Objective: To examine the reporting of AIDS-defining illnesses using t
wo national surveillance systems. Methods: Comparison of AIDS indicato
r diseases reported to the national AIDS reporting system (ARS) for AI
DS cases diagnosed from January 1990-December 1992 among individuals a
ged >13 years in 10 metropolitan areas, with that observed in the Adul
t/Adolescent Spectrum of HIV Disease (ASD) project, a surveillance pro
ject that monitors the clinical diagnoses of HIV-infected individuals
receiving medical care. Results: In the 10 metropolitan areas, 39 265
individuals with AIDS were reported to ARS, and 5969 with AIDS had med
ical record reviews as part of ASD. At initial AIDS diagnosis, the num
ber of indicator diseases reported to ARS was almost identical to the
number observed in ASD (mean number of diagnoses, ARS 1.3; ASD 1.2). H
owever, ASD recorded a greater number of diagnoses over time than ARS
(mean number of indicator diagnoses >12 months after initial diagnosis
, ASD 2.3; ARS 1.4). Conditions that typically occur late in the cours
e of AIDS such as Mycobacterium avium infection and cytomegalovirus di
sease, were more frequently recorded by ASD than by ARS. Conclusion: A
RS provides complete, population-based information on the frequency of
AIDS-defining conditions at initial diagnosis. However, specialized s
urveillance projects such as ASD are needed to accurately describe sub
sequent AIDS-defining conditions.