Table 2.
Risk of bias assessment (EPOC criteria) of included RCTs, SWTs, CBAs and NRCTs
| Munkhuu 2009 [49] (cRCT) | Turan 2012 [36] (cRCT) | Killam 2010 [46] (SWT) | Geelhoed 2013 [44] (CBA) | Rahman 2011 [48] (CBA) | Bronzan 2007 [43] (NRCT) | |
|---|---|---|---|---|---|---|
| Sequence generation |
U |
L |
H |
L |
N/A |
N/A |
| Allocation concealment |
U |
U |
H |
U |
N/A |
N/A |
| Blinding |
U |
L |
L |
U |
U |
U |
| Complete outcome data |
L |
L |
L |
N/A |
N/A |
N/A |
| No selective outcome reporting |
U |
N/A |
U |
U |
U |
U |
| Group comparability |
L |
L |
L |
U |
L |
U |
| Protection against contamination |
U |
L |
U |
U |
U |
H |
| Free from other sources of bias | L | U | L | U | L | H |
cRCT – cluster–randomized controlled trial, CBA – controlled before–and–after trial, NRCT – non–randomised controlled trial, SWT – stepped wedge trial, H – High risk, L – Low risk, U – Unclear, N/A – Not applicable