Development and testing of an electronic frailty index using Canadian electronic medical record data in primary care

BMC Primary Care Thandi et al. BMC Primary Care (2025) 26:359 https://doi.org/10.1186/s12875-025-03075-7 Open Access RESEARCH Development and testing of an electronic frailty index using Canadian electronic medical record data in primary care Manpreet Thandi1*, Andy Gibb2, Morgan Price3, Jennifer Baumbusch1 and Sabrina T. Wong1,2 Abstract Background Frailty is a state of increased vulnerability from physical, social, and cognitive factors and can result in several negative health outcomes at an individual and systemic level. Existing electronic medical record (EMR) data can be optimized to identify patients’ frailty level in primary care to facilitate early intervention and management of frailty in an efficient manner. The purpose of this work was to develop and validate a Canadian electronic frailty index (eFI) using primary care EMR data. Methods We built a Canadian eFI based on the existing UK 36-factor eFI and tested it using EMR data from British Columbia (BC) primary care practices. We used a retrospective cross-sectional design to examine the concurrent criterion validity of the eFI by testing the hypotheses that increasing frailty is associated with (1) higher numbers of primary care visits, (2) increased presence of polypharmacy, and (3) increased presence of cognitive impairment. Hypotheses were tested using Poisson and Logistic regression modelling. The data source for analysis was the BC-Canadian Primary Care Sentinel Surveillance Network. Results Our frailty algorithm was successful in its ability to calculate frailty scores for patients. A total of 15,178 patients met eligibility criteria from 22 primary care practices and 108 care providers. Ages ranged from 65 to 109 (mean 75.7); 54.2% were females. The number of frailty factors detected for patients ranged from 0 to 28 (mean 7.1). Analyses showed significant associations (p < 0.0001) between frailty levels and increasing age, material deprivation, and social deprivation. There were significant associations (p < 0.0001) between increasing frailty scores and our three outcomes. Individuals who were severely frail had nine more annual primary care visits, nine times the odds of concurrent polypharmacy, and approximately double the odds of cognitive impairment than someone who was not frail. Conclusions Our study provides evidence for initial implementation of the eFI in primary care. There is significant potential for EMR data to facilitate early detection of frailty and drive care planning with healthcare teams. Integrating the eFI within primary care provides a tremendous opportunity to screen and manage frailty with the long-term goal of reducing negative patient health outcomes and often unnecessary healthcare costs. Keywords Frailty, Frailty screening, Aging, Primary care, Polypharmacy, Cognitive impairment *Correspondence: Manpreet Thandi 1 School of Nursing, University of British Columbia, T201 2211 Wesbrook Mall, Vancouver, BC V6T 2B5, Canada 2 Centre for Health Services and Policy Research, University of British Columbia, 201-2206 East Mall, Vancouver, BC V6T 1Z3, Canada 3 Department of Family Practice, University of British Columbia, David Strangway Building, Suite 300 5950 University Boulevard, Vancouver, BC V6T 1Z3, Canada © The Author(s) 2025. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. Thandi et al. BMC Primary Care (2025) 26:359 Background Frailty, a state of increased vulnerability from physical, social, and cognitive factors, often results in a greater risk of negative health-related individual and system outcomes [1–4]. Frailty leads to functional decline, decreased quality of life, and loss of independence; systemically, frailty results in increased rates of hospitalizations, long-term care admissions, and premature mortality [2, 4–7]. Moreover, Lavado et al. [8] report an additional healthcare cost of 2.25-fold for frail compared to non-frail patients. Some of these additional costs related to frailty may be avoidable by reversing the condition with appropriate and timely interventions [9–12]. Assessing for frailty early in patients’ health trajectories in primary care settings has the potential to prevent longer term negative health outcomes. However, despite multiple existing frailty identification tools [13], time constraints and competing demands of practice are two of the most significant reasons for limited frailty assessments in primary care [14, 15]. An efficient and consistent approach is necessary to identify individuals who are frail or at risk of frailty [16]. With 93% of primary care practices in Canada [17] and 88% in the USA [18] now using electronic medical records (EMRs), and the uptick in use of digital health, there is high potential to use EMR data in primary care to identify frailty early and work with patients to maintain or improve their functional state. Past work [19] suggests that an automated frailty screening tool can use existing EMR data to calculate frailty scores and support further clinical decision-making. The UK 36-factor electronic frailty index (eFI) [19] was originally developed to address some of the shortcomings of existing frailty assessments. It is based on the accumulation of deficits and calculates frailty scores for patients using routinely collected data that already exist in patients’ EMRs. The eFI is used across 99% of UK primary care settings, has been standardized as part of nationwide guidelines for elderly care across the National Health Service (NHS) [20], and is validated by several authors [21–26], demonstrating its usefulness and applicability in practice. A key strength of the eFI is that clinical terminologies representing frailty deficits can be mapped to international contexts. The purpose of our work was to develop and validate a Canadian electronic frailty index using primary care EMR data. Our specific research question was: Does the criterion validity of the eFI demonstrate evidence to support its implementation in British Columbia (BC) primary care settings? We tested three hypotheses to examine the criterion validity of our eFI: Page 2 of 13 1. Higher eFI scores are associated with higher numbers of primary care v (...truncated)