Table 3.

Critical appraisal of included study designs, risk of bias.

		Items											Remarks
		Treatment allocation			Blinding			Loss to follow-up				Total score
Study	Eligibility criteria	Randomized	Concealed allocation	Baseline comparability	Subjects	Therapists	Assessors	Measures obtained in 85% of all initially included subjects	Intention to treat	Group comparison reported	Point measures and measures of variability reported
van Brussel et al., (2008) [73]	Yes	Yes	Yes	Yes1	No	No	Yes	Yes	Yes	Yes	Yes	8/10	1Corrections were applied
Mintz-Itkin et al., (2009) [61]	Yes	No2	No2	Yes	No	No	No	No3	No	Yes	Yes	3/10	2No randomization method was specified 3No data on loss to follow-up was presented
Kemp et al., (2009) [59]	Yes	Yes	Yes	Yes4	No	No3	No	No	Yes	Yes	Yes	6/10	3Therapists were blinded for disease characteristics 4No formal statistical data was presented in which baseline comparability was ascertained