Kohti terveempää vanhenemista: elektronisen frailtyindeksin analysointi ja kehittäminen
haavoittuvien henkilöiden tunnistamiseksi Suomen terveydenhuollossa varhaisessa
vaiheessa (FINeFI) (FINeFI)

Päärahoittaja

Rahoittajan antama koodi/diaarinumero: 349336

Päärahoittajan myöntämä tuki (€)

Rahoitusohjelma

Hankkeen aikataulu

Hankkeen aloituspäivämäärä: 01.09.2022

Hankkeen päättymispäivämäärä: 31.08.2026

Tiivistelmä

In the era of population aging, finding effective ways to promote healthy aging is a key priority. The degree of frailty is a critical determinant of healthy aging and a predictor of various adverse outcomes, such as mortality, disability and high healthcare use. Ways to assess frailty in routine healthcare practice are however currently limited. This multidisciplinary project aims to develop and validate an electronic frailty index (eFI), a bespoke tool for Finnish healthcare to identify vulnerable at-risk individuals at an early stage, to guide decision-making and aid in preoperative screening and planning proactive care pathways. The eFI allows tracking of health and functioning not only at an individual level but also in populations across different geographical and social areas. We will base the work on the Rockwood frailty index (FI) framework, one of the principal models of frailty used in research. In constructing the eFI, we will first develop and validate the eFI using Finnish nationally representative cohorts (aged >50) that have been linked to health and social care registers. In these cohorts, we have the Rockwood FI and all necessary outcome information available for validation and comparative performance assessment. We will use healthcare use, length of hospital stay, readmissions, in-hospital mortality and long-term all-cause mortality as the outcomes. We will next create an analogous eFI in routinely collected electronic health records (EHRs) in two large healthcare districts in Finland, covering both primary and secondary care. Artificial intelligence will be an integral part of the analytical work. We will device machine learning algorithms (i) alongside conventional methods in identifying the eFI items that yield the highest predictive accuracy and generalizability, and (ii) in identifying eFI items in unstructured data, i.e., free-text clinical notes. The latter approach enables the application of the eFI widely to different EHRs as healthcare providers typically document various aspects of frailty in clinical notes but not as structured, quantitative records. Certain eFI items, such as diagnoses and laboratory tests are however always available in structured data, and combing these items with the unstructured data provides a novel way in assessing frailty in Finnish EHRs. We expect the results of this project to take us one step further in promoting healthy aging in the adult population.

Vastuullinen johtaja

Muut hankkeeseen liittyvät henkilöt (JYU)

Päävastuullinen yksikkö

Seurantakohteet

Liittyvät julkaisut ja muut tuotokset