Review of death certificates to assess completeness of AIDS case reporting.

To assess the level of reporting of acquired immunodeficiency syndrome (AIDS) cases, the authors reviewed death certificates for periods of 3 months during July through December 1985 in each of four cities: Washington, DC, New York City, Boston, and Chicago. Since reporting began in 1981, these cities have reported 38 percent of all AIDS cases in the United States. Death certificates were selected and matched to the AIDS surveillance registries in each city, and medical records of those not on the AIDS registry were reviewed to determine if AIDS had been diagnosed. The estimated completeness of AIDS case reporting to AIDS surveillance systems was high in all four cities (ranging from 83 percent to 100 percent). The unreported cases were similar to reported cases with respect to sex, race, risk factor, and specific diagnosis. Of the causes of death examined, AIDS, Pneumocystis carinii pneumonia, and Kaposi's sarcoma were predictive of AIDS as defined by the CDC case definition. However, 77 of 588 deaths (13 percent) attributed to 1 of these 3 causes occurred in cases that were presumptively AIDS but did not meet the diagnostic requirements to be classified as AIDS for reporting purposes. A review of death certificates provides an easy and rapid means of evaluating surveillance efforts and can be a useful adjunct to other methods of surveillance for AIDS.