Frailty indices based on routinely collected data: a scoping review

A frailty index (FI) is a frailty assessment tool calculated as the proportion of the number of health-related deficits an individual has to the total number of variables in the index. Routinely collected clinical and administrative data can be used as sources of deficits to automatically calculate FIs. This scoping review aimed to evaluate the current research landscape on routine data-based FIs. We searched seven databases to find literature published in 2013–2023. Main inclusion criteria were original research articles on FIs constructed from routine data, with deficits in at least two of the following categories: “symptoms/signs”, “laboratory values”, “diseases”, “disabilities”, and “others”. From 7526 publications screened, 218 were included. Studies were primarily from North America (47.7 %), conducted in the community (35.3 %), and used routine data-based FIs for risk stratification (51.4 %). FIs were calculated using various routine data sources; however, most were initially developed and validated using hospital records. We noted geographical differences in study settings and routine data sources. We identified 611 unique deficits comprising these FIs. Most were either “diseases” (34.4 %) or “symptoms/signs” (32.1 %). Routine data-based FIs are feasible and valid risk stratification tools, but research is confined to high-income countries, their routine adoption is slow, and deficits comprising these FIs emphasise a reactive and overtly medical approach in addressing frailty. Future directions include exploring the feasibility and applicability of using routine databases for frailty assessment in lower- and middle-income countries, and leveraging non-clinical routine data through data linkages to proactively identify and manage frailty.