The Effect of Changes in Hospital Reimbursement on Nurse Staffing Decisions at Safety Net and Nonsafety Net Hospitals

Abstract

The financial savings from the Balanced Budget Act (BBA) are attractive to policy makers, but such savings come at a cost. We measure changes in nurse staffing at hospitals related to potential declines in reimbursement through the BBA. Following Hadley, Zuckerman, and Feder (1989), we define a fiscal pressure index (FPI) to measure the differential effect of the BBA. We estimate the effect of the FPI on the number of full-time equivalent registered nurses (RN) and licensed practical nurses (LPN) per adjusted patient day using American Hospital Association (AHA) data of a panel of hospitals from 1996 to 2001. The AHA data are combined with the Area Resource Files and health maintenance organizations penetration data. We control for hospital heterogeneity using fixed effects. All urban short-term general hospitals that responded to the staffing and uncompensated care questions in the AHA survey between 1996 and 2001. We define safety net hospitals as those with a high ratio of uncompensated costs to total hospital expenses (see, e.g., Zuckerman et al. 2001). We find that the nonsafety net hospitals that were most susceptible to the provisions of the BBA experienced a decline in RN staffing ratios about twice the rate of the nonsafety net hospitals that were least susceptible to the BBA. We are unable to detect an effect of the BBA on staffing at safety net hospitals. RN and LPN staffing levels per adjusted patient day declined, on average, between 1996 and 2001. Within the context of the general decline, we find that RN staffing per adjusted patient day declined even more at nonsafety net hospitals that were most susceptible to lower reimbursement related to the BBA. Thus, there was a small but statistically significant incremental effect of potential BBA losses on RN staffing at hospitals that were expected to be affected most. This incremental decline represented about a 6 percent increase in nurse workload that in isolation might not affect quality. Nevertheless, the BBA contributed to the contemporaneous trends toward higher nurse workloads that could have deleterious effects on quality. In contrast, safety net hospitals did not respond to the provisions of the BBA by reducing staffing ratios. This conclusion is tempered by the fact that we have few safety net hospitals in the sample.