Acupuncture for chronic nonspecific low back pain.
Jinglan Mu,Andrea D Furlan,Wai Yee Lam,Marcos Y Hsu,Zhipeng Ning,Lixing Lao
DOI: https://doi.org/10.1002/14651858.CD013814
IF: 8.4
2020-12-13
Cochrane Database of Systematic Reviews
Abstract:Chronic nonspecific low back pain (LBP) is very common; it is defined as pain without a recognizable etiology that lasts for more than three months. Some clinical practice guidelines suggest that acupuncture can offer an effective alternative therapy. This review is a split from an earlier Cochrane review and it focuses on chronic LBP. To assess the effects of acupuncture compared to sham intervention, no treatment, or usual care for chronic nonspecific LBP. We searched CENTRAL, MEDLINE, Embase, CINAHL, two Chinese databases, and two trial registers to 29 August 2019 without restrictions on language or publication status. We also screened reference lists and LBP guidelines to identify potentially relevant studies. We included only randomized controlled trials (RCTs) of acupuncture for chronic nonspecific LBP in adults. We excluded RCTs that investigated LBP with a specific etiology. We included trials comparing acupuncture with sham intervention, no treatment, and usual care. The primary outcomes were pain, back‐specific functional status, and quality of life; the secondary outcomes were pain‐related disability, global assessment, or adverse events. Two review authors independently screened the studies, assessed the risk of bias and extracted the data. We meta‐analyzed data that were clinically homogeneous using a random‐effects model in Review Manager 5.3. Otherwise, we reported the data qualitatively. We used the GRADE approach to assess the certainty of the evidence. We included 33 studies (37 articles) with 8270 participants. The majority of studies were carried out in Europe, Asia, North and South America. Seven studies (5572 participants) conducted in Germany accounted for 67% of the participants. Sixteen trials compared acupuncture with sham intervention, usual care, or no treatment. Most studies had high risk of performance bias due to lack of blinding of the acupuncturist. A few studies were found to have high risk of detection, attrition, reporting or selection bias. We found low‐certainty evidence (seven trials, 1403 participants) that acupuncture may relieve pain in the immediate term (up to seven days) compared to sham intervention (mean difference (MD) ‐9.22, 95% confidence interval (CI) ‐13.82 to ‐4.61, visual analogue scale (VAS) 0‐100). The difference did not meet the clinically important threshold of 15 points or 30% relative change. Very low‐certainty evidence from five trials (1481 participants) showed that acupuncture was not more effective than sham in improving back‐specific function in the immediate term (standardized mean difference (SMD) ‐0.16, 95% CI ‐0.38 to 0.06; corresponding to the Hannover Function Ability Questionnaire (HFAQ, 0 to 100, higher values better) change (MD 3.33 points; 95% CI ‐1.25 to 7.90)). Three trials (1068 participants) yielded low‐certainty evidence that acupuncture seemed not to be more effective clinically in the short term for quality of life (SMD 0.24, 95% CI 0.03 to 0.45; corresponding to the physical 12‐item Short Form Health Survey (SF‐12, 0‐100, higher values better) change (MD 2.33 points; 95% CI 0.29 to 4.37)). The reasons for downgrading the certainty of the evidence to either low to very low were risk of bias, inconsistency, and imprecision. We found moderate‐certainty evidence that acupuncture produced greater and clinically important pain relief (MD ‐20.32, 95% CI ‐24.50 to ‐16.14; four trials, 366 participants; (VAS, 0 to 100), and improved back function (SMD ‐0.53, 95% CI ‐0.73 to ‐0.34; five trials, 2960 participants; corresponding to the HFAQ change (MD 11.50 points; 95% CI 7.38 to 15.84)) in the immediate term compared to no treatment. The evidence was downgraded to moderate certainty due to risk of bias. No studies reported on quality of life in the short term or adverse events. Low‐certainty evidence (five trials, 1054 participants) suggested that acupuncture may reduce pain (MD ‐10.26, 95% CI ‐17.11 to ‐3.40; not clinically important on 0 to 100 VAS), and improve back‐specific function immediately after treatment (SMD: ‐0.47; 95% CI: ‐0.77 to ‐0.17; five trials, 1381 participants; corresponding to the HFAQ change (MD 9.78 points, 95% CI 3.54 to 16.02)) compared to usual care. Moderate‐certainty evidence from one trial (731 participants) found that acupuncture was more effective in improving physical quality of life (MD 4.20, 95% CI 2.82 to 5.58) but not mental quality of life in the short term (MD 1.90, 95% CI 0.25 to 3.55). The certainty of evidence was downgraded to moderate to low because of risk of bias, inconsistency, and imprecision. Low‐certainty evidence suggested a similar incidence of adverse events immediately after treatment in the acupuncture and sham intervention groups (four trials, 465 participants) (RR 0.68 95% CI 0.46 to 1.01), and the acupuncture and usual care groups (one trial, 74 participants) (RR 3.34, 95% CI 0.36 to 30.68). The certainty of the evidence -Abstract Truncated-
medicine, general & internal