Abstract
Background/Aims Employers have recently seen rapid increases in their cost of providing health insurance benefits for their employees, partly because the traditional health insurance benefit design does not reflect the value of healthcare consumed. Value-based insurance design (VBID) seeks to address this problem by incorporating value as well as cost in each beneficiary’s healthcare decisions. Since 2007, Geisinger Health System (GHS) has implemented its own version of VBID called MyHealth Rewards for its employee population. Key components of the program included zero-copay prescription drugs for patients with certain chronic conditions and a health management program designed to enhance employee self-management. In this study, we examine whether MyHealth Rewards has led to improvements in employee health outcomes in terms of reduced incidence of stroke and MI and also whether it has lowered the cost of care.
Methods A cohort of GHP members who remained as Geisinger employees throughout the study period was identified (N = 4,895). Because MyHealth Rewards program was offered only to Geisinger employees starting in 2007, this cohort was compared against a comparison group consisting of GHP members who remained as non-GHS employees during the same period (N = 12,077). Propensity score matching was used to stratify the sample based on a set of baseline characteristics. Cox proportional hazard model was used to estimate time to first incidence of stroke or myocardial infarction (MI) since 2007. The total medical cost excluding prescription drugs was analyzed using a two-part model consisting of a generalized linear model and a logistic regression model. Covariates included employee age, gender, disease management and chronic condition status, and medical home status.
Results The Geisinger employee cohort experienced stroke or MI later than the non-Geisinger employee group (hazard ratios of 0.73 and 0.56, respectively; P <0.01). There was also reduction in cost of care by about 10% to 13% associated with the Geisinger employee cohort relative to the non-GHS employee group during the second and third years of the program. However, the cost reduction disappeared in subsequent years.
Conclusions VBID has the potential to be an effective tool in achieving the dual goal of improving health outcomes while lowering cost.
