Remodelled epidemiological data published in the April edition of AIDS suggests that by 2004 there may be almost 9,000 more individuals living with HIV in England, Wales and Northern Ireland than previously thought, and that there are more heterosexuals receiving HIV care than gay men.
UK epidemiologists remodelled the data due to their concerns that the impact of HAART and the global HIV pandemic meant that the linear model used to estimate the number of HIV infections in England, Wales and Northern Ireland was no longer accurate. Therefore, using adjusted data collected through the annual survey of prevalent HIV infections diagnosed (SOPHID) between 1996 and 2001, they remodelled likely changes in HIV prevalence in England, Wales and Northern Ireland between 2001 and 2004 using a negative binomial model.
The SOPHID data from 1995-2001 was adjusted upwards to take account of under-reporting and non-attendance at clinics.
Using both the linear regression and binomial models, the investigators calculated the total number of individuals living with HIV, increases due to both heterosexual and homosexual sex, and the proportion of infected individuals living within or outside London.
Adjusted data showed that the total number of individuals with diagnosed HIV increased by 94% between 1995 and 2001 (13,880 to 26,975), by 89% in London (8,529 to 16,136) and by 103% outside the capital (5,351 to 10,839). The number of infections in gay men increased by 62% (8,791 to 14,234), and by 213% in heterosexuals (3,550 to 11,129).
Using the negative binomial model, the investigators estimated that total HIV prevalence increased between 2001 and 2004 by 56% (26,975 to 42,047). However, the linear model only predicted a 25% increase over this period (26,975 to 33,680).
The models also produced different results when the investigators looked at likely increases in HIV prevalence by region and mode of transmission. The binomial model predicted that there would be a slower increase in total HIV prevalence in London than in the rest of England, Wales and Northern Ireland (54% versus 60%). In contrast, the linear model predicted that HIV prevalence would increase by 26% in London and 24% in the rest of England, Wales and Northern Ireland.
New HIV cases caused by sex between men would increase by 32% between 2001-2004, according to the negative binomial model, with a 92% increase in heterosexual cases predicted during the same period. However, the linear model predicted that the increase in new cases amongst gay men would only be 22%, with a 32% increase in new heterosexual HIV infections.
Using the negative binomial model, the investigators estimated that in 2004, for the first time, there would be more heterosexuals than gay men receiving HIV care (21,335 versus 19,011). However, the linear model predicted that in 2004 there would still be a clear majority of gay men receiving HIV care (17,344 versus 14,638).
The investigators conclude that increases in HIV infections following either of these models “will have serious implications for service providers. In particular, the predicted increases in diagnosed HIV will have an impact on GUM clinics in terms of workload, adding to the burden caused by the rising numbers of other sexually transmitted infections.”
Further information on this website
Record 7,000 new HIV diagnoses in UK in 2003 - news story
Rice BD et al. Prevalent diagnosed HIV in England, Wales and Northern Ireland: adjusted totals 1996-2001 and extrapolations to 2004. AIDS 18: 927-932, 2004.