Objective: To evaluate AIDS case reporting in the Spanish regions as compared with the AIDS deaths registered in mortality statistics, in order to identify possible deviations indicative of underreporting.
Methods: We carried out an ecological analysis taking each region as a unit. We compared incidence and mortality obtained from the AIDS reporting system with the HIV/AIDS deaths obtained from the mortality statistics of the Spanish Institute for Statistics. The 1986-1998 period was analysed globally, then the analysis was repeated for the 1995-1998 period.
Results: In the 1986-1998 period there was a good correlation (r = 0.93) between the AIDS incidence rates and HIV/AIDS mortality rates obtained from death statistics. Some regions presented an AIDS incidence lower than expected according to their mortality rate when it was compared with the national average, with Asturias (27%), Comunidad Valenciana (26%), Andalucia (20%), Ceuta (18%) and Cantabria (13%) standing out. Taking as a reference the five regions with the highest completeness, the underreporting in Spain was of 13%. In the 1995-1998 period these deviations increased moderately. In the 1986-1998 period, 18.9% less deaths were notified to the AIDS reporting system in comparison with death statistics, showing a great variability between regions.
Conclusions: In Spain the reporting level of AIDS cases is acceptable for the aims of surveillance, although in some regions measures to improve it should be taken. The notification of deaths to the AIDS reporting system presents great deficiencies.