Every year, the Centers for Medicare & Medicare Services (CMS) publishes in the Federal Register a draft regulation describing how it proposes paying hospitals for the inpatient care they provide to their Medicare patients in the coming fiscal year. The proposed inpatient prospective payment system regulation for FY 2015 was published on May 15, and as always, CMS invited interested parties to submit written comments.
The National Association of Urban Hospitals has always found CMS to be receptive and even responsive to its comments and therefore takes the opportunity to submit detailed comments and suggestions about the agency’s annual proposal.
NAUH is presenting excerpts from its comment letter to CMS. The subjects and the dates they will be published are:
June 26 – the size of the Medicare DSH pool
June 27 – the manner in which CMS uses CBO estimates of changes in insurance status
Today – the methodology for distributing Medicare DSH funds
July 1 – the need for appropriate risk adjustment in the hospital readmissions reduction program
July 2 – a much-needed adjustment to the methodology employed by the hospital readmissions reduction program
July 3 – hospital inpatient rates
July 7 – short hospital stays (two-midnight rule)
NAUH’s complete comment letter to CMS can be found here.
The Methodology for Distributing Medicare DSH Funds
NAUH is pleased about the manner in which CMS has decided to measure the uncompensated care portion of hospitals’ Medicare DSH payments. As we have conveyed to you over the past three years, the data from the S-10 data form of the Medicare cost report that appeared to be the natural foundation for such a calculation is, in NAUH’s view, seriously flawed and has not improved in recent years.
The S-10 seeks to quantify the uncompensated care that hospitals provide, but historically, “uncompensated care” is not a universally agreed-upon concept and hospitals have not reported their uncompensated care uniformly. CMS’s own consultants on how to quantify hospital uncompensated care noted that “…we found variation in how existing programs and entities define uncompensated care.” Those definitions, the consultants noted, vary among federal programs, the states, rating and research organizations, and provider organizations. While charity care and bad debt are always included in such definitions, they noted that “Some entities also include payment shortfalls from government-funded plans, or third party payers.” CMS’s consultants also wrote that “Uncompensated care is most often defined as charity care plus bad debt but may include government and/or commercial payer shortfalls.”
The use of “may” in that observation is significant: it acknowledges that different entities interpret uncompensated care in different ways. Different hospitals have different, and sometimes significantly different, charity care policies, and that, in turn, affects where they report their costs on the S-10. One hospital with a charity care policy that includes care for Medicaid patients whose stays exceeds a day limit may include that Medicaid shortfall in its charity care line on the S-10, for example, while another hospital, with a different policy, may include that shortfall in the Medicaid line of the S-10. In this example, both hospitals are providing the same care for the same reimbursement but only one shows the shortfall as uncompensated care if payer shortfalls are not considered. This is only one of a number of ways, including care involving Medicaid waiver populations and non-patient-specific funding streams, that NAUH believes different hospitals, even neighboring hospitals, end up categorizing and reporting uncompensated care in different ways, and hospitals in different regions and different states do the same.
Numerous examples demonstrate the weakness of S-10 data and its continued unsuitability for use in the implementation of important public policy. NAUH appreciates that CMS appears to agree through its decision to use instead what we believe to be a more tested and reliable proxy for this vital calculation: hospitals’ Medicaid and Medicare SSI days. On behalf of private, non-profit urban safety-net hospitals, we thank you for this decision.
While NAUH appreciates the complexity of the challenge of how best to calculate hospitals’ FY 2015 Medicare DSH payments, we also recognize that this methodology, now in its second year of use, may very well be temporary, leaving us concerned about CMS’s future intentions. NAUH is especially interested in how S-10 data may be part of those future plans. We continue to believe that any use of S-10 data in its current form would pose an enormous problem. Improving the S-10, we believe, will require changes in the form, changes in the instructions for completing the form, and a period of auditing data submitted by hospitals to ensure that these changes are achieving their objectives.
NAUH understands that CMS is still working on this, appreciates your efforts, and welcomes the opportunity to assist in any way we can. Uniformity of uncompensated care reporting, so lacking today, is absolutely essential to this process because of the manner in which the Medicare DSH pool is divided: proportional to eligible hospitals’ uncompensated care costs. If some hospitals report selected costs as uncompensated that other hospitals categorize differently, this could result in an unfair distribution of Medicare DSH payments. The objective of this endeavor should be fairness in the distribution of Medicare DSH resources, and NAUH supports CMS’s decision to use a proxy for uncompensated care until it finds a way to produce the uniformity of reporting that leads to such fairness.
In the meantime, we again wish to express our thanks for CMS’s decision to use the proxy you have chosen instead of S-10 data and urge you to continue using this proxy until you have a better, proven, verifiable method for measuring the uncompensated care hospitals provide.
* * *
Tomorrow – The need for appropriate risk adjustment in the hospital readmissions reduction program
NAUH’s complete comment letter to CMS can be found here.
