Background: Evidence suggests that atypical antipsychotics (AAPs) exert a short-term mortality risk in people with dementia. We assessed whether additional randomized clinical trials influence the current evidence and the potential effect modifiers.
Methods: Electronic databases were systematically searched for randomized controlled trials from their inception through March 2018. A random-effects model was used for analysis. Potential effect modifiers were examined through meta-regression. Trial sequential analysis was performed to quantify the statistical reliability of data in the cumulative meta-analysis with adjustment of significance levels for sparse data and repetitive testing on accumulating data. Certainty of evidence and risk of bias were also evaluated.
Results: We found that compared with placebos, AAPs may increase the risk of mortality (odds ratio [OR], 1.536; 95% confidence intervals [CIs], 1.028-2.296; P = 0.036, high certainty). In the subgroup analysis, the estimated ORs were the highest for olanzapine (1.919; P = 0.232), followed by those for quetiapine (1.663; P = 0.506), aripiprazole (1.649; P = 0.297), and risperidone (1.354; P = 0.277); however, the mortality risk presented by individual AAPs did not exhibit between-group differences. The meta-regression did not identify any effect modifiers, including the chlorpromazine equivalent dose, trial duration, and cognitive status. The trial sequential analysis revealed that future similar trials are unlikely to alter our findings.
Conclusions: Atypical antipsychotics are associated with increased short-term mortality risk, although a disease-drug interaction may contribute to such risk in people with dementia. Patients with dementia may still benefit by AAPs after appropriate assessment of the disease severity as well as the dosage of AAPs, treatment duration, and monitoring of AAPs.