Table 3.
Risk of bias
| Authors | Full-sample randomization | Blinding: participantsa | Blinding: assessorsa | Drop out | Incomplete patient data | Selective outcome reporting | Data substitution | Paymenta | Conflict of interest |
|---|---|---|---|---|---|---|---|---|---|
| Bor et al., 2011 | Yes | No | Yes | 15% | NR | No | NR | Yes | NR |
| Dale et al., 2016 | Yes | Yes | Yes | NR | 18% 3 patients without NPD | No | NR | Yes | One author (consultant for the CR program company) |
| Eack et al., 2010 | Yes | No | No | 13% (33% at 1 year; 37% at 2 years) | 7% | No | using maximum-likelihood estimation | No | Two authors (developed the CR program) |
| Habel et al., 2010 | Yes | No | No | NR | 6 patients without affect recognition and 7 without age discrimination behavioral data | No | NR | No | One author (developed the CR program) |
| Haut et al., 2010 | Yes | No | – | 16% | 12% | No | NR | No | NR |
| Hooker et al., 2012, 2013 | Yes | Yes | Yes | 14% | 7% | No | subject pre to post/post to pre data | Yes | Two authors (consultants for the CR program company) |
| Kariofillis et al., 2014 | Yes | No | Yes | NR | 3 patients without follow-up | No | NR | Yes | NR |
| Luckhaus et al., 2013 | Yes | No | – | 21% | NR | No correlation between behavioral and NBD | for group mean | No | Two authors (developed the CR program) |
| Penadés et al., 2013 | Yes | No | Yes | NR | 14% | No | NR | No | NR |
| Popov et al., 2011, 2012 | 3 patients not randomized | Yes | Yes | 22% | 6% for alpha/gamma analysis 4 patients without NPD | No | NR | Yes | One author (developed the CR program and executive officer of the company) |
| Popov et al., 2015; Popova et al., 2014 | Yes | Yes | No | 24% | 5% | No | NR | Yes | NR |
| Rass et al., 2012 | 10 subjects not randomized in TAU | No | Yes | 13% before the end of therapy but still assessed | 8% | No correlation between NPD and NBD | NR | Yes | NR |
| Subramaniam et al., 2012, 2014 | Yes | Yes | Yes | 3% | 3% for reality-monitoring, 6% for n-back 4 patients without follow-up | Missing data on correlation between NPD and NBD | NR | Yes | Two authors (work for the CR program company) |
| Vianin et al., 2014 | Yes | No | Yes | NR | NR | No correlation between NPD and NBD | NR | No | One author (developed the CR program) |
| Wykes et al., 2002 | Yes | No | No | NR | 5 patients without complete NPD | No correlation between NPD and NBD | NR | No | One author (developed the CR program) |
TAU: treatment as usual; NR: none reported; NPD: neuropsychological data; NBD: neurobiological data; CR: cognitive remediation.
a
Payment and blinding were assumed to be non-existent if not specified in the report or online study record.