Electronic Frailty Index Identifies At-Risk Older Adults

TOPLINE:

An electronic frailty index, derived from electronic health record (EHR) data including diagnostic codes, identified patients who had a higher risk for mortality, acute care visits, and readmissions.

METHODOLOGY:

• Researchers conducted a retrospective cohort study using EHR data from a regional healthcare system in the United States to develop an electronic frailty index and evaluate its performance in older adults.
• They included 518,449 patients (mean age, 71.94 years; 42.9% men; 82% White individuals) who had two outpatient encounters in the 3 years prior or one encounter in the 2 years prior to 2017. A subcohort of 8402 patients from two large primary care clinics was used to validate the performance of the electronic frailty index.
• Diagnostic and procedure codes, retrieved from both inpatient and outpatient encounters over a 2-year period prior to the index date, were used to identify 31 age-related health deficits across various domains.
• The index was calculated as the total number of deficits divided by 31; frailty status was categorized as robust (< 0.1), prefrail (0.1-0.2), frail (0.2-0.3), and very frail (> 0.3).
• The primary outcome was time to death; secondary outcomes included time to acute care visits and hospital readmissions within 90 days.

TAKEAWAY:

• Overall, 72.9% of the cohort were identified as robust, 15.8% as prefrail, 6.9% as frail, and 2.8% as very frail.
• Mortality rates were significantly higher among prefrail (adjusted hazard ratio [aHR], 1.62), frail (aHR, 2.45), and very frail (aHR, 4.14) patients than robust individuals (P < .001 for all).
• Very frail individuals had significantly higher rates of acute care visits (aHR, 5.52) and 90-day readmissions (aHR, 2.11) than robust individuals (P < .001 for both).
• The electronic frailty index demonstrated similar associations with outcomes in the primary care subcohort, with very frail individuals having a higher risk for mortality (aHR, 7.91), acute care utilization (aHR, 4.41), and readmissions (aHR, 2.63) than nonfrail individuals (P < .001 for all).

IN PRACTICE:

“The MGB-eFI [Mass General Brigham-electronic frailty index] demonstrates that measuring frailty even without complete primary care encounter data at the population-level can stratify patients with varying risk of adverse outcomes,” the authors wrote. “Other health systems may also benefit this approach to identify older adults at risk of outcomes that adversely impact overall health and well-being.”

SOURCE:

This study was led by Bharati Kochar, MD, MS, a physician at the Massachusetts General Hospital in Boston. It was published online on February 21, 2025, in Journal of the American Geriatrics Society.

LIMITATIONS:

The index relied on passively collected data. The focus on outpatient encounters may have excluded patients with inpatient data only. Additionally, the absence of detailed socioeconomic data and underrepresentation of minority communities may limit the generalizability of the findings to diverse populations.

DISCLOSURES:

This study was supported by grants from the National Institute on Aging, Päivikki and Sakari Sohlberg Foundation, and Center for Integrated Healthcare of the US Department of Veterans Affairs. One author reported receiving consulting fees from two sources. Another author reported a change in their affiliation.

This article was created using several editorial tools, including AI, as part of the process. Human editors reviewed this content before publication.