How effective is dry needling? an umbrella review

Several previous systematic reviews (SRs) have summarized the evidence on the safety and effectiveness of dry needling (DN). Existing SRs often have methodological limitations, and many are outdated, leading to inconsistent and conflicting conclusions. The aim of this umbrella review (UR) is to evaluate the available SRs on the safety and effectiveness of DN across various clinical conditions, and to re-evaluate and contextualize the previously conflicting evidence. PubMed, MEDLINE (via Ovid), Cochrane Database of Systematic Reviews, Embase, Web of Science, Scopus, PEDro and Epistemonikos were searched (from Jan 2000 to Dec 2025) for relevant SRs. In this UR, SRs investigating DN with or without usual care (UC) performed by various healthcare professionals for treating musculoskeletal or non-musculoskeletal pain or dysfunction in human subjects were considered eligible. Non-systematic or narrative reviews were excluded. We strictly followed the PRIOR guidelines, assessed methodological quality using ROBIS tool, extracted effect sizes from meta-analyses, and evaluated overlap using the Corrected Covered Area (CCA). A total of 75 SRs including 45,032 individuals met the eligibility criteria. Fifty SRs (66.6%) reported author conclusions favoring DN, 17 SRs (22.6%) reported equivocal conclusions, and eight SRs (10.6%) reported negative conclusions; these categories reflect the direction of conclusions stated by the SR authors. Fifty-seven SRs (76%) demonstrated high risk of bias across multiple domains, particularly in study selection, data appraisal, and synthesis. Among the 18 SRs judged to be at low risk of bias, 12 (66.6%) reported favorable conclusions, five (27.7%) equivocal conclusions, and one (5.5%) negative conclusion. Thirty‑six SRs (48%) reported adverse-effects (AEs), while the remaining 39 (52%) failed to provide any AE data. The amount of overlap was slight (CCA = 1.09). This UR provides a comprehensive mapping of the existing systematic review literature on DN. Although many reviews report beneficial effects of DN, the certainty of evidence remains limited due to high risk of bias and substantial heterogeneity. Accordingly, no definitive conclusions regarding clinical effectiveness in reducing pain and disability can be drawn, and the findings should be interpreted with caution. Safety reporting was inconsistent across SRs, with AEs frequently underreported. High-quality, condition-specific SRs and randomized clinical trials are required. CRD420251058215.