ANALYSIS OF MEDICARE PRESCRIPTION DRUG COVERAGE ENROLLMENT

Tengjiao Xiao, Aurelie Thiele

Abstract


Objectives. To identify factors associated with Medicare beneficiaries' choices of prescription drug plans as part of Medicare Advantage (MAPDs) or stand-alone prescription drug plans (PDPs) in order to help policy-makers understand what drives constituents’ choice between the two and potentially refine prediction models of future prescription drug plan enrollment.

Study Design. We propose a methodology based on beta regression to identify factors associated with Medicare enrollees’ choice of MAPDs or PDPs. We consider demographic factors, medical condition factors and plan characteristics.

Data Sources. We use county level MAPD/PDP penetration rates and Medicare population data for all counties in the United States, except in Alaska.

Principal Findings. Our approach documents key differences in factors driving MAPD and PDP penetration rates.

Conclusions. Our methodology provides insights into the different segments of the U.S. Medicare population that MAPDs and PDPs appeal to.


Full Text:

PDF

References


Medicare. “How to get drug coverage”. Retrieved May 5, 2015. http://www.medicare.gov/sign-up-change-plans/get-drug-coverage/get-drug-coverage.html

Kaiser Family Foundation. Research brief. The Medicare Part D Prescription Drug Benefit. September 2014. http://files.kff.org/attachment/medicare-prescription-drug-benefit-fact-sheet

Richard R. Cline, Marcia M. Worley, Stephen W. Schondelmeyer, Jon C. Schommer, Tom A.

Larson, Donald L. Uden, and Ronald S. Hadsall. PDP or MA-PD? Medicare Part D enrollment

decisions in CMS Region 25. Research in Social and Administrative Pharmacy, 6:130-142, 2010.

Silvia L.P. Ferrari and Francisco Cribari-Neto. Beta regression for modelling rates and pro-

portions. Journal of Applied Statistics, 31(7):799-815, 2004.

Kaiser Family Foundation. Medicare advantage: MA-PD plan enrollment as a percent of total Medicare population. Table “MA-PD Plan Enrollment as a Percent of Total Medicare Population” 2014 data. http://kff.org/medicare/state-indicator/enrollees-as-a-of-total-medicare-population/

Centers for Medicare & Medicaid Services. Public use file: New data on geographic variation. http://www.cms.gov/Research-Statistics-Data-and-Systems/Statistics-Trends-and-Reports/Medicare-Geographic-Variation/GV_PUF.html

United States Census Bureau. State & county quickfacts. http://quickfacts.census.gov/qfd/index.html

Centers for Medicare & Medicaid Services. Prescription drug coverage - general information. https://www.cms.gov/Medicare/Prescription-Drug-Coverage/PrescriptionDrugCovGenIn/

Christopher J. Swearingen, Maria S. Melguizo Castro, and Zoran Bursac. Modeling percentage outcomes: The %beta regression macro. In SAS Global Forum 2011, number Paper 335-2011, 2011.

Francisco Cribari-Neto and Achim Zeileis. Beta regression in R. Journal of Statistical Software, 34:1-24, 2010.

Significance codes: 0.000001 '***', 0.001 '**', 0.01 '*', 0.05 '.', 0.1 ' '.

N/A means the variable is not significant in the model.

The actual coefficient estimate of this variable is a very small number close to zero, but not zero exactly.


Refbacks

  • There are currently no refbacks.


©Journal of Health Care Finance