What can we learn from 68 000 clinical frailty scale scores? Evaluating the utility of frailty assessment in emergency departments

Hugh Logan Ellis; Liam Dunnell; Ruth Eyres; Julie Whitney; Cara Jennings; Dan Wilson; Jane Tippett; Dan F. Stein; James Teo; Zina Ibrahim; Kenneth Rockwood

doi:10.1093/ageing/afaf093

What can we learn from 68 000 clinical frailty scale scores? Evaluating the utility of frailty assessment in emergency departments

Hugh Logan Ellis^*, Liam Dunnell, Ruth Eyres, Julie Whitney, Cara Jennings, Dan Wilson, Jane Tippett, Dan F. Stein, James Teo, Zina Ibrahim, Kenneth Rockwood

^*Corresponding author for this work

Research output: Contribution to journal › Article › peer-review

1 Citation (Scopus)

Abstract

Background: Emergency departments (EDs) in England are under significant strain, with increasing attendances and extended wait times, affecting frail older adults. The clinical frailty scale (CFS) has been implemented as a tool to assess frailty in ED settings, but its reliability and predictive accuracy as a screening tool remain debated. Objective: To evaluate the use and variability of the CFS in EDs and its association with patient outcomes, including discharge rates, length of stay, readmission and mortality. Methods: A retrospective cohort study of ED attendances at two London (UK) hospitals from 2017 to 2021. Data included CFS scores, demographics, clinical observations and outcomes. Comparative statistics, logistic regression, Cox proportional hazards models and competing risk regression were applied to examine CFS predictive validity. Results: In a sample of 123 324 ED visits, CFS scores strongly correlated with adverse outcomes: e.g. for long-term mortality (n = 33 475, events = 8871), each CFS single-point increase was associated with a 25% increase in mortality risk (95% CI 1.23–1.26). CFS scores varied significantly between raters and across visits, median difference two levels (interquartile range 1–3). Intraclass correlation coefficient analysis showed that 33.1% of CFS score differences was attributable to between-patient differences, 15.4% to inter-rater differences, with 51.5% residual variance from non-frailty factors, such as acute illness severity. Conclusion: The CFS is associated with crucial patient outcomes in the ED. Inter-rater variability and potentially confounding factors can limit its consistency. Automation to enhance CFS score reliability should be explored as a means to support proactive management.

Original language	English
Article number	afaf093
Journal	Age and Ageing
Volume	54
Issue number	4
DOIs	https://doi.org/10.1093/ageing/afaf093
Publication status	Published - 1 Apr 2025

Keywords

clinical frailty scale (CFS)
electronic health records (EHRs)
emergency department (ED)
frailty assessment
older people

Access to Document

10.1093/ageing/afaf093

Cite this

@article{3782334747954264b34b095e1abe7ce3,

title = "What can we learn from 68 000 clinical frailty scale scores? Evaluating the utility of frailty assessment in emergency departments",

abstract = "Background: Emergency departments (EDs) in England are under significant strain, with increasing attendances and extended wait times, affecting frail older adults. The clinical frailty scale (CFS) has been implemented as a tool to assess frailty in ED settings, but its reliability and predictive accuracy as a screening tool remain debated. Objective: To evaluate the use and variability of the CFS in EDs and its association with patient outcomes, including discharge rates, length of stay, readmission and mortality. Methods: A retrospective cohort study of ED attendances at two London (UK) hospitals from 2017 to 2021. Data included CFS scores, demographics, clinical observations and outcomes. Comparative statistics, logistic regression, Cox proportional hazards models and competing risk regression were applied to examine CFS predictive validity. Results: In a sample of 123 324 ED visits, CFS scores strongly correlated with adverse outcomes: e.g. for long-term mortality (n = 33 475, events = 8871), each CFS single-point increase was associated with a 25% increase in mortality risk (95% CI 1.23–1.26). CFS scores varied significantly between raters and across visits, median difference two levels (interquartile range 1–3). Intraclass correlation coefficient analysis showed that 33.1% of CFS score differences was attributable to between-patient differences, 15.4% to inter-rater differences, with 51.5% residual variance from non-frailty factors, such as acute illness severity. Conclusion: The CFS is associated with crucial patient outcomes in the ED. Inter-rater variability and potentially confounding factors can limit its consistency. Automation to enhance CFS score reliability should be explored as a means to support proactive management.",

keywords = "clinical frailty scale (CFS), electronic health records (EHRs), emergency department (ED), frailty assessment, older people",

author = "Ellis, {Hugh Logan} and Liam Dunnell and Ruth Eyres and Julie Whitney and Cara Jennings and Dan Wilson and Jane Tippett and Stein, {Dan F.} and James Teo and Zina Ibrahim and Kenneth Rockwood",

note = "Publisher Copyright: {\textcopyright} The Author(s) 2025.",

year = "2025",

month = apr,

day = "1",

doi = "10.1093/ageing/afaf093",

language = "English",

volume = "54",

journal = "Age and Ageing",

issn = "0002-0729",

publisher = "Oxford University Press",

number = "4",

}

TY - JOUR

T1 - What can we learn from 68 000 clinical frailty scale scores? Evaluating the utility of frailty assessment in emergency departments

