Table 1.
The studies’ main finding in relation to the severity of AD and its likelihood for bias
| Type of study | Study authors | Main results | Disease severity | Bias risk assessment |
|---|---|---|---|---|
| Blinded RCTs | Tariot et al. [13], 2004 | statistically significant difference in favor of combination therapy for cognition and function | moderate to severe | undetermined(ROB) |
| Porsteinsson et al. [20], 2008 | no statistically significant difference between groups | mild to moderate | undetermined(ROB) | |
| Howard et al. [21], 2012 | for patients previously taking donepezil: continuation with donepezil was significantly better than placebo; continuation with memantine was significantly better than placebo; the interaction of donepezil with memantine was not significant | moderate to severe | low (ROB) | |
| Open-label trials | Dantoine et al. [14], 2006 | improvement in behavior, cognition, and function in favor of combination therapy | moderate to severe | high (ROB) |
| Riepe et al. [24], 2007 | tolerable and statistically significant improvement in cognition in favor of combination therapy | mild to moderate (more impaired appeared to benefit more) | high (ROB) | |
| Olin et al. [25], 2010 | combination therapy is tolerable and safe, and associated with modest changes in cognition and function | moderate | high (ROB) | |
| Shua-Haim et al. [26], 2008 | combination considered well tolerated | mild to moderate | high (ROB) | |
| Farlow et al. [22], 2010 | no significant difference | mild to moderate | high (ROB) | |
| Choi et al. [23], 2011 | no significant difference (except in CMAI-K) | moderate | high (ROB) | |
| Cohort studies | Hartmann and Mobius [29], 2003 | in favor of combination | not reported | 1 star (NOS) |
| Atri et al. [12], 2008 | statistically significant difference in favor of combination therapy for cognition and function | varying severity | 6 stars (NOS) | |
| Lopez et al. [28], 2009 | statistically significant fewer NH admissions | not reported, but mean MMSE 17.4 ± 5.6 | 7 stars (NOS) | |
| Schneider et al. [27], 2011 | statistically significant difference in favor of monotherapy for cognition and function | mild AD | 7 stars (NOS) | |
ROB = Cochrane risk of bias tool; NOS = Newcastle-Ottawa scale (smaller values correspond to higher risks of bias); CMAI-K = Korean version of the Cohen-Mansfield Agitation Inventory; NH = nursing home.