Brinkhaus 2006.

Study characteristics
Methods	Three‐arm RCT Number analyzed/randomized: 284/301 Statistical analysis: Intention‐to‐treat analysis (ITT); analysis of covariance (ANCOVA); power analysis Funding source: German social health insurance funds Ethical approval and informed consent obtained
Participants	Participant recruitment: advertisement on local newspapers and by trial centers Setting: hospital outpatient units in Germany Inclusion criteria: 1) clinical diagnosis of chronic LBP lasting > 6 months; 2) aged 40 to 75; 3) average pain intensity of 40 or more (VAS, 0 to 100) in 7 days before acupuncture; 4) had taken only oral nonsteroidal anti‐inflammatory drugs (NSAIDs) for pain in the 4 weeks before the treatment Exclusion criteria: 1) protrusion or prolapse of one or more intervertebral discs with concurrent neurological symptoms; 2) radicular pain; prior vertebral column surgery; infectious spondylopathy; LBP caused by inflammatory, malignant, or autoimmune disease; 3) congenital deformation of the spine (except for slight lordosis or scoliosis); 4) compression fracture caused by osteoporosis, spinal stenosis, spondylolysis or spondylolisthesis; 5) patients with Chinese medicine diagnoses warranting treatment with moxibustion (determined by trial physicians); 6) any acupuncture treatment during the previous 12 months Age (mean ± SD): 58.8 ± 9.1 years Gender (female): 68% Pain duration (mean ± SD): 14.7 ± 11.1 years Pain intensity (mean ± SD): 64.8 ± 14.0 (VAS, 0 to 100)
Interventions	1) GROUP 1: verum acupuncture (semi‐standardized) Acupuncture points: at least four bilateral local points (BL20 to 34, BL50 to 54, GB30, and GV3 to 6, Huatojiaji, and Shiqizhuixia); and at least two bilateral distant points (SI3, BL40, BL60, BL62, KI3, KI7, GB31, GB34, GB41, LV3, GV14, and GV20). If participants were experiencing local or pseudoradicular sensations, at least two local points were chosen. Other acupuncture points, including ear and trigger points, were chosen individually. Depth: not predefined De Qi: stimulated manually at least once Sessions: 12 sessions (30 min each, 2/week for 4 weeks + 1/week for 4 weeks) Acupuncturist experience: ≥ 140 hours of acupuncture training and ≥ 3 years of clinical practice 2) GROUP 2: minimal acupuncture Same treatment protocol as verum acupuncture, with exceptions for acupoints. At least 6 of 10 predefined non‐acupuncture points were needled bilaterally using a superficial insertion. These points were not in the lower back area. 3) GROUP 3: waiting list Participants received delayed verum acupuncture treatment after 8 weeks Co‐intervention: participants allowed to take oral NSAIDs during the treatment Duration of treatment: 8 wks Duration of follow‐up: 13 months after beginning of treatment
Outcomes	1) Pain intensity: VAS (0 to 100); lower values better 2) Back‐specific function status: HFAQ; higher values better 3) Quality of life: SF‐36, higher values better 4) Pain‐related disability: pain disability index (PDI, 0 to 70), lower values better Assessment times: 8, 26, and 52 weeks after starting treatment Costs: NR Adverse effects: similar incidences between acupuncture groups; hematoma and bleeding most commonly reported
Notes	Conclusion: "Acupuncture was more effective in improving pain than non‐acupuncture treatment in people with chronic LBP, whereas there were no significant differences between acupuncture and minimal acupuncture". Language: English For results, see comparisons 1 and 2.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Randomisation list generated by computer software
Allocation concealment (selection bias)	Low risk	Centralised telephone randomisation procedure used
Blinding of participants (performance bias) All outcomes	Unclear risk	Some degree of un‐blinding had arisen, based on testing, but not clear how much this affected outcomes. Testing suggested some degree of successful blinding of participants (P = 0.04).
Blinding of personnel /care providers (performance bias) All outcomes	Unclear risk	Did not describe the masking methods
Blinding of outcome assessment (detection bias) All outcomes	Unclear risk	Not reported whether the assessor was independent
Incomplete outcome data (attrition bias) All outcomes	Low risk	The dropout rate was 8% for each group and reasons were acceptable.
Intention‐to‐treat‐analysis (attrition bias)	Low risk	ITT analysis based on all available data; robustness analysis conducted
Selective reporting (reporting bias)	Low risk	All outcomes reported
Group similarity at baseline (selection bias)	Low risk	Similar baseline characteristics
Co‐interventions (performance bias)	High risk	Statistically significant differences for days with analgesics during last 3 weeks of acupuncture treatment
Compliance bias (performance bias)	Low risk	The compliance was acceptable and comparable across the groups.
Timing of outcome assessments (detection bias)	Low risk	Data on primary and secondary outcomes collected at the same time across groups
Other bias	Low risk	Not identified