The effect of changes in drug benefit design among individuals with diabetes in large employer-sponsored insurance plans

THE EFFECT OF CHANGES IN DRUG BENEFIT DESIGN AMONG INDIVIDUALS WITH DIABETES IN LARGE EMPLOYER-SPONSORED INSURANCE PLANS By NINEE SHOUA YANG August 2011 Advisor: Dr. Allen C. Goodman Major: Economics Degree: Doctor of Philosophy With spending for prescription drugs rising so rapidly, employers and insurers are seeking different cost-cutting strategies to stem this tide. Given that prescription drugs have become an indispensable means to treat and manage chronic illnesses, the issues of affordability and trade-offs between medications and other health care services are important for chronically ill patients, particularly for patients with diabetes who typically have more than one comorbidity that require drug therapy, and their health insurance plans. In this dissertation, I analyze the effect of prescription drug cost-shifting via changes in drug benefit design on healthcare expenditure among individuals with diabetes; I take into account the comorbid effect of diabetes for the age population ranging from 18 to 62. Study design, data and organization of the report: This study uses a retrospective research design with observational historical data from the MarketScan Research Database from 2000 to 2001. The subjects are individuals enrolled in large employer-sponsored health insurance plans aged 18 to 62 who were diagnosed with diabetes from 2000 to 2001. The analysis of this study relies on the assumption that changes in spending are caused by changes in the drug benefit plan resulting from the action of employers as they seek different ways to cut cost. Regression using GEE estimation technique was used for the analysis. Major findings and conclusions: The overall effect on total healthcare expenditure was a decrease of $1,532.32 on average in the intervention group relative to the comparison group. The average follow-up year incremental out-of-pocket spending for drug following the changes in drug benefit design for an individual in the intervention group was $15.64. Changes in benefit plan design continue to be a significant predictor for drug spending only. Specifically, relative to the comparison group\u27s drug spending, the intervention group is more likely to decrease spending on average by $160.45 for drug services. To a certain extent, geographic region appears to be a significant predictor for inpatient (i.e. south) but not for drug spending and outpatient services spending. The comorbidity score is a significant predictor for increased spending in all three services and total healthcare expenditure and total out-of-pocket expenditure below 1% level. In the models for out-of-pocket expenditure for each service, the plan change conditional on time has a significant effect on only drug and outpatient services out-of-pocket spending, not on inpatient services. The findings suggest that the decrease in total health care expenditure borne mostly by the employers and insurers is explained by changes in drug benefit plan design during the study period from 2000 to 2001. Thus, higher levels of cost sharing transfer a large financial burden to the patient. All things considered, if all changes in healthcare spending are broadly defined as changes in drug benefit design imposed by the employers, then such changes are effective in managing the demand side of healthcare cost even in a chronically ill population who depends on regular drug therapy