Changing pattern of delays in reporting AIDS diagnoses in Australia.

To accurately monitor and predict the progress of the HIV/AIDS epidemic, it is important to adjust reported AIDS counts for reporting delays. This requires estimation of the reporting delay distribution. This paper aims to use a statistical model to identify the main factors influencing reporting delays in Australia and to adjust reported incidence data for these delays among cases of AIDS diagnosed from 1993 and reported before 30 June 1997. Reporting delays were found to vary significantly across states/territories. The influence of calendar time of diagnosis was also significant, with an overall trend toward longer delays over time. AIDS cases diagnosed in the fourth quarter of a year were reported significantly more quickly than those diagnosed in the first or third quarters. No significant differences were found due to sex, age and HIV exposure category, except people with haemophilia, in whom AIDS cases appeared to be reported more slowly. After adjusting for under-reporting and reporting delay, we found that the AIDS incidence in Australia was declining from about 1000 cases per year in 1994 to about 760 cases per year in 1996.