The usefulness of government publication of annual death rates of Medicare patients at all general hospitals in the country has been sharply challenged by a Rand Corp. study published in the Journal of the American Medical Association.

The study, which focused on two conditions, congestive heart failure and acute myocardial infarction, concluded there is little evidence of lower quality of care in hospitals that have higher death rates among Medicare patients than the broad range of hospitals.

The study judged quality of care on the basis of the medical records of selected hospitals in years 1983 and 1984.

The study said half or more of the higher death rates could be explained by purely random variations and a small portion by greater severity of illness, but that quality of care differences were not enough to explain the remaining differences.

For the past four years, the Department of Health and Human Services has been publishing hospital death rates, labeling them as one possible factor in helping determine a hospital's quality of care.

HHS's Health Care Financing Administration, which compiles the death rates, said the Rand study was not comparable to the HHS figures, which cover all disease categories, not just two, and all hospitals, not a sample.

Moreover, HCFA said its figures are more recent, starting with 1986 and then covering later years. The agency said "more work will need to be done to assess the Rand approach."