Cho 2013.

Study characteristics
Methods	Two‐arm RCT Number analyzed/randomized: 116/130 Statistical analysis: per‐protocol analysis; Student's t‐test and mixed model; power analysis Funding source: Korea Health Industry Development Institute (B080048) Ethical approval and informed consent obtained
Participants	Participant recruitment: local newspapers, hospital's magazine, website, and bulletin boards Setting: three medical hospitals in Korea Inclusion criteria: 1) chronic LBP ≥ 3 months before acupuncture treatment; 2) baseline VAS (bothersomeness, 0 to 10) > 5 points; 3) nonspecific and uncomplicated LBP that was intact on neurological examination Exclusion criteria: 1) sciatic pain; 2) pain mainly below knee; 3) serious spinal disorders including malignancy, vertebral fracture, spinal infection, inflammatory spondylitis, and cauda equine compression; 4) history of previous spinal surgery Age (mean ± SD): 42.1 ± 14.0 years Gender (female): 85% Pain duration (mean ± SD): NR Pain intensity (mean ± SD): 6.45 ± 1.30 (VAS, 0 to 10)
Interventions	1) GROUP 1: verum acupuncture (semi‐standardized) Acupuncture points (according to the diagnosis): gallbladder meridian pattern ‐ GB12, GB26, GB30, GB34, GB41; bladder meridian pattern ‐ BL23, BL24, BL25, BL37, BL40; mixed pattern: ST4, ST36, SP13, SP14, GV3, GV4, GV5, GV24, and GV26 Depth: needles inserted vertically to a depth of 5 to 20 mm, depending on the site De Qi: needles left in situ for 15 to 20 min after De Qi sensation elicited by manual stimulation Sessions: 12 sessions (2/week for 6 weeks) Acupuncturist experience: licensed Korean Medicine Doctors with at least 3 years experience 2) GROUP 2: sham acupuncture Same technique and protocol as verum acupuncture except that semi‐blunt needles were used without penetration on 8 predefined non‐acupuncture points (1 cm below BL39, 1 cm lateral to BL18 and BL20, and 2 cm above GB30, all bilaterally) 3) Co‐intervention: participants were requested to do exercises every day according to the given manual; any additional therapy was prohibited. Duration of treatment: 6 weeks Duration of follow‐up: 24 weeks
Outcomes	1) Pain intensity: VAS (0 to 10) 2) Back‐specific function status: Oswestry Disability Index (ODI, 0 to 50) 3) Quality of life: SF‐36, higher values better Assessment times: 6,8,12, and 24 weeks after beginning of treatment Costs: NR Adverse effects: 16/27 participants reported minor to moderate adverse events, none of which persisted longer than 1 weeks.
Notes	Conclusion: “This randomised sham‐controlled trial suggests that acupuncture treatment shows better effect on the reduction of the bothersomeness and pain intensity than sham control in participants with chronic LBP.” Language: English Using the baseline SDs of SF‐36 and ODI as the corresponding SDs of the results in meta‐analysis For results, see comparison 1.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Randomisation code generated by computer
Allocation concealment (selection bias)	Low risk	The random code was kept by a clinician who did not contact participants.
Blinding of participants (performance bias) All outcomes	Low risk	Subjects were blinded and assessment showed adequate blinding.
Blinding of personnel /care providers (performance bias) All outcomes	Unclear risk	Unclear if acupuncturist was successfully blinded or not
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Blinded participants reported the outcomes.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Acceptable and balanced dropout rate (11%) across groups; the reasons were largely acceptable.
Intention‐to‐treat‐analysis (attrition bias)	High risk	The study used per protocol analysis.
Selective reporting (reporting bias)	Low risk	All outcomes were reported.
Group similarity at baseline (selection bias)	Unclear risk	The baseline ODI scores in the acupuncture group were significantly higher.
Co‐interventions (performance bias)	Unclear risk	Unclear risk because the usual care group was visited by a physical therapist twice more than the acupuncture groups.
Compliance bias (performance bias)	Low risk	Subjects were asked to complete > 80% treatments, and they possibly did, according to the report.
Timing of outcome assessments (detection bias)	Low risk	Same assessment time
Other bias	Low risk	Not identified