In 2013, there were 10.7 million people enrolled [PDF] in both Medicare and Medicaid. Dual eligibility depends on age, income, and disability. Dually enrolled beneficiaries are also responsible for a large share of program costs overall; 31% of Medicare fee-for-services spending for 18% of beneficiaries [PDF] who are dually enrolled. Given the additional health challenges [PDF] faced by dual eligibles, this disproportionate cost is not terribly surprising.
Nearly half of dual beneficiaries under age 65 have a serious mental illness, or SMI (bipolar disorder, major depressive disorder, or schizophrenia). People with SMI on public insurance often underuse care, relative to expert recommendations and their needs–potentially increasing longer term costs.
In a recent article published ahead-of-print in Medical Care, Dr. Marguerite Burns and colleagues estimated the effect of transitioning to dual coverage (after Medicaid enrollment) on health care use among adults under age 65 with SMI. The authors used claims and enrollment data from two states, South Carolina (n=1,837) and Missouri (n=6,012), for individuals transitioning between January 2004 to December 2007. A natural experiment framework allowed estimation via interrupted time series (see the paper for details).
Individuals were transitioning from fee-for-service Medicaid-only to dual coverage. Dual coverage meant the following for enrollees: continued low Medicaid copayments, no responsibility for Medicare cost-sharing, increased physician reimbursement rates, and, in South Carolina, no more limits on mental health therapy and outpatient visits.
Effects of transitioning to dual status in Missouri (left) and South Carolina (right)
The authors analyzed the total number of outpatient, ED, and inpatient service use, and the number of visits for mental health and substance use disorder (MHSUD). Adjusted analyses controlled for the baseline Medicaid-only trend (in other words, all comparisons were to expected rates for Medicaid-only enrollees).
In Missouri, the likelihood of an outpatient visit increased by 9% after transitioning to dual coverage, and the average number of visits increased by 7.2%. The probability of having an outpatient MHSUD visit increased by 3.8%. The probability of an ED visit and the number of ED visits increased as well, by 21.6% and 30.9%. For MHSUD ED visits, the parallel increases were 23.3% (probability of any visit) and 32.3% (number of visits). Inpatient admissions also increased overall and for MHSUD.
In South Carolina, there was an increase in the probability of an outpatient visit (2.3 percentage points, 4% relative increase) and no change for MHSUD outpatient visits. The probability and number of ED visits also did not change. However, there was a relative increase in the probability of any inpatient admission of 26.6%; for MHSUD inpatient the relative increase was 42%. The average number of inpatient admissions increased by 29.7% after beneficiaries obtained dual coverage, and the mean inpatient days per month increased by compared to the expected days by 73%. The mean number of MHSUD-specific inpatient days per month increased more than 100%.
This novel longitudinal study shows the increase in utilization across several measure of care in two US states. As noted by the authors, health care use was not lower than expected based on Medicaid-only trends in any category and the increases were not restricted to MHSUD -related services. Certain features of the programs, in particular the “welcome to Medicare visit,” correspond to the increased outpatient use found in the results. Interpreting the increased ED and inpatient care is more complicated, especially because cost-sharing did not change for patients.
The magnitude of Burns and colleagues’ results speak for themselves, but also leave a challenge for future researchers. The questions remain: do these increases in utilization result in changes in health status and outcomes among the newly dually eligible? And what are the mechanisms for the observed changes in ED and inpatient utilization?
Jess Williams