Does Health Insurance Promote Health through Providers and Networks?

OTHER PROJECT INFORMATION – Project Summary/Abstract
Does Health Insurance Promote Health through Providers and Networks?
Through their varying provider networks and coverages, health insurance plans directly and indirectly steer
patients toward some clinicians and hospitals and away from others. If providers differ in their quality of
care and especially if they provide better care for certain types of patients, such network variation can
meaningfully affect health outcomes at the plan level, and not just the provider level. Yet, consumers
selecting health plans lack information about which plans will direct them to better providers, obscuring this
potentially important channel to improving population health.
This project will develop novel outcome-based measures of insurance plan quality based on where
enrollees receive care, and taking account how causal impacts on health vary across provider networks. By
furthering the scientific understanding of the relationship between provider quality and patient outcomes at
the plan level, we aim to equip patients with new information for selecting insurance plans. Our measures
will also provide regulators with new instruments to reward plans that make people healthier.
At the core of our research plan are new, causal estimates of how mortality impacts vary across hospitals
and physicians and, in turn, across the health plans that rely on different provider networks. We focus on
older Americans enrolled in Medicare Advantage plans, leveraging novel data on the provider networks
used by different Medicare Advantage plans. The estimates will characterize which hospitals and providers
produce the best outcomes for beneficiaries with different demographics and comorbidities and thus which
plans are suitable for each type of patient by directing them to providers who produce good outcomes for
conditions they have or are likely to develop.
The project has four primary aims:
Aim 1: Measure and Validate Mortality Effects for Healthcare Providers
Aim 2: Link Physician and Hospital Mortality Effects to Plan Mortality Effects
Aim 3: Explore Heterogeneity in Mortality Effects Across Beneficiaries
Aim 4: Evaluate the Impact of Policy Proposals

Grant Number: 5R01AG075073-04
NIH Institute/Center: NIH
Principal Investigator: Jason Abaluck