Potential Savings from Optimized Plan Choice in Medicare Beneficiaries with Schizophrenia Abstract The Medicare prescription drug benefit (Part D) relies on private market mechanisms to deliver care: a typical beneficiary can choose from between 45 and 57 private stand-alone Part D plans (PDP) in 2009, depending on his/her region of residence. It is known that beneficiaries do not choose plans based on their medication needs;and that the ability to choose plans is poorer among beneficiaries with mental disorders. The purpose of this study is to simulate personalized plan choice based on patient medication needs in schizophrenia and quantify potential improvement in patient drug coverage and potential savings to the government. Part D provides substantial premium and cost-sharing assistance to beneficiaries qualifying for the low-income subsidy (LIS) program. About 93% of PDP enrollees with schizophrenia received the LIS in 2007, whereas 40% of all PDP enrollees did. The majority of LIS enrollees are randomly assigned to PDP plans with premiums at or below the regional average. Random assignment does not assign enrollees to plans based on their medication needs and has caused severe problems including disruptions in plan coverage for 5.9 million between 2007 and 2010; and high beneficiary and government spending for those assigned to plans requiring high copayments. Our study will develop intelligent assignment algorithms based on beneficiaries' medication needs and dynamics of plan features in the Part D market. The algorithms can be easily implemented each year after an initial setup of software without substantial costs. The intelligent assignment method can be used by the government to assign/reassign beneficiaries or provide personalized assistance to help beneficiaries with schizophrenia to choose plans. The intelligent assignment method has the potential to substantially reduce government spending while improving patient outcomes.

Public Health Relevance

The prevalence rate of schizophrenia is 2.6% among Medicare beneficiaries enrolled in stand-alone prescription drug plans (PDP), much higher than the 1.1% among the US general population. About 93% of Medicare PDP enrollees with schizophrenia received the low-income- subsidy (LIS) on the plan premium and cost-sharing for medications, whereas 40% of overall Medicare Part D enrollees did. The majority of LIS enrollees are randomly assigned to PDP plans with premiums at or below the regional average. Random assignment does not assign enrollees to plans based on their medication needs and this leaves room for substantial savings and improvement in drug coverage under alternative assignments. It is known that beneficiaries are unable to choose plans based on their medication needs. Compared with average beneficiaries, patients with schizophrenia are poorer, less educated, and less likely to make rational plan choices. If we find that enrollees wit schizophrenia are assigned in ill-fitting plans that require utilization reviews for many of their psychiatric and non-psychiatric drugs and are unable to switch to better plans, alternative assignments based on their medication use or personalized assistance in switching plans are necessary. The intelligent assignment method has the potential to substantially reduce government spending while improving patient outcomes.

Agency
National Institute of Health (NIH)
Institute
National Institute of Mental Health (NIMH)
Type
Exploratory/Developmental Grants (R21)
Project #
1R21MH100721-01A1
Application #
8635037
Study Section
Special Emphasis Panel (SERV)
Program Officer
Rupp, Agnes
Project Start
2013-12-01
Project End
2015-11-30
Budget Start
2013-12-01
Budget End
2014-11-30
Support Year
1
Fiscal Year
2014
Total Cost
$187,870
Indirect Cost
$62,870
Name
University of Pittsburgh
Department
Public Health & Prev Medicine
Type
Schools of Public Health
DUNS #
004514360
City
Pittsburgh
State
PA
Country
United States
Zip Code
15213
Zhang, Yuting; Zhou, Chao; Baik, Seo Hyon (2014) A simple change to the medicare part D low-income subsidy program could save $5 billion. Health Aff (Millwood) 33:940-5