Development of frailty measurement tools using administrative health data: A systematic review

Son Nghiem; Disna Sajeewani; Katrina Henderson; Clifford Afoakwah; Joshua Byrnes; Wendy Moyle; Paul Scuffham

doi:10.1016/j.archger.2020.104102

Development of frailty measurement tools using administrative health data: A systematic review

Arch Gerontol Geriatr. 2020 Jul-Aug:89:104102. doi: 10.1016/j.archger.2020.104102. Epub 2020 May 12.

Authors

Son Nghiem¹, Disna Sajeewani², Katrina Henderson³, Clifford Afoakwah², Joshua Byrnes², Wendy Moyle⁴, Paul Scuffham⁵

Affiliations

¹ Centre for Applied Health Economics, Griffith University, Queensland, Australia. Electronic address: s.nghiem@griffith.edu.au.
² Centre for Applied Health Economics, Griffith University, Queensland, Australia.
³ Health Library, Griffith University, Queensland, Australia.
⁴ Menzies Health Institute Queensland, Griffith University, Queensland, Australia; School of Nursing and Midwifery, Griffith University, Queensland, Australia.
⁵ Centre for Applied Health Economics, Griffith University, Queensland, Australia; Menzies Health Institute Queensland, Griffith University, Queensland, Australia.

PMID: 32464423
DOI: 10.1016/j.archger.2020.104102

Abstract

Objectives: Frailty is an increasingly common health condition and is seen more often due to the ageing population. This study reviews the evidence on the development and validation of these automated frailty measurement tools.

Design: Six databases: PubMed, EMBASE, MEDLINE, CINAHL, Scopus, and Web of Science were electronically searched. Selected studies must have developed and validated a new frailty measurement tool using administrative health data and published results in a peer-reviewed, English-language journal. Selected studies were synthesized narratively.

Setting and participants: The review focused on large scale studies using administrative health data in developed countries. Participants included older people aged 65 years and above.

Measures: The main measures of review studies include discrimination power and the prediction ability of adverse health outcomes; performance against established frailty measures; and validation records.

Results: Five studies were selected for narrative synthesis after screening the full-text. All frailty measurement tools in the selected five studies produced strong discrimination power with C-statistics ranging from 0.61-97. Two studies were independently validated in studies by other authors or conducted in other locations; one study developed an early prediction model, and no study has been applied in practice.

Conclusions and implications: Automated frailty measurement tools using administrative health data are still in the early development stage with five tools developed since 2016. Selected studies have strong prediction of adverse health outcomes. Future studies should include validation and refinement of these tools in other countries and assessment of their clinical utility and capacity to inform cost-effective policy and practice.

Keywords: Administrative data; Automate; Frailty; Healthcare; Measurement; Prevention.

Publication types

Research Support, Non-U.S. Gov't
Systematic Review

MeSH terms

Aged
Aged, 80 and over
Cost-Benefit Analysis
Frailty* / diagnosis
Frailty* / epidemiology
Geriatric Assessment
Humans