Table 3.
Quality assessment of the individual studies using the Cochrane Collaboration’s tool.
| Study | Random sequence generationa | Allocation concealmenta | Performance biasb | Detection bias | Attrition biasc | Reporting biasd | Overall gradee | |
| Agyapong et al, 2012 | + | − | − | − | + | + | 3 | |
| Ahtinen et al, 2013 | N/A | N/A | − | − | + | + | 4 | |
| Aikens et al, 2015 | − | − | − | − | + | + | 2 | |
| Askins et al, 2009 | + | ? | − | − | − | + | 2 | |
| Ben-Zeev et al, 2014 | N/A | N/A | − | − | + | + | 4 | |
| Burns et al, 2011 | N/A | N/A | − | ? | + | + | 4 | |
| Carissoli et al, 2015 | ? | ? | − | − | + | + | 2 | |
| Dagöö et al, 2014 | + | + | − | − | + | + | 4 | |
| Depp et al, 2015 | + | + | + | − | − | + | 4 | |
| Enock et al, 2014 | ? | ? | ? | − | − | + | 1 | |
| Gorini et al, 2010f | ? | ? | − | − | ? | − | 0 | |
| Granholm et al, 2012 | N/A | N/A | − | − | − | + | 3 | |
| Grassi et al, 2011f | ? | ? | − | − | ? | − | 0 | |
| Grassi et al, 2007 | ? | ? | − | − | ? | − | 0 | |
| Harrison et al, 2011 | N/A | N/A | − | − | − | + | 3 |
|
| Huffziger et al, 2013 | + | ? | − | − | + | + | 3 | |
| Kenardy et al, 2003 | ? | ? | − | − | ? | + | 1 | |
| Lappalainen et al, 2013 | ? | ? | − | − | + | + | 2 | |
| Ly et al, 2014 | + | + | − | − | + | + | 4 | |
| Ly et al, 2012 | N/A | N/A | − | − | + | + | 4 | |
| Newman et al, 2014 | ? | ? | − | − | + | + | 2 | |
| Newman et al, 1997 | ? | ? | − | − | + | + | 2 | |
| Pallavicini et al, 2009 | + | ? | − | − | + | − | 2 | |
| Preziosa et al, 2009f (studies 1 and 2) | ? | ? | − | − | ? | − | 0 | |
| Proudfoot et al, 2013 | + | + | + | − | − | + | 4 | |
| Repetto et al, 2013 | + | ? | − | − | + | − | 2 | |
| Riva et al, 2006f | ? | ? | − | − | ? | − | 0 | |
| Rizvi et al, 2011 | N/A | N/A | − | − | + | + | 4 | |
| Shapiro et al, 2010 | N/A | N/A | − | − | − | + | 3 |
|
| Watts et al. 2013 | + | + | − | − | − | + | 3 | |
| Wenze et al, 2014 | N/A | N/A | − | ? | + | + | 4 | |
| Zautra et al, 2012f | ? | ? | − | − | + | + | 2 | |
|
|
|
|
|
|
|
|
|
aThe label “not applicable” (N/A) is used in 1-armed studies.
bThe risk for performance bias is rated low if personnel are blinded irrespective of whether participants were blinded.
cThe bias for attrition is considered high when the attrition from pre-intervention to post-intervention is 20% or more.
dThe bias for selective reporting is labeled low if all prespecified outcomes are reported, it is not necessary that all statistical information is reported per outcome (eg, means, standard deviation, CI, P values).
eThe overall grade is determined by summing the number of low-risk categories and the number of N/A categories; +=low risk of bias; −=high risk of bias; ?=unclear risk of bias.
fStudy is not included in the meta-analysis.