Table 1.

Risk of bias assessment, based on Higgins et al ¹³

Author, year	Random sequence generation	Allocation concealment	Blinding of participants and personnel	Blinding of outcome assessment	Incomplete outcome data	Selective reporting	Other bias
Ackermann, 200516	+	+	+	+	+	+	+
Burton, 199515	?	?	?	?	?	?	–
Christian, 200832	+	+	+	?	+	+	?
Dubbert, 200229	?	?	?	+	?	+	?
Galaviz, 201317	?	?	–	?	+	+	+
Galaviz, 201718	?	?	–	?	?	+	–
Goldstein, 199933	?	?	–	?	?	+	–
Grandes, 200919	+	–	–	+	+	+	?
Harris, 201536	+	–	–	–	+	+	+
Harris, 2017a38	?	?	–	?	+	+	+
Harris, 2017b34	+	+	–	–	+	+	+
Jansink, 201320	?	?	–	?	–	+	?
Jolly, 201830	?	–	–	+	+	+	–
Kerse, 199939	?	+	–	+	+	+	–
Koelewijn-van Loon, 201021	?	?	–	+	+	+	?
Lakerveld, 201322	+	+	–	+	–	+	?
Leonhardt, 200823	?	?	–	?	+	+	+
Little, 200435	?	?	?	+	+	+	?
Marshall, 200536	?	?	–	+	+	+	–
McCallum, 200724	+	+	–	+	+	+	–
Mehring, 201331	+	+	–	–	+	+	–
Sims, 199925	?	?	?	?	?	+	?
Valve, 201326	+	?	–	?	+	+	?
Van der Weegen, 201527	?	+	–	+	+	+	+
Westland, 202028	+	+	–	–	+	+	–

+ = low risk of bias. – = high risk of bias. ? = unclear risk of bias. Allocation concealment, as well as blinding of participants and clinicians delivering the intervention, was not possible in most study designs. All studies reporting follow-up data had reported data loss. If data loss was <15% and loss is even across groups or the loss was accounted for conservatively in data analysis (for example, intention to treat with replacing missing follow-up data with baseline values), the data loss was rated as low risk of bias. If adherence to the intervention was either not reported or <80% it was rated as high risk of bias in ‘other bias’.