Caring for the Uninsured with Prostate Cancer: A Comparison of Four Policy Alternatives in California

Abstract

The IMPACT Program seeks to improve access to prostate cancer care for low-income, uninsured men. The objective of the current study was to compare the cost-effectiveness of four policy alternatives in treating this population. We analyzed the cost-effectiveness of four policy alternatives for providing care to low-income, uninsured men with prostate cancer: (1) IMPACT as originally envisioned, (2) a version of IMPACT with reduced physician fees, (3) a hypothetical Medicaid prostate cancer treatment program, and (4) the existing county safety net. We calculated cost-effectiveness based on incremental cost-effectiveness ratios (ICERs) with the formula ICER = (Cost(alternative strategy) - Cost(baseline strategy)) / (QALY(alternative strategy) - QALY(baseline strategy)). We measured outcomes as quality-adjusted life years (QALYs). "Best-case" scenarios assumed timely access to care in 50% of cases in the county system and 70% of cases in any system that reimbursed providers at Medicaid fee-for-service rates. "Worst-case" scenarios assumed timely access in 35 and 50% of corresponding cases. In fiscal year 2004-2005, IMPACT allocated 11% of total expenditures to administrative functions and 23% to fixed clinical costs, with an overall budget of $5.9 million. The ICERs ($/QALY) assuming "best-case" scenarios for original IMPACT, modified IMPACT, and a hypothetical Medicaid program were $32,091; $64,663; and $10,376; respectively. ICERs assuming "worst-case" scenarios were $27,189; $84,236; and $10,714; respectively. County safety net was used as a baseline. In conclusion, IMPACT provides underserved Californians with prostate cancer care and value-added services with only 11% of funds allocated to administrative fixed costs. Both the original IMPACT program and the hypothetical Medicaid prostate cancer program were cost-effective compared to the county safety net, while the reduced-fees version of IMPACT was not.