Objectives To measure the potential savings from medical nutrition therapy (MNT) and to estimate the net cost to Medicare of covering these services for Medicare enrollees. This includes developing an estimate of the cost of providing medical nutrition services to the Medicare population and estimating the savings in hospital and other spending resulting from the use of these services.

Design Analysis of longitudinal data from the Group Health Cooperative of Puget Sound (Seattle, Wash) for persons aged 55 years and older who have coverage for MNT services.

Subjects/setting Persons aged 55 years and older who had diabetes (n=12,308), cardiovascular disease (n=l0,895), or renal disease (n=3,328) and who were covered under the Group Health Cooperative of Puget Sound, including Medicare beneficiaries enrolled in the plan's Medicare risk contract program. Extrapolation to the US Medicare population is based on data for persons served by the Group Health Cooperative of Puget Sound.

Intervention The use of MNT.

Main outcomes measure Differences in health care utilization levels of persons with diabetes, cardiovascular disease, and renal disease who do and do not receive MNT. Differences in utilization were estimated for hospital discharges per calendar quarter, physician visits per quarter, and other outpatient visits per quarter.

Statistical analyses performed Multivariate regression models of changes in utilization for persons after they receive MNT services.

Results Our analysis showed that MNT was associated with a reduction in utilization of hospital services of 9.5% for patients with diabetes and 8.6% for patients with cardiovascular disease. Also, utilization of physician services declined by 23.5% for MNT users with diabetes and 16.9% for MNT users with cardiovascular disease. The net cost of covering MNT under Medicare is estimated to be $369.7 million over the 1998 through 2004 period. The total cost of benefits is estimated to be $2.7 billion over this period. This would be partially offset by estimated savings of $2.3 billion resulting in net costs of $369.7 million. The program would actually yield net savings after the third year of the program, which would continue through 2004 and beyond.

Conclusion After an initial period of implementation, coverage for MNT can result in a net reduction in health services utilization and costs for at least some populations. In the case of persons aged 55 years and older, the savings in utilization of hospital and other services will actually exceed the cost of providing the MNT benefit. These results suggest that Medicare coverage of MNT has the potential to pay for itself with savings in utilization for other services. J Am Diet Assoc. 1999;99:428-435.