Duke 2008 | ASHE | Equity and Efficiency in Health and Healthcare

Background Consumer-directed health plans (CDHPs) may encourage patients to eliminate unnecessary care and seek lower-cost, higher-quality providers. Yet if cost-sharing leads patients to forgo necessary care—particularly preventive care—costs and quality could suffer in the long run. To better understand these trade-offs, we aim to describe and test a conceptual model of preventive service use within CDHPs.

High deductibles create incentives to use fewer medical services. Because acute and chronic care visits are subject to the deductible in CDHPs, their use may decrease. Most CDHPs attempt to preserve the use of preventive care through exemptions from the deductible and employer contributions to medical spending accounts. However, up to 80% of preventive care is provided in the context of acute and chronic care visits. Thus, CDHP benefit design elements may work in opposition to one another with regards to preventive care, particularly for services provided during routine primary care encounters. These indirect effects of high deductibles may be most pronounced for people with the lowest incomes.

Methods We will test these effects by analyzing claims and human resources data from a large, self-insured U.S. employer that began offering CDHPs in 2001. The study population will be comprised of adults 21-64 years covered under employer-sponsored plans. The outcomes of interest will be the unadjusted and adjusted use of 5 preventive services by eligible adults as defined by U.S. Preventive Services Task Force guidelines. The 5 services are: routine preventive visits, influenza vaccination, and screening for cervical, breast, and colorectal cancer. The main independent variable will be plan type dichotomized as CDHP or traditional insurance product (HMO, PPO, POS).

Using 2001 data, we have created multivariable logistic regression models of preventive service use adjusting for age, income, gender, and health status. To determine whether high deductibles indirectly affect preventive service via fewer acute/chronic visits, we then adjusted for the number of outpatient visits. With the acquisition of 2000 data, we will add a difference-in-difference analysis to control for service use levels prior to the offer of CDHPs.

Results In our sample, the traditional insurance plans required small co-payments for 3 of the 5 preventive services. Deductibles for a family contract ranged from $0-$500 and the employee premium contributions ranged from $1300-$1950. The CDHPs exempted all 5 preventive services from the deductible. 85% of those who chose a CDHP opted for the plan with the lowest deductible ($3000 for a family plan) and lowest premium contribution ($930). In this case, $2000 was placed into a Health Reimbursement Account (HRA) by the employer on behalf of the employee.

For the 9100 insured adults in our sample, CDHP take-up was 27% and CDHP enrollees were younger and wealthier than those in traditional plans. There was no difference in the average number of outpatient visits by plan type. Unadjusted and adjusted rates of preventive service use were substantively similar. There was no difference in the use of breast or colorectal cancer screening among eligible adults by plan type. However, CDHP enrollees were more likely to use PAP smears (AOR 2.06, 1.78-2.38) and less likely to use preventive visits (AOR 0.82, 0.71-0.96) and influenza vaccination (AOR 0.47, 0.30-0.76). Adjusting for the number of outpatient visits did not change the association between plan type and preventive service use.

Conclusion/Implications CDHP enrollees may use certain preventive services at a different rate than those in traditional plans, but the mechanism explaining these differences remains unclear. Further analyses will attempt to adjust more fully for selection effects and model more precisely the indirect effects of high deductibles on preventive service use.

If the differences revealed in preliminary analyses are confirmed, cervical cancer screening could be seen as a model of consumer activation under CDHPs. On the other hand, decreased use of influenza vaccination and preventive visits would raise concern about the unintended negative effects of these plans. The results of this study should inform the policy debate regarding exemptions from the deductible for primary care services and value-based cost-sharing.