AU - Ellis, Hugh Logan

AU - Dunnell, Liam

AU - Eyres, Ruth

AU - Whitney, Julie

AU - Jennings, Cara

AU - Wilson, Dan

AU - Tippett, Jane

AU - Stein, Dan F.

AU - Teo, James

AU - Ibrahim, Zina

AU - Rockwood, Kenneth

N1 - Publisher Copyright: © The Author(s) 2025.

PY - 2025/4/1

Y1 - 2025/4/1

N2 - Background: Emergency departments (EDs) in England are under significant strain, with increasing attendances and extended wait times, affecting frail older adults. The clinical frailty scale (CFS) has been implemented as a tool to assess frailty in ED settings, but its reliability and predictive accuracy as a screening tool remain debated. Objective: To evaluate the use and variability of the CFS in EDs and its association with patient outcomes, including discharge rates, length of stay, readmission and mortality. Methods: A retrospective cohort study of ED attendances at two London (UK) hospitals from 2017 to 2021. Data included CFS scores, demographics, clinical observations and outcomes. Comparative statistics, logistic regression, Cox proportional hazards models and competing risk regression were applied to examine CFS predictive validity. Results: In a sample of 123 324 ED visits, CFS scores strongly correlated with adverse outcomes: e.g. for long-term mortality (n = 33 475, events = 8871), each CFS single-point increase was associated with a 25% increase in mortality risk (95% CI 1.23–1.26). CFS scores varied significantly between raters and across visits, median difference two levels (interquartile range 1–3). Intraclass correlation coefficient analysis showed that 33.1% of CFS score differences was attributable to between-patient differences, 15.4% to inter-rater differences, with 51.5% residual variance from non-frailty factors, such as acute illness severity. Conclusion: The CFS is associated with crucial patient outcomes in the ED. Inter-rater variability and potentially confounding factors can limit its consistency. Automation to enhance CFS score reliability should be explored as a means to support proactive management.

AB - Background: Emergency departments (EDs) in England are under significant strain, with increasing attendances and extended wait times, affecting frail older adults. The clinical frailty scale (CFS) has been implemented as a tool to assess frailty in ED settings, but its reliability and predictive accuracy as a screening tool remain debated. Objective: To evaluate the use and variability of the CFS in EDs and its association with patient outcomes, including discharge rates, length of stay, readmission and mortality. Methods: A retrospective cohort study of ED attendances at two London (UK) hospitals from 2017 to 2021. Data included CFS scores, demographics, clinical observations and outcomes. Comparative statistics, logistic regression, Cox proportional hazards models and competing risk regression were applied to examine CFS predictive validity. Results: In a sample of 123 324 ED visits, CFS scores strongly correlated with adverse outcomes: e.g. for long-term mortality (n = 33 475, events = 8871), each CFS single-point increase was associated with a 25% increase in mortality risk (95% CI 1.23–1.26). CFS scores varied significantly between raters and across visits, median difference two levels (interquartile range 1–3). Intraclass correlation coefficient analysis showed that 33.1% of CFS score differences was attributable to between-patient differences, 15.4% to inter-rater differences, with 51.5% residual variance from non-frailty factors, such as acute illness severity. Conclusion: The CFS is associated with crucial patient outcomes in the ED. Inter-rater variability and potentially confounding factors can limit its consistency. Automation to enhance CFS score reliability should be explored as a means to support proactive management.

KW - clinical frailty scale (CFS)

KW - electronic health records (EHRs)

KW - emergency department (ED)

KW - frailty assessment

KW - older people

UR - http://www.scopus.com/inward/record.url?scp=105003116956&partnerID=8YFLogxK

U2 - 10.1093/ageing/afaf093

DO - 10.1093/ageing/afaf093

M3 - Article

AN - SCOPUS:105003116956

SN - 0002-0729

VL - 54

JO - Age and Ageing

JF - Age and Ageing

IS - 4

M1 - afaf093

ER -

What can we learn from 68 000 clinical frailty scale scores? Evaluating the utility of frailty assessment in emergency departments

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this