Hospitals and Health Reform: What’s at Stake?

    • August 22, 2014
Kathy Hempstead

Katherine Hempstead, PhD, MA, director and senior program officer, leads RWJF's work on health insurance coverage.

Policymakers and health care leaders aren’t the only ones watching hospital trends to gauge the early impact of the Affordable Care Act (ACA). Investors are also paying close attention. The Wall Street Journal reports that in their quarterly earnings calls, big hospital systems (HCA, Tenet) have reported that uncompensated care has declined, particularly in states that expanded Medicaid, and that overall volume has increased.

Yet the impact of health reform on hospitals will not be uniform. There is considerable variation in exposure to uncompensated care both within and between states. As we await the release of comprehensive 2014 discharge data on volume by payer, it is useful to look at the extent of this variation.

Differences Between States

According to data compiled by the Hospital ACA Monitoring Project (HAMP), there are big differences between and within states in the distribution of self-pay patients as a proportion of all inpatient admissions, meaning that some hospitals stand to gain considerably more under the ACA. For example in Minnesota, which expanded Medicaid, most hospitals (90%) had less than 7 percent of inpatient admissions from self-pay patients. On the other hand in Nevada, another expanding state, more than half of hospitals had more than 7 percent of their inpatient admissions from self-pay patients. For non-expanding states, there is a similar spread, as seen in the comparison between Wisconsin and South Carolina. While nine out of 10 of Wisconsin’s hospitals have less than 8 percent of their admissions from self-pay patients, half of South Carolina’s hospitals have 8 percent or more of admission from self-pay patients, including 10 percent who have more than 15 percent. The differences are even greater in the case of emergency room visits.  

Variation within States

The extent of variation within states differs as well. While 50 percent of Minnesota hospitals are compressed within the fairly narrow range of between 2 and 8 percent self-pay patients, in Nevada, the inter-quartile range extends from 18 to 30 percent. Since individual hospitals vary considerably in their exposure to uncompensated care, they do not all have the same opportunity to gain or lose from health reform in general and Medicaid expansion in particular. In some states that expanded Medicaid, hospitals have relatively little uncompensated care and thus will probably gain little, and in some non-expanding states, hospitals have a great deal of exposure and face a huge missed opportunity. Also, in states where there is more variation in payer mix, Medicaid expansion may have greater implications for the relative competitive position of hospitals (e.g., Virginia and Maryland), as compared with states where there is less variation (e.g., Kentucky and Tennessee).

Safety Net Scale

We grouped hospitals into four categories, according to the proportion of their inpatient admissions which are either self-pay patients or Medicaid enrollees. Slightly more than 20 percent of inpatient admissions occur in hospitals where with 30 percent of inpatient admissions are self-pay patients or Medicaid enrollees. These are hospitals that have the most to gain from health reform. At the other end of the spectrum are approximately 10 percent of hospitals—but only about seven percent of all admissions—where less than 10 percent of inpatient admissions comes from Medicaid or self-pay patients. These hospitals will have less expectation of change. When comparing the safety net scale for states by their Medicaid expansion status, we find they are basically similar, but non-expanding states have a higher proportion of discharges in hospitals where at least 20 percent of admissions are from Medicaid or self-pay patients.

The HAMP data also show how payer mix differs by these safety score categories for both inpatient admissions as well as admissions to the emergency department. Additionally, data show the payer mix by safety net score level for a subset of diagnoses that are considered to be sensitive to insurance status. Hospitalizations for short-term complications of diabetes, for example, are designated as a Prevention Quality Index (PQI) by the Agency for Healthcare Research and Quality, meaning that appropriate ambulatory care could potentially prevent this type of hospitalization. While this diagnosis overall accounts for less than 1 percent of inpatient admissions, it comprises twice as high a share of admissions in hospitals in the highest safety net category. Knee replacements on the other hand, are overwhelmingly funded by Medicare and commercial insurance, and comprise nearly 4 percent of admissions in hospitals with the lowest safety net score, as compared with less than one percent in hospitals with the highest score.

A subset of reasons for emergency department (ED) visits which are considered primary care treatable are also being monitored—headaches, urinary tract infections, and upper respiratory issues. Together these three categories comprise nearly 5 percent of ED visits, and interestingly this share does not differ greatly by safety net score, although as can be seen, the payer mix varies greatly by safety net score.  

When the first quarter of 2014 discharge data for these hospitals becomes available, we will be interested in monitoring trends in volume by payer, for all admissions and visits, but also for these specific diagnoses. We can see from these baseline data that there is significant variation, both between states and among hospitals within states.

Hospitals with relatively little uncompensated care may be less likely to participate in networks for Marketplace plans, which would further limit the impact of coverage expansion on their volume and payer mix. Since hospitals differ greatly in payer mix in general and their exposure to uncompensated care in particular, the impact of health reform will not be experienced equally.