Diagnosis accuracy of touchscreen-based testings for major neurocognitive disorders: a systematic review and meta-analysis.

Facing an increasing prevalence, diagnosis and management of dementia has become a global health challenge. While most cognitive assessments are based on paper tests, the current trend is to digitise them, using new technologies. We aimed to achieve a systematic review on touchscreen-based assessments for neurocognitive disorders. A search on four databases (PubMed, Embase, IEEE Xplore and Web of Science) was performed. Two investigators independently selected the articles and assessed their quality using the QUADAS-2 tool. We included articles whose participants were over 65 years, classified according to the presence/absence of major neurocognitive disorder (M-NCD) determined by conventional assessment of cognition, examined with a novel tool using a touchscreen device (tablet or smartphone). We finally retained 35 articles for the systematic review and 24 for the meta-analysis. Of the 35 articles included in the systematic review, participants had Alzheimer's disease, Parkinson's disease, vascular dementia, Lewy body disease or any type of dementia. Pooled sensitivity and specificity were 0.89 (95% CI: 0.86-0.91) and 0.88 (95% CI: 0.82-0.92), respectively. Performances of self-administered or brief assessment testings were similar to those of assessor-administered or longer duration testings. The major limitation of this review and meta-analysis is the multiplicity and diversity of methods (tools, cognition assessment and type of M-NCD), which make the comparison difficult. We conclude that brief and self-assessment touchscreen-based cognitive tests are appropriate and simple tools to screen major cognitive disorders that can be used in primary care. The study was registered in Prospero (CRD42022358725).