Estimates of HIV infection among injecting drug users in Glasgow, 1985-1990.

Abstract

To use research and surveillance studies in Glasgow (Scotland, UK) to estimate the number of current injectors infected with HIV, the total number of injectors infected up to the end of 1990 and the recent incidence of infection. (A) Prevalence of injecting drug use was estimated using log-linear modelling. (B) Prevalence of HIV infection was determined from voluntary testing of a community-wide sample of injectors. (C) The number of infected current injectors was predicted by combining the distributions generated by (A) and (B). (D) Data on known HIV-positive injectors were used in conjunction with (C) to forecast the cumulative number of infected injectors. The number of current injectors was estimated to be 9400; the prevalence of HIV infection among 447 injectors recruited to the HIV prevalence study during 1990 was 1.1%. From these data, the number of HIV-positive current injectors in 1990 was estimated to be between 52 and 138. Between 1985 and 1990, 110 known HIV-positive injectors were registered or received treatment in Glasgow for HIV-related diseases; the total number of cases estimated to have occurred during this period was between 110 and 300. The incidence of infection in Glasgow during 1990 was likely to have been low in light of the finding that only one case in the prevalence study had not previously been diagnosed HIV-positive. Linkage of datasets from a variety of sources and studies has enabled the substantial refinement of estimates of the number of injectors and the proportion infected with HIV in Glasgow up to 1990.