Systematic Review and Meta-Analysis of Prevalence in Post-Stroke Neurocognitive Disorders in Hospital-Based Studies.

Barbay, Mélanie; Diouf, Momar; Roussel, Martine; Godefroy, Olivier

BACKGROUND/AIMS Post-stroke neurocognitive disorders (post-stroke NCD) have been reported with a very variable prevalence. METHODS Based on a systematic literature search, hospital-based studies published between January 1990 and September 2015 were selected when they reported the prevalence of total, mild, and major post-stroke NCD diagnosed by using specified criteria. Factors affecting prevalence were assessed using meta-regression analysis. RESULTS Among the 7,440 references evaluated, 16 hospital-based studies were selected, corresponding to a total of 3,087 patients. The overall prevalence of total post-stroke NCD was 53.4% (95% CI: 46.9-59.8): 36.4% for mild post-stroke NCD (95% CI: 29-43.8) and 16.5% (95% CI: 12.1-20.8) for major post-stroke NCD. The overall prevalence was mainly influenced by the threshold score used for categorization (p = 0.0001) and, in the subgroup of studies using a conservative threshold (i.e., ≤7th percentile), by the recurrent stroke rate (p = 0.0005). The prevalence of major post-stroke NCD was mainly influenced by age (p = 0.003). CONCLUSION More than half of stroke survivors experience post-stroke NCD, corresponding to mild post-stroke NCD in two-thirds of cases and major post-stroke NCD in one-third of cases. Harmonization of stroke assessment and cognitive score thresholds is urgently needed to allow more accurate estimation of post-stroke NCD prevalence, especially mild post-stroke NCD.