In 2013, traditional Medicare—which covered roughly 38 million beneficiaries—spent about $179 billion on hospital services. To control costs and reward efficiency, Medicare pays the majority of hospitals using an approach known as the inpatient and outpatient prospective payment systems (PPS). Under a PPS, hospitals are paid a predetermined amount based on the clinical classification of each service they provide to beneficiaries. However, beginning in 1983, in response to concern that certain cancer hospitals would experience payment reductions under such a system, Congress required the establishment of criteria under which 11 cancer hospitals (see figure) are currently exempted from the inpatient PPS and receive payment adjustments under the outpatient PPS. Generally, Medicare pays these PPS-exempt cancer hospitals (PCH)—which are all teaching hospitals—based on their reported costs.
Prospective Payment System (PPS)-Exempt Cancer Hospitals (PCH), Location and Effective Date
Since PCHs were first established in the early 1980s, cancer care and Medicare’s payment system have changed significantly. Advances in techniques and drugs have increased treatment options and allowed for more localized delivery of care. Along with these developments, the primary setting for cancer care has shifted from the inpatient setting to the outpatient setting. For example, patients now typically have chemotherapy and radiation treatments without staying overnight in a hospital. In addition, in 2007, the Centers for Medicare & Medicaid Services (CMS)—the agency within the Department of Health and Human Services (HHS) that administers the Medicare program—refined the inpatient PPS methodology to better account for variation in the severity and complexity of beneficiaries in its payment calculations. Medicare’s current payment system better recognizes the resource intensity of hospital care than the system put in place in 1983.
In a February 2015 report, GAO’s analysis of 2012 Medicare data found a number of similarities and differences between PCHs and PPS teaching hospitals. For instance, unlike beneficiaries treated at PPS teaching hospitals, nearly all beneficiaries treated at PCHs had a diagnosis of cancer. However, at both PCHs and PPS teaching hospitals the health status of Medicare beneficiaries with cancer was not markedly different and the resources needed to care for inpatient cancer beneficiaries were also clinically similar.
In addition, compared with how PPS teaching hospitals are paid, the methodologies for paying PCHs provide little incentive for efficiency. Under a PPS, Medicare pays hospitals a predetermined amount based on the clinical classification of each service they provide. PPS hospitals can retain any amount Medicare pays that exceeds their costs. In contrast, as required by the exemption, Medicare pays PCHs for inpatient services based on their reported costs, subject to an upper limit, as well as potential add-on payments. For outpatient care, Medicare payments to PCHs are composed of PPS service-specific rates and an upward payment adjustment based on reported costs.
GAO reported that, had PCH beneficiaries received inpatient and outpatient services at nearby PPS teaching hospitals, Medicare may have realized substantial savings in 2012. GAO estimated that for inpatient care that year, CMS paid PCHs 42.3 percent more per discharge, on average, than it would have typically paid PPS teaching hospitals in the same geographic area to treat equally complex cancer beneficiaries. The estimated differences in Medicare payment between PCHs and local PPS teaching hospitals varied greatly across the PCHs, with the largest payment difference at 90.9 percent and the smallest payment difference at 6.7 percent. Overall, the difference between the amount Medicare paid PCHs and the estimated amount Medicare would have paid PPS hospitals for treating comparable cancer patients suggests that Medicare would have saved about $166 million in 2012.
In the outpatient setting, Medicare payment adjustments to PCHs resulted in overall reimbursements that were 37 percent higher, on average, than payments Medicare would have made to teaching hospitals under the outpatient PPS for the same set of services. Again, the size of the payment adjustment varied widely across PCHs, ranging from 13 percent to 51 percent. For the majority of PCHs, Medicare increased payments by more than 25 percent over PPS rates. GAO calculated that, if PCHs were paid for outpatient services in the same way as PPS teaching hospitals—and forgone payment adjustment amounts had been returned to the Supplementary Medical Insurance Trust Fund rather than redistributed to PPS hospitals under current requirements—Medicare would have saved about $303 million in 2012.
