Ann Meadow, ScD¹, Marian V. Wrobel, PhD², and Henry B. Goldberg². (1) Office of Research, Development, and Information (ORDI), Center for Medicare and Medicaid Services, 7500 Security Boulevard, C3-19-26, Baltimore, MD 21244-1850, 410-786-6602, AMeadow@cms.hhs.gov, (2) Abt Associates, Inc., 55 Wheeler St., Cambridge, MA 02138

Medicare’s prospective payment system for home health services (HH PPS), implemented in 2000, allowed for cost outlier payments when patients’ service needs are very high. Outliers are approximately 3% of the 60-day payment episodes. Outliers should result from unpredictable clinical circumstances not captured by the PPS’s case mix model. Information about outliers can help policymakers understand whether the payment system is operating as intended, and can help Medicare develop avenues for payment refinements.

Data and Methods

We examine rate ratios for outlier and nonoutlier episodes for selected characteristics of patients and counties where patients reside. For the nonoutlier sample, we use all 2001 claims for a random 2% sample of Medicare home health users who did not have any outlier episodes that year (n=74,274). For the outlier sample, we use a 10% random sample of outlier episodes (n=10,655). Claims are linked to the Area Resource File and other administrative databases.

Results

Outlier episodes involve beneficiaries covered under disability insurance (who are younger than 65) 75% more frequently than do nonoutlier episodes. Outliers are 25% less likely to involve beneficiaries age 85+. Racial and Hispanic minorities are 44% more likely in outlier episodes. Outliers are associated with large metropolitan areas of 1 million population or more (rate ratio=1.27), and rural county residence is half as likely in outlier cases. Outliers are 50% more likely than nonoutliers to occur in counties with poverty rates of 16.8% or higher. Regionally, the New England Census Division, New York, and California are associated with outlier status (RR>=1.50).

Primary diagnosis indicates outliers are strongly associated with medical care complications such as surgical wound infections (RR=6.25), diabetes (RR=5.16), and skin conditions including chronic ulcers and cellulites (RR=2.92). Summary functional status does not differ by outlier status, but a measure of medical care needs indicates outliers are twice as likely to be in the most-severe category. Services are primarily oriented towards skilled nursing and home health aides for outlier episodes. Despite the similar functional profile for both groups, the use rate for home health aide services is 50% higher for outlier cases. The use rate for physical therapy is about one-third lower.

Conclusion

Some characteristics of outliers are associated with social disadvantage. Clinical and services data suggest outlier cases arise primarily from medical as opposed to rehabilitative needs. Certain variations suggest health system, social, and financing factors may also be involved in outlier incidence.

Learning Objectives:

• Understand Medicare's outlier payment policy for home health services.
• Describe important clinical and social characteristics associated with outlier cases.

Keywords: Medicare, Financing

Presenting author's disclosure statement:
Organization/institution whose products or services will be discussed: federal Medicare insurance program
I do not have any significant financial interest/arrangement or affiliation with any organization/institution whose products or services are being discussed in this session.