There is substantial variation and low correlation in healthcare spending across Medicare, Medicaid, and private insurance plans within different US regions, a study published in JAMA Network Open found.
The US spends around $3.8 trillion per year on healthcare funded by Medicare, Medicaid, and private insurers. However, spending and healthcare utilization are rarely the same for all three payers.
Researchers used data from the Healthcare Cost Institute (HCCI), Medicare fee-for-service claims, and the Transformed Medicaid Statistical Information System Analytic Files to determine the correlation in regional spending and utilization across the three payers.
The study reflects data from 25.4 million individuals with employer-sponsored coverage in 2017, 69.8 million with Medicaid coverage in 2016 and 2017, and 26.7 million with Medicare fee-for-service coverage. Researchers looked at spending patterns in 241 hospital referral regions (HRRs).
Healthcare spending per beneficiary varied across payers. The mean private insurance spending per beneficiary was $4,441, while the Medicare mean was $10,281, and the Medicaid mean was $6,127 per beneficiary. The overall mean for the three payers was $5,782 per beneficiary.
Medicaid had the most variation in spending per beneficiary across the HRRs, with a coefficient of variation of 0.233, according to the study. Private insurance had a coefficient of variation of 0.160 and Medicare had a coefficient of variation of 0.126.
Medicaid and private insurance plans likely saw more variation in spending than Medicare because Medicare relies on regulated payments to hospitals, while Medicaid and privately insured prices are generally determined by the region’s market.
In addition to spending variation among the payers individually, there was low spending correlation within regions across all three payers.
For example, the correlation coefficient between HRR level spending was 0.020 for private insurance and Medicare, 0.213 for private insurance and Medicaid, and 0.162 for Medicare and Medicaid.
According to researchers, past studies have documented low correlation between Medicare and private insurance spending. The low correlation between Medicare and Medicaid may be because Medicaid programs are administered by states, which can take their own regulatory approaches.
The low correlation across payers within regions led to only three HRRs in the country falling in the bottom quartile of spending for all three payers and four HRRs failing the highest quartile.
Healthcare utilization also varied by payer across HRRs.
The mean number of inpatient days was 1.30 for Medicare beneficiaries, 0.56 for Medicaid beneficiaries, and 0.21 for privately insured individuals.
Medicaid saw the most variation again, with a coefficient of variation of 0.301. The coefficient of variation was 0.207 for Medicare beneficiaries and 0.180 for privately insured individuals.
There was a higher correlation across payers within regions for healthcare utilization, researchers found. The correlation coefficient was 0.465 for private insurance and Medicare, 0.527 for private insurance and Medicaid, and 0.278 for Medicare and Medicaid.
Similarly, the correlates of inpatient utilization, including indicators of population health, hospital beds per capita, and reimbursement rates, were similar across the three payers on a regional level.
For Medicare, overall healthcare spending and utilization were correlated due to federally regulated reimbursement rates. However, there was low correlation between expenditures and utilization for private insurance because the market determines reimbursement rates. Correlation for Medicaid fell somewhere in the middle, as some prices are regulated in certain areas and negotiated in others, the study noted.
The study results revealed no correlates of spending variation across the three payers, indicating that payer-specific factors likely lead to variation in healthcare spending among individuals with different health plans.
Researchers suggested that policymakers consider payer-specific policies and be mindful of the spillover effects their policies could have on other beneficiary populations.
Source: Revcycle intelligence
