The authors hypothesized that non-specific causes of death are likely to be assigned to persons with unknown medical history or those not receiving regular health care; therefore, this category may be more frequently given as a cause of death for minorities in the U.S. than whites. To investigate these hypotheses, the authors examined death certificate data for New Mexico for the period from 1958 to 1982.

Death rates from non-specific causes, for both men and women, were higher in almost all age-gender strata for Hispanic and American Indian persons compared with non-Hispanic whites. Similarly, the age-adjusted mortality rates attributed to non-specific causes were consistently higher for minority men and women than for whites in each four-year time period.

The authors make several arguments based on these data which will require further investigation. First, it is possible that this reflects lower utilization of medical care by minorities compared with whites. Second, it is possible that this reflects fewer autopsies among ethnic minorities in New Mexico compared with whites. Finally, these data suggest that there might be substantial bias in ethnic/racial comparisons that make use of mortality data.