headache in 2012 (38); the Scottish Intercollegiate Guidelines Network guideline for the Management of Chronic Pain, inclusion of acupuncture (Grade A recommendation) for chronic low back pain and osteoarthritis (39); and the National German Gynaecologic Oncology Association’s (Arbeitsgemeinschaft Gynakologische Onkologie) inclusion of acupuncture for 12 symptoms associated with breast cancer treatment in 2015 (36, 37). Table 2. Statistical summary of findings of this review Evidence Level Number of Conditions Changes in Level of Evidence Number of Conditions Evidence of Positive effect 8 Increase to positive effect 5 Evidence of Potential positive effect 38 Increase to potential positive effect 18 Unclear/insufficient evidence 71 Increase to unclear/insufficient evidence 1 No evidence of effect 5 Decreased evidence level 2 Total conditions with some evidence of effect (any level) 117 _______________________ ____ Total conditions reviewed 122 Total increases in evidence level since prior reviews 24 1.6 Interpreting the findings This review set out to identify the current state of evidence regarding acupuncture, and has done so principally by examining systematic reviews and meta-analyses. Most of these systematic reviews were restricted to only randomised controlled clinical trials which examine efficacy, not effectiveness (NHMRC level I evidence); however, some systematic reviews did include pragmatic trials and other uncontrolled studies. Consequently, studies which focused on effectiveness rather than efficacy such as cohort studies, case-control studies, or case series (NHMRC level III-IV evidence) were generally excluded from this review. Earlier discussion has demonstrated the limitations of using placebo controlled clinical trials to assess acupuncture’s efficacy, with the NIH’s acknowledgement of this reflected in its current programmatic priorities which encourage Placebo controlled clinical trials consistently underestimate the true effect size of acupuncture, yet over 1000 trials have still demonstrated NHMRC Level I evidence for the effectiveness of acupuncture for 117 conditions. The Acupuncture Evidence Project © AACMA 6 pragmatic trials at the expense of RCTs (31). Regardless of the limitations of acupuncture RCTs that under-report the true effect size of acupuncture (23, 28), for eight conditions across four clinical areas there is no longer any doubt of acupuncture’s efficacy. For the reasons outlined above, it is not correct to infer that acupuncture is ineffective for conditions which fall outside of the ‘evidence of positive effect’ category. For a further 109 conditions examined, positive results have been achieved from some trials in every case, with varying levels in the quality of the evidence. Acupuncture has already been incorporated into clinical guidelines for over 100 conditions even though its current evidence level is rated below ‘evidence of positive effect’ in most of these cases. This reflects the weighting of other factors in guideline development, and recognition that the quality of levels of evidence are adequate to support a given recommendation, even though they may fall below the highest level in some cases (8). As recommended by both the Society for Acupuncture Research and the NIH, future acupuncture research to inform clinical practice guidelines should be focused on pragmatic trials, head-to-head comparison studies with other interventions (especially currently-recommended usual care interventions) to compare effectiveness, safety and cost-effectiveness, underpinned by further mechanism studies (25, 31). It is accurate to state from this report that there is NHMRC Level 1 evidence for acupuncture’s effectiveness for 117 conditions. The number of conditions included in this review is significantly larger than in previous reviews for two reasons. Firstly, acupuncture researchers have just begun to research the effectiveness of acupuncture for conditions not previously studied. Secondly research has become more nuanced and targeted. For example, what was previously reviewed as ‘cancer adverse effects’ is now separated into 12 different cancer-related conditions such as pain, fatigue, insomnia and xerostomia. 1.7 Limitations This review has several limitations. Literature before March 2013 was obtained by pooling results from the DVA 2010 and USVA 2013 reviews. Any MA or SR overlooked by those reviews were not identified in this review. MAs and SRs from March 2013 to September 2016 were identified via a PubMed search. MAs and SRs which were not indexed by PubMed were not identified in this review with one exception. Three additional SRs and MAs published between October 2016 and January 2017 were included in the final revision. The findings of this review are limited by the mounting evidence that sham/placebo controls used in acupuncture trials are not inert, which is likely to lead to a consistent underestimation of the true effect size of acupuncture interventions (see sections 1.3 and 1.4 above). By design this review aimed to update previous reviews, and focused on MAs and SRs. Studies which focus solely on effectiveness rather than efficacy such as cohort studies, case-control studies and case series were excluded by most of the SRs. This results in under-reporting studies on the effectiveness of acupuncture. For example, a PubMed search on 3 November 2016 for the following key