Development and validation of a frailty index derived from routinely
COLLECTED AND ELECTRONICALLY RECORDED HEALTH INFORMATION
Authors List Aickin, H., Waitematā District Health Board, Auckland, New Zealand (presenter) Bloomfield, K., Waitematā District Health Board & University of Auckland, Auckland, New Zealand Wu Z., University of Auckland, Auckland, New Zealand Connolly, MJ., Waitematā District Health Board & University of Auckland, Auckland, New Zealand
Introduction: Frailty is associated with adverse outcomes and increasing dependency. Rapid detection of frailty from routinely-collected electronic data is desirable. There are no studies utilising electronic data to assess frailty in New Zealand (NZ) inpatients.
Aim: Derive and validate a frailty index (FI) from routinely-collected electronic data in hospitalised older adults to predict adverse health outcomes.
Methods: A retrospective cohort of inpatients aged >65 years at Waitematā District Health Board (WDHB), NZ, with electronically recorded Global Geriatric Assessment (GGA) collected between May-2018 and October-2018. A FI was derived, applied to each participant then outcomes determined by level of frailty within 1 year of follow-up. The primary outcome was 6-month re-hospitalisation rate. Secondary outcomes were mortality and entrance into Residential Aged Care (RAC) at 1 year. T-tests determined difference of mean FIs among different outcomes. Chi-square tests assessed the relationship between outcomes and pre-specified Frailty Score (FS) groups (0-2, 3, 4, 5).
Results: The sample comprised 162 consecutive GGAs. Participants had a mean (SD) age of 86 (8.2) years, 108 (66.7%) were female and 143 (88%) identified as European/Pākehā. The mean (SD) FI was 0.42 (0.12). The 6-month hospitalisation rate for the lowest frailty group (FS 0-2) and highest frailty group (FS 5) was 35.7% and 66.7% respectively (p=0.04). The one-year mortality of FS 0-2 and FS 5 was 3.6% and 28.9% respectively (p=0.008). The one-year RAC admission of FS 0-2 and FS 5 was 14.3% and 54.1% respectively (p=0.001).
Conclusions: This study derived and operationalised a FI using routinely-collected electronic data from hospitalised older adults in WDHB. Higher FI scores were associated with higher risk of 6-month hospitalisation, one-year mortality, and one-year RAC admission. This is a novel application of the FI in NZ. This tool improves clinical identification of frailty and has significant potential for research purposes.
The ANZSGM Retreat is proudly bought to you by The Australian & New Zealand Society for Geriatric Medicine