The effect of cuts in medicare reimbursement on hospital mortality

To determine if patients treated at hospitals under different levels of financial strain from the Balanced Budget Act (BBA) of 1997 had differential changes in 30-day mortality, and whether vulnerable patient populations such as the uninsured were disproportionately affected.

Hospital discharge data from all general acute care hospitals in Pennsylvania from 1997 to 2001.

A multivariate regression analysis was performed retrospectively on 30-day mortality rates, using hospital discharge data, hospital financial data, and death certificate information from Pennsylvania.

We used 370,017 hospital episodes with one of four conditions identified by the Agency for Healthcare Research and Quality as inpatient quality indicators were extracted.

The average magnitude of Medicare payment reduction on overall net revenues was estimated at 1.8 percent for hospitals with low BBA impact and 3.6 percent for hospitals with a high impact in 1998, worsening to 2 and 4.8 percent, respectively, by 2001. Operating margins decreased significantly over the time period for all hospitals (p<.05). While unadjusted mortality rates demonstrated a disproportionate rise in mortality for patients from high impact hospitals from 1997 to 2000, adjusted analyses show no consistent, significant difference in the rate of change in mortality between high-impact and low-impact hospitals (p=.04-.94). Similarly, uninsured patients did not experience greater increases in mortality in high-impact hospitals relative to low-impact hospitals.

An analysis of hospitalizations in the Commonwealth of Pennsylvania did not find an adverse impact of increased financial strain from the BBA on patient mortality either among all patients or among the uninsured.