Frailty Trends in an Older Veteran Subpopulation 1 Year Prior and Into the COVID-19 Pandemic Using CAN Scores
Background: We studied the effects of the first year of the COVID-19 pandemic on frailty trends in a subset of older veterans at the Phoenix Veterans Affairs Health Care System.
Methods: We identified 3538 and 6103 veterans aged 70 to 75 years as of February 8, 2019, with a calculated Care Assessment Need (CAN) score of ≥ 75 for 1-year mortality and hospitalization, respectively. After excluding veterans with insufficient 2020 and 2021 data, we compared the difference in 1-year mortality and hospitalization CAN scores from 2019 to 2020 with 2020 to 2021 using a paired t test.
Results: The difference in mean (SD) 1-year mortality CAN scores from 2020 to 2021 was 0.2 (13.4) when compared with the previous year’s -4.9 (12.5) (P < .0001), indicating increased frailty. The difference in 1-year hospitalization CAN scores from 2020 to 2021 was -1.5 (12.0) when compared with the previous year’s -2.8 (9.9) (P < .0001).
Conclusions: Frailty in our veteran subpopulation as calculated by 1-year mortality CAN scores increased in the first year of the COVID-19 pandemic when compared with a recovering trend the previous year.
In the hospitalization group, there were 6046 veterans in the analysis; 57 veterans missing a 1-year hospitalization CAN score that were excluded. The mean age was 71.7 (1.3) years and included 5874 male (97.2%) and 172 female (2.8%) veterans. There was a decline in mean 1-year hospitalization CAN scores in our subset of frail older veterans by 2.8 (95% CI, -3.1 to -2.6) in the year preceding the COVID-19 pandemic. This mean decline slowed significantly to 1.5 (95% CI, -1.8 to -1.2; P < .0001) after the first year of the COVID-19 pandemic. Mean CAN scores for 1-year hospitalization were 84.6 (95% CI, 84.4 to 84.8), 81.8 (95% CI, 81.5 to 82.1), and 80.2 (95% CI, 79.9 to 80.6)
We also calculated the number of veterans with increasing, stable, and decreasing CAN scores across each of our defined periods in both the 1-year mortality and hospitalization groups. 
A previous study used a 1-year combined hospitalization or mortality event CAN score as the most all-inclusive measure of frailty but determined that it was possible that 1 of the other 5 CAN risk measures could perform better in predicting frailty.10 We collected and presented data for 1-year mortality and hospitalization CAN scores. There were declines in pandemic-related US hospitalizations for illnesses not related to COVID-19 during the first few months of the pandemic.18 This may or may not have affected the 1-year hospitalization CAN score data; thus, we used the 1-year mortality CAN score data to predict frailty.
Discussion
We studied frailty trends in an older veteran subpopulation enrolled at the PVAHCS 1 year prior and into the COVID-19 pandemic using CAN scores. Frailty is a dynamic state. Previous frailty assessments aimed to identify patients at the highest risk of death. With the advent of advanced therapeutics for several diseases, the number of medical conditions that are now managed as chronic illnesses continues to grow. There is a role for repeated measures of frailty to try to identify frailty trends.19 These trends may assist us in resource allocation, identifying interventions that work and those that do not.
Some studies have shown an overall declining lethality of frailty. This may reflect improvements in the care and management of chronic conditions, screening tests, and increased awareness of healthy lifestyles.20 Another study of frailty trajectories in a veteran population in the 5 years preceding death showed multiple trajectories (stable, gradually increasing, rapidly increasing, and recovering).19
The PACT is a primary care model implemented at VA medical centers in April 2010. It is a patient-centered medical home model (PCMH) with several components. The VA treats a population of socioeconomically vulnerable patients with complex chronic illness management needs. Some of the components of a PACT model relevant to our study include facilitated self-management support for veterans in between practitioner visits via care partners, peer-to-peer and transitional care programs, physical activity and diet programs, primary care mental health, integration between primary and specialty care, and telehealth.21 A previous study has shown that VA primary care clinics with the most PCMH components in place had greater improvements in several chronic disease quality measures than in clinics with a lower number of PCMH components.22
