eCite Digital Repository

Utility of the hospital frailty risk score derived from administrative data and the association with stroke outcomes

Citation

Kilkenny, MF and Phan, HT and Lindley, RI and Kim, J and Lopez, D and Dalli, LL and Grimley, R and Sundararajan, V and Thrift, AG and Andrew, NE and Donnan, GA and Cadilhac, DA, Stroke123 Investigators and the AuSCR Consortium, Utility of the hospital frailty risk score derived from administrative data and the association with stroke outcomes, Stroke, 52, (9) pp. 2874-2881. ISSN 0039-2499 (2021) [Refereed Article]


Preview
PDF
Not available
215Kb
  

DOI: doi:10.1161/STROKEAHA.120.033648

Abstract

Background and purpose: Conditions associated with frailty are common in people experiencing stroke and may explain differences in outcomes. We assessed associations between a published, generic frailty risk score, derived from administrative data, and patient outcomes following stroke/transient ischemic attack; and its accuracy for stroke in predicting mortality compared with other measures of clinical status using coded data.

Methods: Patient-level data from the Australian Stroke Clinical Registry (20092013) were linked with hospital admissions data. We used International Statistical Classification of Diseases and Related Health Problems, Tenth Revision codes with a 5-year look-back period to calculate the Hospital Frailty Risk Score (termed Frailty Score hereafter) and summarized results into 4 groups: no-risk (0), low-risk (15), intermediate-risk (515), and high-risk (>15). Multilevel models, accounting for hospital clustering, were used to assess associations between the Frailty Score and outcomes, including mortality (Cox regression) and readmissions up to 90 days, prolonged acute length of stay (>20 days; logistic regression), and health-related quality of life at 90 to 180 days (quantile regression). The performance of the Frailty Score was then compared with the Charlson and Elixhauser Indices using multiple tests (eg, C statistics) for predicting 30-day mortality. Models were adjusted for covariates including sociodemographics and stroke-related factors.

Results: Among 15 468 adult patients, 15% died ≤90 days. The frailty scores were 9% no risk; 23% low, 45% intermediate, and 22% high. A 1-point increase in frailty (continuous variable) was associated with greater length of stay (ORadjusted, 1.05 [95% CI, 1.04 to 1.06), 90-day mortality (HRadjusted, 1.04 [95% CI, 1.03 to 1.05]), readmissions (ORadjusted, 1.02 [95% CI, 1.02 to 1.03]; and worse health-related quality of life (median difference, −0.010 [95% CI −0.012 to −0.010]). Adjusting for the Frailty Score provided a slightly better explanation of 30-day mortality (eg, larger C statistics) compared with other indices.

Conclusions: Greater frailty was associated with worse outcomes following stroke/transient ischemic attack. The Frailty Score provides equivalent precision compared with the Charlson and Elixhauser indices for assessing risk-adjusted outcomes following stroke/transient ischemic attack.

Item Details

Item Type:Refereed Article
Keywords:hospitalization, ischemic attack, transient, mortality, register, risk factor
Research Division:Biomedical and Clinical Sciences
Research Group:Neurosciences
Research Field:Neurology and neuromuscular diseases
Objective Division:Health
Objective Group:Evaluation of health and support services
Objective Field:Evaluation of health outcomes
UTAS Author:Phan, HT (Dr Hoang Phan)
ID Code:151624
Year Published:2021
Web of Science® Times Cited:7
Deposited By:Menzies Institute for Medical Research
Deposited On:2022-08-02
Last Modified:2022-08-02
Downloads:0

Repository Staff Only: item control page