60 million people are enrolled in Medicare and 75 million in Medicaid, making these programs among the
most influential in the US health care system. Through them, federal and state governments regularly
experiment with new models of payment and delivery, trying out policy ideas on millions of beneficiaries. A
key laboratory for this experimentation is the Center for Medicare and Medicaid Innovation (the Innovation
Center), which has recently committed to “make equity a centerpiece of every model.”
New payment and delivery models may either widen disparities (e.g., through inadequate risk adjustment and
selection), ameliorate them (e.g., by addressing social determinants of health with flexible reimbursement), or
both (across different outcomes and subgroups). However, existing methods are not well suited to estimate
these programs' impacts on health equity.
Medicare and Medicaid programs and policies, because they are rarely randomized, are evaluated using
observational causal inference methods. One popular such method is difference-in-differences (diff-in-diff),
which compares changes in outcomes for people exposed to a new model versus changes over the same time
in a comparison group of people not exposed to the model. While useful for overall program impacts,
standard diff-in-diff methods are not suited to equity evaluations. The target estimands are not formulated
for equity questions, they do not accommodate varying responses across subgroups, and they do not account
for variation in treatment effects. In addition, the methods are not grounded in frameworks for health equity
and thus fail to use causal assumptions tailored for groups who experience disproportionate disadvantage.
Partly because of these methodological shortcomings, existing evidence on equity impacts of Medicare
alternative payment models is sparse and mixed. Primary care transformation models have received even less
attention. We propose to develop and apply novel methods to study how Comprehensive Primary Care Plus
(CPC+) affects health care for Black Medicare beneficiaries. This national advanced primary care medical
home model was introduced across 18 regions in 2017-18 and will continue through 2021. Our specific aims
are 1) To estimate impacts of CPC+ on health care and health equity for Black Medicare beneficiaries, 2) To
estimate variation in responses to CPC+ within subgroups of Black Medicare beneficiaries with intersecting
identities of gender, disability, and dual eligibility for Medicaid, and 3) To explore variation in responses to
CPC+ across two contextual dimensions: primary care practice characteristics and the local health care and
socioeconomic environment. This application is in response to NOT-HS-21-014 Special Emphasis Notice
(SEN): AHRQ Announces Interest in Health Services Research to Advance Health Equity.
