Venue: The Fuqua School of Business, Duke University, 1 Towerview Drive, Durham, NC 27708-0120
Presentation
The Effects of Consumer-Directed Health Plans on Utilization and Cost of Care
Consumer-directed health plan (CDHP) has been lauded by some observers as the next great idea in health care financing and pilloried by others as the latest bad idea in health care financing. Despite the often extreme differences of opinion on the subject, two things are clear: CDHP has attracted considerable attention from employers and individuals, and there has been a striking lack of research on the effects of pairing a high deductible policy with a health spending account covering a portion of that deductible on use of services and overall cost of care. The purpose of this research is to determine empirically what the effects are of CDHP on health care utilization.
Our research benefits from a unique data source and setting. We have acquired all administrative claims data and enrollment records from a consumer-directed health insurance company operating in select states in the US. The company operates almost exclusively in the small group market where its products are offered only as a total replacement option to small employers. The total replacement aspect of the setting (versus as an add-on to other health insurance options) enables us to bypass the selection issue that plagues the few other efforts to examine the impact of CDHP on health care use. We observe spending account levels, deductible levels, and cost-sharing requirements for the 4614 companies offering the insurance company's products. Detailed claims are available for all enrollees, which in addition to traditional claims information (ICD-9 and CPT codes) also provide a running total of the remaining spending account balance. Moreover, 2144 (46%) of the companies are observed for multiple years.
Statistical identification of the effects of spending account levels and other plan characteristics will be achieved by exploiting within-company variation over time in plan characteristics; as such, companies and their enrollees will serve implicitly as their own controls given the plausibly exogenous changes in plan characteristics over time. Indeed, nearly 40% of companies observed for two or more years altered deductible or copayment levels; in addition, given roll-over of the unused spending account balances at the end of the contract year, spending accounts at the start of the contract year (and hence the amount that must be paid at full cost out-of-pocket between the deductible and the spending account) are variable as they represent past spending.
Our estimates will represent the first credible estimates of the pure effect of CDHP on health care use and cost without being clouded by uncertainties regarding the effect of selection. The policy implications of our findings are potentially quite significant as HSA legislation is likely to be closely scrutinized by the new Congress and employers and consumers are urgently looking for means of reducing their exposure to high health insurance premiums. Our findings will also provide the empirical basis needed to forecast trends in health care cost and utilization as more and more workers switch to CDHP-type plans, including the imminent introduction of Medicare Medical Savings Account plans.