GAO also found that, compared with PPS teaching hospitals, the PCH group had a higher median Medicare inpatient profit margin and a similar median outpatient profit margin. In addition, GAO examined whether an increased focus on treating cancer patients affects hospitals’ profitability. Its analysis of the PPS inpatient and outpatient data showed no relationship between the share of Medicare payment derived from treating cancer beneficiaries and Medicare profit margins. Furthermore, PCHs generally had positive all payer margins—profit margins that include both private and public payers—which could be attributable, in part, to how PCHs account for administrative costs.
In its February 2015 report, GAO concluded that Medicare PPS or an alternative payment system may be reasonable for PCHs. Because Medicare’s payment methodology for PCHs lacks strong incentives for cost containment, Medicare expenditures in 2012 were substantially higher than they would have been had PCH cancer beneficiaries been treated at local PPS teaching hospitals. Until Medicare pays PCHs in a way that encourages greater efficiency, Medicare remains at risk for overspending.
 GAO used a regression model to predict what the typical Medicare payment per discharge for PCH beneficiaries would have been if they had been treated at PPS teaching hospitals in the same core based statistical area (CBSA). The average estimated payment per discharge had a margin of error of 1.0 percent at a 95 percent confidence level, with no individual CBSA margin of error exceeding 1.6 percent. The 42.3 percent overall payment difference had a margin of error at the 95 percent confidence level of plus or minus 1.5 percentage points.
 GAO estimated this savings amount within a range of plus or minus $4 million at a 95 percent confidence level. This savings estimate covers 9 of the 11 PCHs due to missing 2012 data for 2 PCHs.
 The comparison group of PPS teaching hospitals consists of those without National Institutes of Health-designated comprehensive cancer centers.
To help HHS better control Medicare spending and encourage efficient delivery of care, and to generate cost savings from any reductions in outpatient payments to PCHs, GAO recommended that Congress consider taking the following action:
GAO estimated that if, in 2012, PCH beneficiaries had received inpatient and outpatient services at nearby PPS teaching hospitals, and if the forgone outpatient adjustments were returned to the Supplementary Medical Insurance Trust Fund, Medicare may have realized annual savings of almost $500 million.
This savings estimate is composed of $166 million for inpatient services, which is estimated within a range of plus or minus $4 million at a 95 percent confidence level, and $303 million for outpatient services. The inpatient estimate did not include all PCHs due to missing data for 2012.
The information contained in this analysis is based on findings from the product in the related GAO product section. In developing those findings, GAO divided Medicare fee-for-service beneficiaries into two groups—beneficiaries with cancer and beneficiaries without cancer—using cancer diagnosis codes provided by CMS to analyze 2011 and 2012 Medicare claims data. To compare hospital characteristics, GAO analyzed 2012 Medicare cost report data. Because data limitations precluded a direct comparison of Medicare inpatient payments, GAO used a regression analysis to model payment per discharge for Medicare beneficiaries who were treated for cancer at PPS teaching hospitals located in the same geographic locations as PCHs. GAO included each beneficiary’s Medicare severity diagnosis-related group and the core based statistical area (CBSA) in which the hospital was located as independent variables. GAO used this model to predict what the typical Medicare payment per discharge for PCH beneficiaries would have been if they had been treated at PPS teaching hospitals in the same CBSA and compared the results to inpatient PCH payments per discharge. To compare Medicare outpatient payments, GAO calculated the overall percentage payment adjustment to PCHs, which represented the difference in Medicare payments between PCHs and local PPS teaching hospitals. Furthermore, GAO determined Medicare inpatient and outpatient profit margins, as well as all payer margins, for PCHs and PPS teaching hospitals. Table 14 in appendix V lists the program GAO identified that may have opportunities for cost savings.
Other than technical comments, which were incorporated as appropriate, HHS did not comment on GAO’s February 2015 report. In addition, GAO provided a draft of this report section to HHS for review and comment. HHS did not have comments on this issue.
For additional information about this area, contact James Cosgrove at (202) 512-7114 or firstname.lastname@example.org.
Unlike beneficiaries seen at teaching hospitals paid under Medicare's prospective payment systems (PPS) in 2012, nearly all beneficiaries seen at PPS-exempt cancer hospitals (PCH)—a group of 11 facilities having met certain statutory criteria—had a diagnosis of cancer. However, the health status of Medicare beneficiaries with cancer who were treated at PCHs and PPS teaching hospitals was not m...