Trivial Pursuit. (View on Washington).

There was a lot of hope last fall that the sudden drop in Medicare reimbursement rates in October would be only a temporary setback for long-term care providers. Nursing homes were mobilized to educate their legislators. Every member of the Massachusetts delegations to the House and Senate signed a letter petitioning their congressional colleagues to act quickly to save their SNFs. With such a broad-based effort, it seemed that reason would prevail and a quick fix would emerge from Congress, despite the victory of budget-conscious Republicans in the midterm elections.

Some of that optimism evaporated in November when the General Accounting Office (GAO), a congressional agency that exists to audit and evaluate the effects of legislation, released a report entitled Skilled Nursing Facilities: Available Data Show Average Nursing Staff Time Changed Little after Medicare Payment Increase. The report, accessible online at www.gao.gov/new.items/d03176.pdf, can be interpreted as implying that SNFs squander increased reimbursements authorized by Congress. The American Health Care Association (AHCA) responded immediately with its own analysis casting doubt on the accuracy of the GAO's conclusions, but the damage had been done. Restoration of the reimbursement rates has again become a controversial issue with a doubtful outcome.

The dispute over the GAO report has been described as a debate over research methods. In reality, it reflects the often irrational way that policies are sold to Congress.

In 2000, the long-term care industry was fighting to reverse the devastating cuts in Medicare reimbursement levels imposed by the Balanced Budget Act of 1997. The industry had already enjoyed some success in lobbying Congress to raise reimbursement levels for the most expensive care--patients classified in "high-cost" Resource Utilization Groups (RUGs). One argument offered to Congress in support of raising other reimbursement rates was that they were too low to permit sufficient staffing to deliver quality care.

Congress responded to this argument with a provision of the Benefits Improvement and Protection Act (BIPA) that authorized a 16.6% increase in the nursing component of the SNF Medicare reimbursement rate. The nursing component varied at the time from 26 to 74% of the total daily reimbursement, depending on patient RUG classification, so that the actual increase in reimbursement could be anything between 4 and 12.3% per Medicare patient. However, rather than "protecting" even this increase, BIPA provided a temporary solution: The new rate would be in effect only from April 1, 2001, until October 31, 2002. During that time, the GAO would study how increased reimbursement affected patient care.

In 2000, few stakeholders seriously expected that the GAO study might influence the outcome of sustaining legislation in 2002. Even under the best of circumstances, the GAO could not have completed its study until the early fall of 2002. Most congressmen at the time assumed that the economy would still be generating large budget surpluses that would easily support higher nursing-staff reimbursement and other Medicare reforms. As it turned out, the GAO did not release its report until after the expiration of the BIPA increase, when Congress had been expected to have already "resolved" the reimbursement problem with permanently higher rates.

Not only was the report mistimed, as it turned out, but its findings were controversial. The GAO study relied exclusively on data provided by the CMS Online Survey Certification and Reporting System (OSCAR), a source that is generally considered to be inaccurate for individual nursing homes but "reasonably" accurate for the industry in general. Because OSCAR reports the total number of nursing hours and of patient days, the GAO calculated that SNFs initially provided an average of 3 1/2 hours of nursing time per patient day. Performing the same calculation for OSCAR reports between April and December 2001, the GAO generally found very small changes in the average nursing time:

* slightly more than 2 minutes per day for all SNFs

* roughly 3 minutes per day for not-for-profit, nongovernmental SNFs

* slightly more than 5 minutes per day for government-operated SNFs

The GAO reported two important, but not necessarily helpful, exceptions to these unimpressive results. First, SNFs that initially had unusually low staffing increased the amount of nursing time spent with each patient by an average of 15 minutes per day; however, the GAO report notes that a similar increase was observed among poorly staffed SNFs before the reimbursement change occurred, indicating that Medicare reimbursement itself was not the primary factor. Second, SNFs in the states of Arkansas, Nebraska, North Dakota, and Oklahoma also increased their average per-patient nursing time by 15 or more minutes per day, but in three of these four states, the change was attributed to statewide Medicaid reimbursement policies.

Representatives of AHCA and, to a lesser extent, the American Association of Homes and Services for the Aging and the American Hospital Association have raised several questions about the validity of the GAO study. AHCA has even duplicated the GAO study methods, cornparing OSCAR data from reporting periods through August 2002. AHCA concluded, "Eighty percent of the CMS data GAO relied on for its report is from the period between May and August 2001--long before nursing homes, given the time frames of the Medicare claims and payment process, would have received reasonable amounts of monies from the add-on." Even AHCA's "corrected" data, however, found that the typical freestanding SNF added only one nursing staff member per 100 patients, resulting in an average of 4.8 minutes of more nursing contact per patient-day.

The reality is that Congress probably was never serious about using increased Medicare reimbursement to improve the staff/patient ratio in the long-term care field. Medicare pays for barely one of nine SNF residents. Reimbursement for most long-term care patients is through Medicaid and self-pay, which explains why changes in statewide Medicaid policies had a significant effect on staffing levels in Arkansas, North Dakota, and Oklahoma (and perhaps elsewhere). Even if BIPA had doubled the nursing staff reimbursement, this would not have applied to nearly 90% of all SNF residents. 'What's more, the most active stakeholders certainly understood that BIPA might not affect staffing levels at all because the legislation never required SNFs to spend any of the increased reimbursement to improve the nurse/patient ratio.

It seems doubtful that Congress understands the reimbursement/quality connection. The truth behind the controversy over how BIPA reimbursement was used is that the long-term care field's major need at this moment is survival rather than quality improvement. Nursing facilities are heavily reliant upon Medicaid; for too many, Medicaid is a money loser. If much of the BIPA increase was used by providers to pay dividends to stockholders or pay off loans, it could reasonably be interpreted as an investment in solvency. With government-financed nursing care a marginal business at best, the debate over whether SNFs achieved an average of 2 or 5 minutes of extra nursing time per patient day would seem to be a game of Trivial Pursuit[R].

To comment on this article, please send e-mail to [e-mail].