Investigating the Impact of Medicare, Part D on the Diabetes Medications Using Enterprise Miner and Survival Analysis
The purpose of this paper is to examine the influence of Medicare, Part D on diabetes medications since its introduction in 2006, with the prescribed medicines data from the Medical Expenditure Panel Survey collected by the Agency for Healthcare Research and Quality.
In this paper, Summary Statistics, Kernel Density Estimation and Survival Analysis in Base SAS® and SAS/STAT® were used. The association analysis and the text mining analysis in Enterprise Miner™ were also utilized. Before that, SAS SQL, some SAS data step procedures or functions such as the transpose procedure were used to preprocess the data. Summary Statistics show that Medicare, part D increases the average Medicare payment for diabetes medications from $20 in 2005 to $130 in 2006. The kernel density estimation demonstrates that in 2005, the male beneficiaries pay more for the drugs, metformin and glyburide, while the female beneficiaries spend more on insulin and supplies. In 2006, the costs of the diabetes supplies and metformin remain higher than the other costs. The association analysis provides such information that the Medicare enrollees have fewer choices of their drugs than those with private insurance. Survival analysis indicates that insulin and metformin are more stable in terms of usage than the other drugs in general, while glyburide itself is very unstable in 2005. Moreover, the diabetes medications in 2005 are more stable than those in 2006.
View the pdf for.
This paper was written by Xiao Wang, Department of Mathematics, University of Louisville, KY.