Making too much of too little: interpreting HIV surveillance data

In late 1994 monthly data on Victorian HIV diagnoses for the year to the end of November 1994 was made available to a community organisation in response to a formal request. These data showed an increase in the number of HIV diagnoses towards the end of 1994 and was used by a community organisation as the basis of a press release highlighting the potential for an 'explosive' rise in HIV infection. The objective of this paper is to examine, using a case study approach, interpretation and use of routine published and unpublished surveillance data on HIV diagnoses and incidence in Victoria. Quarterly (January 91 to December 96) and monthly (January 94 to December 96) surveillance data were reviewed, by exposure category and seroconversion history and compared with findings from late 1994. There was an increase in HIV diagnoses in Victoria towards the end of 1994. However the number of new HIV diagnoses determined monthly, by virtue of small numbers, is subject to inevitable fluctuations. There continues to be confusion about the meaning of HIV surveillance data, even among community organisations involved with AIDS. Care needs to be taken in interpreting surveillance data to ensure that its limitations are understood. (author abstract)