Donzelli 2006.

Methods	quasi‐RCT
Participants	43 participants. 1. Back School group n = 22. 2. Pilates group n = 21. Inclusion criteria: chronic LBP without peripheral irradiation for at least 3 months; neurological values within the normal range; negative Lasegue’s test and Wassermann test. Exclusion criteria: clinical history of spinal surgery; neurological values outside the normal range; radicular pain with positive Lasegue’s and Wassermann’s signs and straight leg raise test; structural deformities such as spondylolisthesis; stenosis of the vertebral canal; computed tomography or nuclear magnetic resonance documented disc hernia; rheumatoid arthritis or other rheumatologically related pathologies; conditions unrelated to the spinal column that mimic lumbalgic symptoms.
Interventions	1. Back School group: 10 sessions/60 minutes. During each session, participants performed all the exercises listed in the protocol. The protocol included postural education exercises, respiratory education, muscular extension and strengthening exercises of the paravertebral muscles and lower limbs, mobilising exercises for the spinal column and antalgic postures. During each treatment session, the therapist taught the participants some theoretical notions of the anatomy and pathology of the spinal column and in the principles of postural education. 2. Pilates group: 10 sessions/60 minutes. The protocol comprised a programme of exercise modules that made it easier to adapt the exercise to the requirements of each participant in each group. The protocol comprised postural education, stretching exercises, and breathing education.
Outcomes	1. Pain: visual analogue scale. 2. Disability: Roland‐Morris Disability Questionnaire.
Notes	Secondary care setting. Funding: N/A.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	High risk	Used a quasi‐random procedure
Allocation concealment (selection bias)	Unclear risk	The sequence generation procedure or the method of allocation were not mentioned. The title and abstract indicate that it is an RCT.
Blinding of participants and personnel (performance bias) All outcomes	High risk	No mention of any attempts to blind the participants
Blinding (performance bias and detection bias) All outcomes	High risk	No mention of any attempts to blind the care providers
Blinding of outcome assessment (detection bias) All outcomes	High risk	No mention of any attempts to blind the assessors
Incomplete outcome data (attrition bias) All outcomes	Low risk	The percentage of withdrawals and dropouts was within the acceptable rate (less than 20%).
Intention‐to‐treat Analysis	Unclear risk	Not mentioned
Selective reporting (reporting bias)	Unclear risk	It was unclear if the published report included all expected outcomes.
Similarity of baseline characteristics?	Low risk	Demographic and clinical baseline characteristics were similar for both groups.
Co‐interventions avoided or similar?	Unclear risk	There were few reported co‐interventions in the study.
Compliance acceptable?	Low risk	Compliance was acceptable based on the descriptions of both groups.
Timing outcome assessments similar?	Low risk	All important outcomes assessments for both groups were measured at the same time.