Table 3. Summary of risk of bias assessment [61].

Studies	Components of risk of bias							Summary risk of bias	Overall RoB	Comments, high risk components
	1	2	3	4	5a	5b	6
Aigner et al. 2006 [29]	U	U	U	L	H	H	H	High (3) Unclear (3) Low (1)	H	Three high component: 5a, 5b, 6 (5a: No primary outcomes pre-specified, 5b: No primary outcomes pre-specified, 6: No primary endpoint specified and No ITT reported)
Bonk et al. 2000 [30]	U	U	H	L	H	N/A	H	High (3) Unclear (2) Low (1) N/A (1)	H	Three high component: 3, 5a, 6 (3: Unblind assessors, 5a: No primary outcomes pre-specified, 6: No primary endpoint specified and No ITT reported)
Borchgrevink et al. 1998 [31]	U	U	L	H	U	N/A	H	High (2) Unclear (3) Low (1) N/A (1)	H	Two high component: 4, 6 (4: loss follow-up >16 without stating of which group, 6: No primary endpoint specified and No ITT reported)
Conforti and Fachinetti 2013 [54]	U	U	L	L	H	N/A	H	High (2) Unclear (2) Low (2) N/A (1)	H	Two high component: 5a, 6 (5a: No primary outcomes pre-specified, 6: No primary endpoint specified, No ITT reported and No statistical analysis reported)
Dehner et al. 2006 [55]	L	L	U	L	H	N/A	H	High (2) Unclear (1) Low (3) N/A (1)	H	Two high component: 5a, 6 (5a: No primary outcomes pre-specified, 6: No primary endpoint specified and No ITT reported)
Dehner et al. 2009 [32]	L	L	U	U	H	N/A	H	High (2) Unclear (2) Low (2) N/A (1)	H	Two high component: 5a, 6 (5a: No primary outcomes pre-specified, 6: No primary endpoint specified and No ITT reported)
Ferrari et al. 2005 [33]	L	L	L	H	L	N/A	H	High (2) Unclear (1) Low (4) N/A (1)	H	Two high component: 4, 6 (4: reasons of drop-out were not reported, 6: No primary endpoint specified and No ITT reported)
Foley-Nolan et al. 1992 [34]	U	U	U	L	H	N/A	H	High (2) Unclear (3) Low (1) N/A (1)	H	Two high component: 5a, 6 (5a: No primary outcomes pre-specified, 6: No primary endpoint specified and No ITT reported)
Jull et al. 2013 [56]	L	L	L	U	L	L	H	High (1) Unclear (1) Low (5)	H	One high component: 6 (No ITT and Pain (VAS) has reported in significant difference between group at baseline)
Ottosson et al. 2007 [35]	L	L	H	L	N/A	L	H	High (2) Low (4) N/A (1)	H	Two high component: 3, 6 (3: Unblind, 6: No primary endpoint specified)
Picelli et al. 2011 [36]	L	U	L	L	H	N/A	H	High (2) Unclear (2) Low (3) N/A (1)	H	Two high component: 5a, 6 (5a: P-value did not report in NDI and PPT, 6: No primary endpoint specified and No ITT reported)
Rosenfeld et al. 2003 [37][Rosenfeld et al. 2006 reporting same trail][57]	L	L	L	H	H	H	H	High (4) Low (3)	H	Four high component: 4, 5a, 5b, 6 (4: drop-out difference between groups, 5a: No primary outcomes pre-specified, 5b: No primary outcomes pre-specified, Reporting sick leave but have not stated, 6: No primary endpoint specified and No ITT reported)
Schnabel et al. 2004 [58]	U	U	U	H	H	N/A	H	High (3) Unclear (3) N/A (1)	H	Three high component: 4, 5a, 6 (4: Loss of follow-up: A = 36%, B = 15%, 5a: No primary outcomes pre-specified, 6: No primary endpoint specified and No ITT reported)
Scholten-Peeters et al. 2006 [59][Scholten-Peeters et al. 2003 trial protocol][62]	L	L	L	L	L	L	H	High (1) Low (6)	H	One high component: 6 (No primary endpoint specified)
Vassiliou et al. 2006 [38]	L	L	L	L	L	N/A	H	High (1) Low (4) N/A (5)	H	One high component: 6(No primary endpoint specified)

1 = Sequence generation, 2 = Allocation concealment, 3 = Blinding of participants, personnel and assessors, 4 = Incomplete outcome data, 5a = Selecting outcome reporting (Short term), 5b = Selecting outcome reporting (Long term), 6 = Other potential threats to validity, L = low risk of bias, H = high risk of bias, U = unclear risk of bias, N/A = not applicable