Study: Medicaid Expansion Under the ACA Has Not Improved Quality in Safety-Net Hospitals

As some governors and state legislatures have pushed ahead since the passage of the Affordable Care Act (ACA) to expand Medicaid coverage, the main focus has naturally been on expanding access to care for Medicaid recipients. Until now, few researchers have looked at quality outcomes concerns.

Now, a new research study published in JAMA Internal Medicine online has found that, essentially, the answer of whether Medicaid expansion has led to the enhancement of patient outcomes, the diminution of hospital-acquired infections, or improved HCAHPS (Hospital Consumer Assessment of Healthcare Providers and Systems Survey) scores, in the safety-net hospitals that deliver a huge portion of the inpatient care to Medicaid recipients, is clear: there has been no documented improvement in quality along any of those dimensions.

In “Association of Medicaid Expansion With Quality in Safety-Net Hospitals,” Paula Chatterjee, M.D., M.P.H., Mingyu Qi, MSc., and Rachel M. Werner, M.D., Ph.D., write that “Safety-net hospitals (SNHs) operate under limited financial resources and have had challenges providing high-quality care. Medicaid expansion under the Affordable Care Act led to improvements in hospital finances, but whether this was associated with better hospital quality, particularly among SNHs given their baseline financial constraints, remains unknown.” As a result, the three researchers decided to statistically compare the performance of safety-net hospitals along several key dimensions, during the period from 2012 to 2018, comparing the performance of 495 safety-net hospitals in states that had expanded Medicaid with the performance of 316 safety-net hospitals in states that had not expanded Medicaid, for a total of 811 safety-net hospitals studied. The study was conducted from January to November 2020.

The researchers note that “The primary outcome was hospital quality measured by patient-reported experience (Hospital Consumer Assessment of Healthcare Providers and Systems Survey), health care–associated infections (central line–associated bloodstream infections, catheter-associated urinary tract infections, and surgical site infections following colon surgery) and patient outcomes (30-day mortality and readmission rates for acute myocardial infarction, heart failure, and pneumonia). Secondary outcomes included hospital financial measures (uncompensated care and operating margins), adoption of electronic health records, provision of safety-net services (enabling, linguistic/translation, and transportation services), or safety-net service lines (trauma, burn, obstetrics, neonatal intensive, and psychiatric care).”

The researchers write that, “In this difference-in-differences analysis of a cohort of 811 SNHs, no differential changes in patient-reported experience, health care–associated infections, readmissions, or mortality were noted, regardless of Medicaid expansion status after the Affordable Care Act.”

On the other hand, they write, “There were modest differential increases between 2012 and 2016 in the adoption of electronic health records”: they found that safety-net hospitals went from a level of 99.4-percent EHR adoption to 99.99 percent in non-expansion states, while safety-net hospitals went from a level of 94.6 percent to 100 percent in expansion states. They also found that while the average number of inpatient psychiatric beds available to patients actually declined from 24.7 to 23.9, in non-expansion states, the number of inpatient psychiatric beds available to patients increased from 29.3 to 31.4, in expansion states.

Looking at clinical outcomes in some core areas, the researchers write that “The SNHs in expansion states demonstrated a small improvement in readmissions owing to pneumonia in the preexpansion period vs SNHs in nonexpansion states (preexpansion period mean [SD], 17.8 [1.5]; coefficient for trend, −0.16; P = .09) (eFigure 3C and eTable 5 in the Supplement). In addition, SNHs in nonexpansion states showed a small preexpansion period improvement in mortality associated with acute myocardial infarction compared with those in expansion states (preexpansion period mean [SD], 14.9 [1.4]; coefficient for trend, 0.22; P = .02). Formal tests of preexpansion trends did not reach statistical significance.”

Further, they write, “We found no statistically significant differential changes in patient-reported experience, health care–associated infections, readmissions, or mortality between SNHs in expansion and nonexpansion states after Medicaid expansion.”

The researchers also take care to frame carefully what they’ve found “This study has several limitations,” they note. “First, there is no universal definition of an SNH. This lack of a formal definition remains a limitation of studies reported and for policymakers by making it difficult to design policies that support hospitals caring for the bulk of uninsured and underinsured patients. Nonetheless, we tested 3 definitions reported in the literature and found largely consistent findings. Second, our definition of quality may be limited, as we focused on measures that have improved among some hospitals after incentivization. Third, Medicaid expansion may have had multiple simultaneous downstream consequences that affect hospital care, which could have limited our ability to detect changes in quality. Safety-net hospitals may have dedicated any increase in financial resources toward supporting patient and community services in ways that are not reflected in available data. The SNHs may be increasing staffing capabilities, financing debts, or dedicating more spending toward community benefits. Our study was limited in its capacity to evaluate these other dimensions that may have also improved patient care. In addition, data on hospital finances are often incomplete and skewed, which may have prevented us from detecting changes. However, earlier work comparing HCRIS [data derived from the Rand Corporation’s Healthcare Provider Cost Reporting Information System] with internal hospital audits suggests that since 2010, HCRIS provides reliable reflections of hospitals’ financial circumstances.”

In sum, the researchers write, “In this difference-in-differences study, we found little evidence that SNHs in expansion states improved quality compared with SNHs in nonexpansion states since Medicaid expansion. Fostering high-quality safety-net systems will require greater understanding of the factors influencing their challenging circumstances.” And, they add, “The results of this study suggest that safety-net hospitals may require ongoing support for quality improvement in the post-Medicaid expansion era.”