From cognitive screening to digital phenotyping: rethinking early detection of cognitive impairment in primary care.

Early detection of cognitive impairment has become a clinical and public-health priority in the therapeutic era of Alzheimer's disease (AD). Brief instruments such as the Mini-Mental State Examination, Montreal Cognitive Assessment, Mini-Cog, Clock Drawing Test, Fototest, Memory Alteration Test, Eurotest, and AD8 remain widely used because they are inexpensive, clinically interpretable, and feasible in routine care. However, most were developed to identify established cognitive impairment rather than subtle mild cognitive impairment or biologically defined early AD. Ceiling effects, educational and cultural bias, examiner variability, and limited integration with functional, neuropsychiatric, subjective, and biomarker data restrict their value as stand-alone tools in contemporary diagnostic pathways. Digital cognitive assessment may address selected limitations of conventional screening by standardizing administration, reducing scoring variability, enabling repeated measurement, and capturing process-level features such as response latency, intra-individual variability, learning effects, speech and language markers, graphomotor dynamics, gaze, and ecologically sampled behavior. These signals may support earlier risk stratification and longitudinal monitoring, but they do not resolve diagnostic uncertainty or establish AD etiology on their own. This narrative review examines the transition from traditional cognitive screening to digital cognitive phenotyping. It considers established brief and contextual instruments, digitized conventional tests, remote repeated assessments, speech and language-derived digital cognitive biomarkers, digital clock drawing, prospective-memory tools, virtual reality and serious games, oculomotor and graphomotor metrics, passive sensing, multimodal platforms, and their integration with structural MRI, blood-based biomarkers, cerebrospinal fluid markers, and amyloid/tau PET. In the context of evolving AD criteria and disease-modifying therapies, digital screening should be understood as a governed triage and phenotyping layer rather than a stand-alone diagnostic label. A staged pathway is proposed that combines analog cognitive instruments, informant and functional measures, neuropsychiatric assessment, digital signals, and biological markers to support earlier, more equitable, and clinically actionable detection of cognitive impairment.