Using a Claims-Based Frailty Index to Investigate Frailty, Survival, and Healthcare Expenditures among Older Adults Hospitalized for COVID-19 at an Academic Medical Center

Background Frailty is associated with mortality in older adults hospitalized with COVID-19, yet few studies have quantified healthcare utilization and spending following COVID-19 hospitalization. Objective To evaluate whether survival and follow-up healthcare utilization and expenditures varied as a function of claims-based frailty status for older adults hospitalized with COVID-19. Design Retrospective cohort study Participants 136 patients aged 65 and older enrolled in an Accountable Care Organization (ACO) risk contract at an academic medical center and hospitalized for COVID-19 between March 11, 2020 - June 3, 2020 Measurements We linked a COVID-19 Registry with administrative claims data to quantify a frailty index and its relationship to mortality, healthcare utilization, and expenditures over 6 months following hospital discharge. Kaplan Meier curves and Cox Proportional Hazards models were used to evaluate survival by frailty. Kruskal-Wallis tests were used to compare utilization. A generalized linear model with a gamma distribution was used to evaluate differences in monthly Medicare expenditures. Results Much of the cohort was classified as moderate to severely frail (65.4%), 24.3% mildly frail, and 10.3% robust or pre-frail. Overall, 27.2% (n=37) of the cohort died (n=26 during hospitalization, n=11 after discharge) and survival did not significantly differ by frailty. Among survivors, inpatient hospitalizations during the 6-month follow-up period varied significantly by frailty (p=0.02). Mean cost over follow-up was $856.37 for the mild and $4914.16 for the moderate to severe frailty group, and monthly expenditures increased with higher frailty classification (p <.001). Conclusions In this cohort, claims-based frailty was not significantly associated with survival but was associated with follow-up hospitalizations and Medicare expenditures. Electronic Supplementary Material Supplementary material is available for this article at 10.14283/jfa.2023.15 and is accessible for authorized users.

Using a Claims-Based Frailty Index to Investigate Frailty, Survival, and Healthcare Expenditures among Older Adults Hospitalized for COVID-19 at an Academic Medical Center.

BACKGROUND: Frailty is associated with mortality in older adults hospitalized with COVID-19, yet few studies have quantified healthcare utilization and spending following COVID-19 hospitalization. OBJECTIVE: To evaluate whether survival and follow-up healthcare utilization and expenditures varied as a function of claims-based frailty status for older adults hospitalized with COVID-19. DESIGN: Retrospective cohort study. PARTICIPANTS: 136 patients aged 65 and older enrolled in an Accountable Care Organization (ACO) risk contract at an academic medical center and hospitalized for COVID-19 between March 11, 2020 - June 3, 2020. MEASUREMENTS: We linked a COVID-19 Registry with administrative claims data to quantify a frailty index and its relationship to mortality, healthcare utilization, and expenditures over 6 months following hospital discharge. Kaplan Meier curves and Cox Proportional Hazards models were used to evaluate survival by frailty. Kruskal-Wallis tests were used to compare utilization. A generalized linear model with a gamma distribution was used to evaluate differences in monthly Medicare expenditures. RESULTS: Much of the cohort was classified as moderate to severely frail (65.4%), 24.3% mildly frail, and 10.3% robust or pre-frail. Overall, 27.2% (n=37) of the cohort died (n=26 during hospitalization, n=11 after discharge) and survival did not significantly differ by frailty. Among survivors, inpatient hospitalizations during the 6-month follow-up period varied significantly by frailty (p=0.02). Mean cost over follow-up was $856.37 for the mild and $4914.16 for the moderate to severe frailty group, and monthly expenditures increased with higher frailty classification (p <.001). CONCLUSIONS: In this cohort, claims-based frailty was not significantly associated with survival but was associated with follow-up hospitalizations and Medicare expenditures.

MIXEDMETHODS EVALUATIONOF THE INITIAL IMPLEMENTATION OF A MEDICAID ACCOUNTABLE CARE ORGANIZATION PROGRAM TO ADDRESS FOOD AND HOUSING INSECURITY

BACKGROUND: States and health systems are investing in programs to address patients' unmet social needs, such as food and housing insecurity, but there has been limited evaluation of the implementation of these programs. In 2020, Massachusetts initiated the Flexible Services (Flex) program to provide funding to Medicaid accountable care organizations (ACOs) to address food and housing insecurity through community resources. The study objective was to examine initial implementation of Flex (March 2020-July 2021), using the Reach, Efficacy, Adoption, Implementation, Maintenance (RE-AIM) framework. METHODS: This mixed-methods evaluation was part of LiveWell, a longitudinal study assessing the impact of Flex on community health center patients aligned with two large hospitals within Mass General Brigham (MGB) in Boston, MA. ACO participants were screened annually for food and housing insecurity. To assess reach, we examined Flex enrollment using electronic health record data of enrollees ≥21 years old. Eligibility criteria for Flex included: 1) enrollment in MGB Medicaid ACO, 2) food or housing insecurity identified by screening or clinical encounter, and 3) a complex health condition (e.g., uncontrolled diabetes, depression). To assess implementation, adoption, and effectiveness, we conducted qualitative interviews with Flex enrollees (N=16) and health system staff (N=15). Interviews were analyzed using the Framework Method. RESULTS: Of 44,417 ACO enrollees, 693 (2%) were enrolled in Flex in the first 17 months. A total of 19,275 (43%) of ACO enrollees and 521 (75%) of Flex enrollees completed annual screening for food/housing insecurity. Mean ACO participant age was 40 years (SD: 14);62% were female;32% were Hispanic. Mean Flex enrollee age was 46 years (SD: 13);81% were female;54% were Hispanic. Implementation challenges included complex eligibility requirements, administrative burden (e.g., tracking, documentation), COVID- 19 factors (e.g., reduced clinic visits), and coordinating with community organizations. Facilitators included raising staff awareness to increase referrals, administrative funding for enrollment staff, adaptive strategies to identify eligible patients, and streamlined communication with community organizations. Flex enrollees reported improvements in healthy eating and food security. Patients who were able to select food or meals based on their preferences reported higher satisfaction. Patient-reported housing support included assistance with utility bills and affordable housing applications. CONCLUSIONS: To improve reach, adoption, and effectiveness in diverse populations, states and health systems implementing programs to address social needs should consider expanding screening for food and housing insecurity, minimizing administrative burden, providing funding for enrollment staff, and tailoring programs to patient preferences.