The authors concluded that acupuncture may reduce pain in patients with peripheral osteoarthritis (OA) and could be considered for patients with knee OA. Overall, this was a well-conducted and clearly reported review. The authors’ conclusions reflect limited evidence from a small number of studies and are likely to be reliable.

To evaluate the effectiveness of acupuncture in patients with osteoarthritis (OA) of the peripheral joints.

MEDLINE, EMBASE, AMED, CINAHL, British Nursing Index, PsycINFO, University of Maryland's CAMPAIN database and the Cochrane Library were searched to July 2005; the search terms were reported. No language restrictions were applied. In addition, the authors’ own files and conference proceedings were searched and reference lists were screened.

Randomised controlled trials (RCTs) that compared needle acupuncture, with or without electrical stimulation, with a control (sham, placebo or another non-acupuncture intervention) in patients with peripheral joint OA were eligible for inclusion. Studies of other forms of acupuncture were excluded. Studies had to report data or statistical comparisons between treatments. The primary review outcome was the change from baseline in pain scores on a 100-mm visual analogue scale (VAS) and on the Western Ontario MacMaster Osteoarthritis Index (WOMAC).

Most of the included studies were in patients with knee OA; other studies were in patients with hip or thumb OA. The included studies compared manual and electro-acupuncture with a wide variety of control treatments, such as waiting list, sham/placebo acupuncture, superficial needling, medications (e.g. non-steroidal anti-inflammatory drugs), ice, transcutaneous electrical nerve stimulation (TENS), exercise, hydrotherapy, and education or advice. Treatment sessions generally lasted 20 to 30 minutes.

The authors did not state how the papers were selected for the review, or how many reviewers performed the selection.

Validity was assessed using a modified Jadad scale, which considered the reporting and handling of randomisation, blinding and the handling of withdrawals. Studies scoring 4 or 5 out of the maximum 5 points were considered to be high quality. In addition, studies of knee and hip OA were classified as adequate if at least 4 points were needled for at least 20 minutes, with at least 6 treatments given at least once weekly, and there was sufficient electrical stimulation for a minimal sensation or deqi (needle sensation) to be achieved.

Two reviewers independently assessed validity. Any differences were resolved through discussion.

For each study, mean changes from baseline of eligible pain scores (VAS and WOMAC) were extracted. Post-treatment means of treatment differences and 95% confidence intervals (CIs) were calculated. Authors were contacted for missing information.

Two reviewers independently extracted the data onto a standardised form. Differences were resolved through discussion.

The studies were grouped by type of intervention and comparator and combined in narrative. For homogeneous studies, pooled standardised mean differences (SMDs) with 95% CIs were calculated using a fixed-effect model. Statistical heterogeneity was assessed using the χ² test. There were insufficient data for sensitivity analyses. Differences between the studies in terms of study quality, sample size, analysis, and characteristics of treatment and control groups, were assessed.

Eighteen RCTs (n=1,891) were included. Some studies had more than one control group. Deqi or numbness was reported in 13 studies.

Two sham-controlled studies reported blinding of the assessor and patient and scored 5 points on the Jadad scale. Eight studies scored 4 points. Six studies reported patient blinding and 13 studies reported blinding of the assessor. In all but one study acupuncture was considered adequate.

Overall, 10 of the 18 studies reported reduced pain in acupuncture groups compared with control. Seven of the 10 high-quality studies reported reduced pain in acupuncture groups compared with control.

Manual acupuncture.

Two studies reported significantly reduced pain in manual acupuncture groups compared with a waiting-list control. Three of the 4 studies that compared manual acupuncture with superficial needling reported reduced pain in acupuncture groups. The meta-analysis (3 studies, n=329) showed a significant reduction in pain in acupuncture groups compared with sham (SMD 0.24, 95% CI: 0.01, 0.47, p=0.04). One study reported no difference between manual acupuncture and mock TENS. Three studies reported mixed results for manual acupuncture compared with conventional treatment. One study reported insufficient data.

Electro-acupuncture.

Three studies reported significantly reduced pain in electro-acupuncture groups compared with sham or placebo acupuncture groups.
Three studies reported reduced pain in electro-acupuncture groups compared with oral treatments. Two studies reported no difference between electro-acupuncture and TENS. One of the 3 studies that compared electro-acupuncture with education reported reduced pain in acupuncture groups.

Nine out of 14 studies in patients with knee OA reported reduced pain in acupuncture groups. A meta-analysis of 2 studies of knee OA (n=264) showed a significant reduction in pain in acupuncture groups compared with sham (SMD 5.40, 95% CI: 0.91, 9.99, p=0.02).

Sham-controlled RCTs suggest that acupuncture may reduce pain in patients with peripheral OA and acupuncture could be considered for patients with knee OA.

The review question was stated clearly. Several relevant sources were searched and attempts were made to minimise publication and language bias. Methods were used to minimise reviewer error and bias in the assessment of validity and extraction of data, but it is unclear whether similar steps were taken when selecting the studies. Only RCTs were included, and validity was assessed and the results reported. In view of the differences between the studies, a narrative synthesis with pooling of only clinically and statistically homogeneous data was appropriate. The results were also discussed with respect to quality. Overall, this was a well-conducted and clearly reported review. The authors’ conclusions reflect limited evidence from a small number of studies and are likely to be reliable.

Practice: The authors stated that acupuncture should be considered for knee OA.

Research: The authors stated the need for more studies to evaluate manual and electro-acupuncture for hip OA, and more studies to compare acupuncture with non-steroidal anti-inflammatory drugs.

Wonkwang University, Soongsan grant.

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.