Abstract
Objective: To assess the effects of an alternative method of paying home health agencies for services to Medicare beneficiaries, based on a demonstration program. Data Sources/Study Setting. Primary and secondary data collected on participating home health agencies in five states and their patients during the three-year demonstration period. Primary data included patient surveys at discharge and six months later, and two rounds of interviews with executive staff of the agencies. Secondary data included agencies' Medicare cost reports, quality assurance reviews, Medicare claims data, demonstration claims data, demonstration patient intake forms and plan of treatment forms. Study Design: The 47 agencies volunteering to participate in the demonstration were each randomly assigned to the treatment or control group. Treatment group agencies were paid a predetermined rate based on their inflation-adjusted cost per visit during the year preceding the demonstration; control group agencies were paid under Medicare's conventional cost reimbursement method. Demonstration impacts were estimated by comparing outcomes for the two groups of agencies and their respective patients, using regression models to control for any remaining differences. Principal Findings. Agencies paid under prospective rate setting were slightly better at holding per-visit cost increases below inflation than were control group agencies. The change in payment method had no effect on agencies' volume of Medicare visits or quality of care, nor on patients' use of Medicare services or other formal or informal care services. Conclusion: Changing from cost- based reimbursement to predetermined payment rates for Medicare home healthcare visits would not lead to large savings for the Medicare program, but would not increase costs to Medicare or adversely affect patients or their caregivers.