The Affordable Care Act (ACA) has greatly increased the number of Medicaid enrollees since it expanded eligibility criteria on Jan. 1, 2014. Nearly 9 million additional U.S. adults now have Medicaid coverage, mostly in the 31 states and Washington, which have opted into Medicaid expansion.

The ACA has also had an important impact on hospital payer mix, mainly by decreasing the amount of uncompensated care in Medicaid-expansion states. Previous studies have shown disparities in the quality of inpatient care based on insurance type. Patients with Medicaid insurance often have longer hospitalizations and higher in-hospital mortality than commercially-insured patients and occasionally even than uninsured patients.

As hospitalists, we were interested in how Medicaid expansion under the ACA might have affected these inpatient outcomes.

In our study published in the Journal of Hospital Medicine ( 2016 Dec. doi: 10.1002/jhm.2649 ), we evaluated the impact of state Medicaid expansion status on payer mix, length of stay, and in-hospital mortality for general medicine patients discharged from U.S. academic medical centers. We considered Jan. 1, 2014, to be the date of ACA implementation for all states except Michigan, New Hampshire, Pennsylvania, and Indiana, which had unique dates of Medicaid expansion.

We were able to identify 3,144,488 discharges from 156 hospitals in 24 Medicaid-expansion states and Washington, and 1,114,464 discharges from 55 hospitals in 14 nonexpansion states between October 2012 and September 2015.

As expected, hospitals in Medicaid-expansion states experienced a significant 3.7% increase in Medicaid discharges and 2.9% decrease in uninsured discharges after ACA implementation. This represented a relative 19% jump in Medicaid and 60% drop in uninsured discharges. Hospitals in nonexpansion states did not see a significant change in payer mix after ACA implementation.

Despite this difference in payer mix trends, state Medicaid expansion status was not associated with differences in overall length of stay or in-hospital mortality in our study. More precisely, we looked at the length of stay and mortality indices, or ratio of observed to expected values, to control for such potential confounders as disease severity and comorbid conditions.

One possible explanation for our findings is that the higher proportion of Medicare and commercially-insured patients overshadowed the contribution of Medicaid patients to the overall length of stay and mortality indices.

To our knowledge, our study is the first to look at the effect of ACA implementation on inpatient outcomes. Early evidence suggests that Medicaid expansion has improved outpatient outcomes. Low-income adults in Medicaid-expansion states have shown greater gains in access to primary care clinics and medications and in the diagnosis of certain chronic health conditions than those in non-expansion states. However, these changes would not necessarily lead to improvements in the length of stay or mortality indices for Medicaid-expansion hospitals, since the measures account for patient acuity on admission.

The take-home message from our study for health policy makers is that state Medicaid expansion status had a neutral effect on both length of stay and mortality indices. This should be reassuring for states considering expansion of their Medicaid programs in the future.

As a next step, it would be useful to see research on the impact of ACA implementation on other inpatient outcomes that may vary with insurance type, such as readmissions or hospital-acquired complications.

The take-home message for hospitalists is that there is more work to be done in reducing disparities in inpatient care based on payer status. Though not a primary focus of our study, we did see variation in the length of stay and mortality indices based on insurance type.

It is unclear whether these differences occurred because of variation in the expertise of inpatient providers, access to invasive procedures or medical therapies, the timeliness of discharge to post-acute care facilities, or other patient- or system-level factors. However, these disparities warrant our improvement efforts moving forward.

Mary Anderson, MD, and Christine Jones, MD, are hospitalists at the University of Colorado (Aurora) Hospital and assistant professors in the department of medicine at the University of Colorado.