Phone-based frailty screening matches clinical scale accuracy without internet or in-person assessment

A South Korean study developed and validated a rule-based frailty classification algorithm delivered entirely by automated telephone call (interactive voice response, IVR), requiring no internet access or human contact Primary Study. Across two rural cohorts (n=154 with full geriatric assessment; n=1,532 with comparator scoring) and an external validation cohort of adults aged 90 and over recruited across five urban and rural sites, IVR-based frailty classification showed strong, graded associations with a validated 45-item Frailty Index, disability, gait speed and health-related quality of life.

The tool’s discriminative accuracy for detecting frailty was comparable to the Clinical Frailty Scale, the benchmark in widespread UK and international use: area under the curve of 0.77 for the IVR tool versus 0.76 for the Clinical Frailty Scale, a difference the authors report as not statistically significant. External validation in the oldest-old cohort supported generalisability beyond the original study population.

The study was conducted entirely within the South Korean phone system and healthcare context; the authors do not test, and this evidence does not establish, how the tool would perform if adapted to a UK telephony and NHS service context.

For services seeking to extend frailty case-finding to populations who are hard to reach through in-person or digital-first channels, this offers a proof of concept for a scalable, low-cost screening modality — with the important caveat that UK-context validation, including consideration of how it would perform for landline-only, hearing-impaired or non-native-language populations, would be needed before any adoption.