Comparing Frailty Assessment Methods and their Ability to Predict Adverse Outcomes in Patients with Advanced Chronic Kidney Disease.

BACKGROUND

Frailty is common in patients with chronic kidney disease (CKD), and those affected by both are at increased risk of adverse outcomes including disability, hospitalization, and death. Collecting data on frailty as part of clinical care could enhance care by identifying patients at risk of adverse events. However, clinical assessment of frailty requires time and resources. Frailty definitions based on administrative data might provide an efficient alternative. The primary objective was to compare agreement between administrative claims-based definitions versus objectively measured frailty in adults with advanced, non-dialysis CKD, and to examine their associations with adverse outcomes.

METHODS

The cohort consisted of Manitoba participants from the Canadian Frailty Observation and Interventions Trial (CanFIT). This multicentre cohort study followed adults with advanced CKD longitudinally. Every visit, assessments were conducted to determine frailty status using the Fried Frailty Phenotype, Short Physical Performance Battery, and healthcare providers' impression. The CanFIT database was linked to administrative databases at the Manitoba Centre for Healthy Policy to calculate two claims-based frailty indicators, the Segal and modified pre-operative frailty indices, which have been validated in the non-CKD literature.

RESULTS

Of the 442 participants included, the mean age was 66±14 years and 58% were male; 88% had hypertension, 61% dyslipidemia, and 58% diabetes. The prevalence of frailty varied from 19% to 70% depending on definition. Agreement between frailty definitions was poor (κ 0.09-0.33); however, individuals considered frail, using both administrative or measured definitions had a higher risk of all-cause mortality and hospitalization, except for those identified by the Segal Frailty Indicator.

CONCLUSIONS

This study suggests that those identified as frail by nearly all measures were at higher risk of adverse outcomes. Thus, most frailty models in this study can be used to identify high risk advanced non-dialysis CKD populations, allowing us to target individuals for interventions that aim to improve outcomes.