How Hospitals Respond to Changes in Medicare Reimbursements

Since 1984, hospitals have been reimbursed for care provided to Medicare patients under the Prospective Payment System (PPS). Under this system, hospitals receive a fixed payment for each patient that is determined by the patient's diagnosis-related group (DRG) at the time of admission; thus, reimbursement is unaffected by the hospital's actual expenditures on the patient.

The motivation for switching from the old cost-plus system, in which hospitals were reimbursed based on actual expenditures, to the PPS was to provide hospitals with better incentives to contain costs and increase efficiency. Yet the PPS may also provide incentives for hospitals to behave in undesirable ways. For example, if certain diagnoses are more profitable because the reimbursement is larger relative to typical treatment costs, hospitals may respond by encouraging some types of admission and discouraging others, a potentially dangerous practice. Hospitals may also practice "upcoding," switching patients from low- to high-paying DRGs, which would not affect patient outcomes but would increase Medicare costs. On the other hand, hospitals may respond by providing higher-quality care in high-paying DRGs in order to attract lucrative patients.

In How Do Hospitals Respond to Price Changes? (NBER Working Paper 9972), Leemore Dafny examines the effect of changes in DRG-specific reimbursement levels (DRG prices) on coding, admissions volume, and intensity of care. Typically, it is quite difficult to accurately measure the effect of DRG price increases, as prices are usually adjusted in response to changes in hospital costs, generating a relationship between DRG prices and expenditures or intensity that is not necessarily causal.

In her analysis, the author makes use of a change in DRG prices that was unrelated to changes in costs. Forty percent of DRG codes come in pairs - for each diagnosis such as cardiac arrhythmia, there was one code for patients aged 70 and above or with complications and a second code for patients under 70 without complications. Analysis suggested that costs for complication-free patients were similar regardless of age, so the age criterion was eliminated in 1988 and DRG prices were recalibrated. This resulted in an average 11% increase in DRG prices for the top codes in each pair (those with complications) and an average 6% decrease in DRG prices for the bottom codes.

Using the 1985-1991 Medicare Provider Analysis and Review data, a 20% sample of all hospitalizations of Medicare enrollees, the author examines how hospitals responded to these relative price changes. She finds that hospitals upcoded patients to DRG codes with large price increases, garnering an estimated $330-$425 million in additional reimbursement annually. This response was sophisticated, with more upcoding in DRGs where the spread between top and bottom codes had increased more. For-profit hospitals were found to be more likely to engage in upcoding.

Next, the author examines whether hospitals responded by altering admissions volume or intensity of care, as measured by total costs, length of stay, number of surgeries, number of ICU days, and in-hospital deaths. She finds little evidence of changes in admissions volume or intensity of care for patients in the affected DRGs as a result of the price changes. However, she finds strong evidence that hospitals spent the extra reimbursement funds on patient care in all DRGs. This suggests that hospitals may find it difficult to improve quality in specific diagnoses only, implying that hospitals will compete on overall quality rather than trying to specialize.

This research was supported by the National Science Foundation, the National Bureau of Economic Research, and the National Institute on Aging. It was summarized by Courtney Coile.

National Bureau of Economic Research, 1050 Massachusetts Ave., Cambridge, MA 02138; 617-868-3900; email:

Contact Us