Geographic variation in health spending, the extent to which it exists, and the causes, seem to be a never-ending inspiration for research. Most has used Medicare data, which has been the primary nationally available data on a common benefit plan. Although Medicaid is highly variable across the states, some sub-populations have federally mandated common eligibility requirements and more similar benefits. An article in Health Affairs used the Medicaid disabled population to discern state and regional variations in health spending from 2001 to 2005 and adjusted for case mix to attempt to uncover adjusted differences due to utilization and unit price. (HA Article) The results suggest interesting regional and state patterns of care that often have mixed effects on spending.
The study covered only those who were eligible for Medicaid, not dual eligibles, and only those persons in fee-for-service Medicaid. It also focused on acute, not long-term, care services, such as hospital inpatient, hospital outpatient, physician office, prescription drug, DME and similar categories. The relationship of spending to market and Medicaid program characteristics was examined. Although data was analyzed on a state and hospital referral region level, to get a higher scale picture, the nine Census Bureau regions were also looked at. Overall, per beneficiary spending was high in New York, much of New England, Minnesota, Maryland and Alaska and low in the South and Washington State. Spending per beneficiary was about 67% higher in New York than in California and more than twice as high in Connecticut as in Oklahoma or Mississippi.
There was a wide range in hospitalization rates, per day hospital payments, number of physician visits, payments for physician services and use and cost of prescription drugs. Widely variant regional patterns emerge; for example New England had high overall spending but low hospitalization rates, while spending tended to be low in the South but some states had high hospitalization rates. Prescriptions per year ranged from 55 in Connecticut to 18 in Texas. The ten highest spending states were $1650 above the national yearly average, of which 72% was due to higher utilization. The ten lowest states were $1,161 below the average, of which 58% was volume related. Some states showed patterns of greater outpatient and drug use and fewer hospitalizations. The Pacific area had low outpatient prices and fewer hospitalizations and prescription fills, leading to overall lower costs.
More hospitalizations were associated with more hospital beds in a region and fewer admissions with more primary care resources. Higher prices for hospitalizations were correlated with more hospital use. Less healthy states had more admissions and those with a high managed-care presence had fewer. Washington State was singled out for having low spending, largely due to significantly lower inpatient use, with some compensation in outpatient and drug use. This data is very useful for states to evaluate what combination of benefits and reimbursement seems to produce the lowest spending and to ascertain where abuse may be occurring.