Cesare 2011.

Study characteristics
Methods	Two‐arm RCT Number analyzed/randomized: 60/62 Statistical analysis: ITT analysis; ANCOVA; power analysis Funding source: NR Ethical approval and informed consent obtained
Participants	Participant recruitment: at outpatient clinic directly Setting: outpatient clinic in Italy Inclusion criteria: 1) LBP > 6 months; 2) aged > 45 years; 3) no pharmacological therapy for other systemic pathologies; 4) no previous treatment with mesotherapy for chronic LBP Exclusion criteria: 1) presence of neurological pathologies; concomitant severe rheumatic disease; systemic pathologies (e.g. diabetes); 2) surgical intervention in 3 months prior to the study; 3) psychiatric pathologies or patients undergoing psychotherapeutic treatment or physical treatment in 5 weeks prior to the study Age (mean): 52.5 years Gender (female): 55% Pain duration (mean): 4.8 years Pain intensity (mean): 7.4 (VAS, 0 to 10)
Interventions	1) GROUP 1: acupuncture mesotherapy (standardized) Local anesthetic (a drop of lidocaine, 2%) injected at 18 points for both groups before acupuncture. Acupuncture points: local bilateral points (GB30, BL31, BL52, GV3, Ashi points); bilateral distal points (GB34, GB41, BL60, KI4, TE5) Depth: NR De Qi: NR Sessions: 4 sessions (1/week for 4 weeks) Acupuncturist experience: 3 years of acupuncture training and 8 years of clinical experience 2) GROUP 2: trigger point mesotherapy A number of the 18 local superficial trigger points of the most important accessible muscles in lumbar tract and hip were treated (quadratus lumborum – 2 bilateral trigger points: one at the angle between the crest of the ilium and transverse processes of L4 and the other at the transverse processes of L1‐L2; iliopsoas muscle ‐ 2 paravertebral bilateral points: one on the D12 thoracic vertebrae and the other on L2 lumbar vertebrae; gluteus maximus – 3 bilateral trigger points: all on the medial superior portion of sacrum region; piriformis muscle – 2 bilateral points: one at the junction between the middle and the outer third of the piriformis line and the other one at medial end of piriformis line) Co‐intervention: none Duration of treatment: 4 weeks Duration of follow‐up: 12 weeks after starting treatment
Outcomes	1) Pain intensity: VAS (0 to 10) and VRS (verbal rating scale, 0 to 6); lower values better 2) The quality and intensity of pain: McGill Pain Questionnaire (MPQ) Short Form, lower values better 3) Back‐specific function status: Roland Morris Disability Questionnaire (RMDQ, 0 to 24), and Oswestry Low Back Pain Disability Questionaire Index (ODI, 0 to 50); for both, lower values better Assessment times: 4 and 12 weeks after beginning of treatment Costs: NR Adverse effects: five participants in acupuncture group reported slight neck pain
Notes	Conclusion: "Our results suggest that the response to ACP mesotherapy may be greater than the response to TRP mesotherapy in the short term follow‐up." Language: English For results, see comparison 5. Be cautious to the result related with the paper due to extremelly small SD values of pain reported.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Used random numbers from a statistics textbook
Allocation concealment (selection bias)	Low risk	Used sealed, sequentially numbered, opaque envelopes
Blinding of participants (performance bias) All outcomes	Unclear risk	Authors claimed "blinding was maintained until the end of the study", but did not describe the details.
Blinding of personnel /care providers (performance bias) All outcomes	High risk	Not possible
Blinding of outcome assessment (detection bias) All outcomes	Low risk	VAS (pain) scores were recorded by participants reported to be blinded; other outcomes recorded by blinded assessor.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Dropout rate was low (< 5%) and the reasons were acceptable.
Intention‐to‐treat‐analysis (attrition bias)	Low risk	ITT analysis using last value carried forward for missing values
Selective reporting (reporting bias)	Low risk	All of the outcomes were reported.
Group similarity at baseline (selection bias)	Low risk	Similar baseline characteristics
Co‐interventions (performance bias)	Low risk	No co‐intervention was used.
Compliance bias (performance bias)	Low risk	All subjects completed the treatment.
Timing of outcome assessments (detection bias)	Low risk	The outcomes were collected at 4 and 12 weeks.
Other bias	Low risk	Not identified