Table 2.

Characteristics and risk of bias assessment of included trials

Author	Licciardone, 2003 [21]	Chown et al., 2008 [26]
Participants	N=91, 21-69yo	N=239, 18-65yo
Randomised	Yes	Yes
Blinding	Patients	Patients and Assessors
Inclusion	NSLBP for 3 months	NSLBP for 3 months
Exclusion	Red flags, neurological signs, surgery, workers comp, pregnancy, recent manipulation	Red flags, radiculopathy, surgery, anti-coagulants
Intervention detail	Senior osteopathic students	One osteopath
	Choice of soft tissue, MET, Art, HVT, SCS, cranio-sacral, myofascial technique	Choice of soft tissue, MET, Art, HVT, functional, exercise, education, psychosocial, nutritional advice
	Seven sessions over 5 months	Five sessions over 3 months
	Follow up at 1, 3 and 6 months	Follow up 6 weeks and 12 months
Control	Sham or no treatment	Manipulative PT or group exercise
Outcome measures	SF-36, VAS, RM, ODI, satisfaction questionnaire	ODI, EuroQoL, Shuttle walk test, satisfaction questionnaire
Main results	SF-36:	(For osteopathy only)
	1 month OMT >control (p=0.03)	ODI - 5.0 (95% CI 1.6 – 8.4; SD 10.5; p<0.01):
	3 months Sham > OMT/control (p=0.01)	EQ-5D 0.11 (CI 0.02 to 0.19; SD 0.24; p<0.05):
	6 months Sham > OMT/control (p=0.03)	Group comparison not done
	VAS pain:
	1 month OMT/Sham >control (p=0.01/0.003)
	3 months OMT/Sham >control (p=0.001/0.001)
	6 months OMT/Sham >control (p=0.02/0.02)
	RM no differences
	OMT less co-treatments (p=0.03)
Risk of Bias score /12 Detail of point loss	7	9
	Randomisation process not fully described	Patients not blinded
	Care provider not blinded	Care provider not blinded
	Drop out rate not fully described	Compliance not acceptable
	Co-interventions not avoided
	Compliance not acceptable
Quality Issues	Confounders in sham techniques, co-treatments	Sample size reduced
		Statistical analysis incomplete