Accountable Care Organizations Are Associated with Savings Among Medicare Beneficiaries with Frailty
- PDF / 156,943 Bytes
- 3 Pages / 595.276 x 790.866 pts Page_size
- 99 Downloads / 186 Views
J Gen Intern Med DOI: 10.1007/s11606-020-06166-6 © Society of General Internal Medicine 2020
INTRODUCTION
Efforts to address high health care costs in the USA are increasingly focusing on targeting care for high-need, highcost individuals.1 A recent National Academy of Medicine report identified older adults with frailty and a high burden of chronic conditions as a key population that may benefit from specific targeting of programs and interventions.2 These individuals are at high risk for hospitalizations, especially related to potentially avoidable conditions, and for prolonged hospital and post-acute care courses.3 Accountable Care Organizations (ACOs) have received a lot of attention as a promising model for organizing care delivery with more intensive care management and care coordination strategies to help manage and control costs of health care. Early results suggest ACOs have achieved modest savings for the general Medicare population4; however, the degree to which ACOs may have reduced spending for older people with frailty and a high comorbidity burden is less known. Therefore, we sought to determine whether ACOs are associated with reduced spending for older adults with frailty.
METHODS
Using Medicare claims data from 2009 to 2016, first identified beneficiaries (65 years or older) with frailty, defined as those with at least 2 of 12 claims-based indicators as proposed by Kim and Schneeweiss (gait abnormality, malnutrition, failure to thrive, cachexia, debility, difficulty walking, history of fall, muscle wasting, muscle weakness, decubitus ulcer, senility, or certain durable medical equipment use like canes/ walkers).3, 5 We then performed a difference-in-differences analysis to compare changes in Medicare spending for older adults with frailty before and after entry in to the Medicare Shared Savings Program (MSSP) relative to a group of control patients in the same hospital-referral Received August 5, 2020 Accepted August 14, 2020
region (HRR) served by providers not participating in ACOs. We focused on cohorts of patients in practices entering the MSSP in 2012, 2013, and 2014. Our primary outcome was total annual Medicare spending and secondary outcomes were spending by type of setting. Given that spending is highly right-skewed, we used log gamma models, which adjusted for age, sex, race/ethnicity, dual status, reason for Medicare eligibility, comorbidities (using Chronic Condition Warehouse algorithms), and community-level characteristics from the American Community Survey (i.e., educational attainment and % below federal poverty line). The models also included fixed effects for ACO practices (with all non-ACO practices identified as “Control”), and each combination of hospital referral region and year. An indicator for “ACO Post” identifies all patients in MSSP practices in post years (2013–2016, depending on when the practice joined the MSSP). This study was performed using SAS Statistical Package version 9.4. Two-sided P values were used and considered significant at P < 0.05. This study was
Data Loading...
