Association for Public Policy Analysis & Management

Patient cost-sharing has traditionally been used to offset moral hazard and give patients incentive not to seek health care when it is of low value to them. Yet the impact of cost-sharing among low-income populations and for high-value services may run counter to these goals, and there is limited research regarding the impacts of cost-sharing among low-income populations. This study investigates the relationship between state Medicaid co-payment policies and use of cancer screening among non-elderly non-disabled adult women enrolled in Medicaid. While cost-sharing in Medicaid is typically low, states have more flexibility with patient cost-sharing in adult populations than among children and most charge co-payments to at least some groups of adults in Medicaid. The ACA provides incentives for states to eliminate cost-sharing for preventive care, yet states still have leeway in determining whether to charge co-payments.

We examine the impact of state-level cost-sharing policies for general outpatient visits and for preventive services on breast and cervical cancer screening among non-elderly non-disabled fee-for-service Medicaid enrollees prior to implementation of the ACA. Data on the use of screening services come from the Medicaid Analytic eXtract (MAX) files for 2003, 2008 and 2010 across 46 states. MAX data contain enrollee-level information on eligibility, demographics, enrollment details, and claims. We use information from diagnosis and procedure codes from outpatient claims to identify breast and cervical cancer screening. These are merged with state-year level data on co-payment policies (whether the state Medicaid program charged co-payments for general visits to outpatient providers and whether co-payments were waived for preventive services), as well as county-level controls capturing healthcare capacity and population demographics from the Area Health Resources File. We consider receipt of mammograms among women ages 50-64 and Pap tests among women ages 21-64, using state and year fixed effect models to assess the impact of changes in cost-sharing policies on utilization of screening services.

Among women who have cost-sharing for general visits, those without cost-sharing for preventive services (women for whom standard co-payments are waived for a preventive visit) have higher rates of both breast and cervical cancer screening. Predicted rates of screening based on regression models are similar for women who have no co-payments for any outpatient visits and those with co-payments for all visits (general and preventive). For both types of screening we examine, women who have cost-sharing for most visits but not for preventive care are approximately three percentage points more likely to be screened than are women for whom there is no difference in cost-sharing policies between general and preventive services. These results may reflect a behavioral response on the part of patients and/or providers since the lack of co-payment for preventive services may be more salient in the presence of co-payments for other services, or preventive services may be more likely to be provided in order to waive a co-payment that would be required otherwise. Our results suggest that full coverage of preventive services with no cost-sharing can increase use of screening among Medicaid enrollees.