map for this reason are listed in Appendix B.

DATA ABSTRACTION AND ASSESSMENT

Each included systematic review had data abstracted by 1 reviewer and verified by a second reviewer. Abstracted data included: number of studies included in the review that had acupuncture as the intervention, condition, type of acupuncture, comparators, certainty of evidence statement(s), and main findings relevant to acupuncture as treatment for condition.

SYNTHESIS

Our evidence mapping process resulted in a visual depiction of the evidence for acupuncture, as well as an accompanying narrative with ancillary figures and tables. The visual depiction or evidence map uses a bubble plot format to display information on 4 dimensions: bubble size, bubble label, x-axis, and y-axis. This allowed us to provide the following types of information about each included systematic review, as follows:

Number of articles in systematic review (bubble size): Each systematic review bubble’s size is proportional to the number of primary research studies included in that systematic review related to the effect of acupuncture.

Condition (bubble label): Each bubble is labeled with the condition discussed by that systematic review.

Shapes and colors: Intervention characteristics for each condition are presented in the form of shapes (type of acupuncture) and colors (comparators). For type of acupuncture, rectangle denotes electro-acupuncture only and circle denotes all other types (manual/standard, electro-acupuncture). For comparators, the color red represents sham/placebo, blue for active/usual care, purple for mixed comparators with subgroups, and gray for mixed comparators with no subgroups. A condition can show up more than once if multiple systematic reviews had included either different acupuncture interventions and/or different comparators.

Strength of findings (rows): Each condition is plotted on the map based on the certainty of evidence statement as reported in the systematic review. Many reviews report more than 1 conclusion. Thus, to keep reviews mutually exclusive, we have 3 categories: “All conclusions are rated as low or very low certainty,” “at least 1 conclusion rated as moderate certainty,” and “at least one conclusion rated as high or strong certainty.” For reviews with multiple certainty of evidence statements, we selected the highest certainty of evidence statement.

Effect of acupuncture (columns): Each condition is plotted in either “benefit” or “no benefit” as effect of acupuncture based on conclusion of systematic review.

Narrative synthesis: The narrative synthesis expands upon the visual evidence map to provide overarching conclusions from the maps. Details about the conclusions in individual reviews are included in an appendix.