The impacts of the 340B Program on health care quality for low-income patients.

OBJECTIVE: To assess the effects of hospital 340B eligibility on quality of inpatient care provided to Medicaid and uninsured patients and for all patients. DATA: Agency for Health Care Research and Quality's Healthcare Cost and Utilization Project State Inpatient Data, Hospital Cost Reporting Information System Data, Office of Pharmacy Affairs Information System Data, and American Hospital Association Annual Survey. DESIGN: Regression discontinuity design comparing hospitals just above the DSH percentage program eligibility threshold to those just below. Quality measures include all-cause mortality and 30-day readmission rates as well as condition-specific measures. DATA EXTRACTION: Inpatient data from general acute care hospitals from 2008 to 2014 in 15 states. Data linked on hospital 340B eligibility and participation. PRINCIPAL FINDINGS: We did not find discontinuities in inpatient care quality across the Program eligibility threshold for Medicaid and uninsured patients; specifically, on all-cause mortality (beta = -0.04 percentage points, 95% CI: -0.16, 0.08), 30-day readmission rates (beta = -0.16 percentage points, 95% CI: -0.81, 0.5), or other measures. Among insured and non-Medicaid patients, we found discontinuities for acute myocardial infarction (beta = -0.87 percentage points, 95% CI: -1.55, -0.2) and postoperative sepsis (beta = -0.15 percentage points, 95% CI: -0.23, -0.07) mortality. CONCLUSIONS: 340B Program participation has not demonstrated improved quality of inpatient care among Medicaid or uninsured patients.

Hospital concentration and low-income populations: Evidence from New York State Medicaid.

While a large body of evidence has examined hospital concentration, its effects on health care for low-income populations are less explored. We use comprehensive discharge data from New York State to measure the effects of changes in market concentration on hospital-level inpatient Medicaid volumes. Holding fixed hospital factors constant, a one percent increase in HHI leads to a 0.6% (s.e. = 0.28%) decrease in the number of Medicaid admissions for the average hospital. The strongest effects are on admissions for birth (-1.3%, s.e. = 0.58%). These average hospital-level decreases largely reflect redistribution of Medicaid patients across hospitals, rather than overall reductions in hospitalizations for Medicaid patients. In particular, hospital concentration leads to a redistribution of admissions from non-profit hospitals to public hospitals. We find evidence that for births, physicians serving high shares of Medicaid beneficiaries in particular experience reduced admissions as concentration increased. These reductions may reflect preferences among these physicians or reduced admitting privileges by hospitals as a means to screen out Medicaid patients.