New HIV infections via injecting drug use (IDU) appear to be on the increase in England and Wales, according to a collaborative study from the UK's Health Protection Agency (HPA) and Imperial College London, published in the July 22nd issue of the journal AIDS. The study, which combines anonymous HIV testing data with community surveys for the first time, suggests that recent increases in HIV IDU transmission are most pronounced in younger, recent IDUs, in London. This increase in new infections coincides with a shift in UK drugs policy away from public health concerns towards a stronger focus on crime.
In the UK, harm reduction initiatives such as the provision of clean needles through needle exchange programmes (NEPs) have been relatively effective in limiting the spread of HIV among injecting drug users (IDUs). By the end of 2002, only 7% of the 56,000 diagnosed HIV infections were associated with IDU. However, there has been some recent evidence of an increase in risky injecting practices suggesting that new HIV infections amongst IDUs may be on the increase.
In order to examine trends in HIV prevalence amongst IDUs, researchers from the HPA and Imperial College, London combined data from two voluntary unlinked-anonymous survey programmes that included adults (aged 15-49) who had injected drugs in the previous four weeks.
The first is an annual survey of IDUs via drug agencies in England and Wales (ranging in number over the years between 29-59; providing advice, support, harm-reduction and/or treatment services) has been ongoing since 1990, and includes a brief self-completed questionnaire and oral fluid samples for HIV testing.
The second was a series of community-based surveys in London (1990-1993); London and seven other English cities (1997-1998); and London and Brighton (2001-2002). This was conducted in the field (e.g. street locations, homes and social venues) and included an interviewer-administered questionnaire and oral fluid samples for HIV testing. This provided the researchers with a cross-sectional data set, including almost 28,000 oral fluid samples on which to test anonymously for HIV.
Evidence of increase in HIV prevalence
HIV prevalence among IDUs in England and Wales declined from a peak of 5.9% (67 positive HIV antibody tests out of a total of 1132 samples) in 1990 to a low of 0.6% (14/2270) in 1996. It then remained stable until 2000, after which there was, say the researchers, "some evidence of an increase" to 1.4% (21/1529) in 2003.
Individuals who had been injecting for the shortest period of time (less than three years; 1.2%) and those who had been injecting for the longest period of time (more than twelve years; 2.9%) had the highest HIV prevalence in 2003. In contrast, those who had been injecting drugs between three and five years, or six and eleven years, had lower HIV prevalence (0.3% and 0.7%, respectively).
HIV prevalence was found to be higher in London (5%) compared with elsewhere in England and Wales (0.4%) and similar in women (1.8%) and men (1.6%).
Five factors were included in multivariate modelling after adjustment: survey year; recruitment location; length of injecting career; recruitment setting; and having had a voluntary confidential HIV test.
The odds of being HIV-positive were higher for the survey years 1990-95 and 2001-2003 compared with 1996 (p=0.001); higher for recruitment in London compared with outside London (Adjusted Odds Ratio 7.33; 95% CI, 5.60-9.59); highest for those injecting for 15 years or more (AOR 2.3; 95% CI, 1.61-3.28); higher for those recruited in the community versus those from the agency survey (AOR 1.76; 95% CI, 1.37-2.24); and higher for those who had ever had a voluntary HIV test outside of the survey (AOR 2.49; 95% CI,1.95-3.18).
Younger IDUs in London at highest risk of new HIV infection
The investigators used an adjusted model (adjusted for number of years injecting, recruitment setting and having had a voluntary HIV test outside of the survey) to fit location and survey year together, and the results sugggested that the recent increase in HIV prevalence was mainly occuring in London (p=0.025).
To examine this futher, force of infection in and outside of London, defined as the yearly rate at which HIV-negative IDUs become HIV-positive, was estimated by fitting a model to prevalence data by calendar year and injecting career length. The results suggest that force of infection in London is higher amongst novice IDUs (those injecting for less than one year) and has increased over time.
Between 1992-1997, the force of infection amongst novice IDUs in London was 0.008 (95% CI, 0.002-0.02), whereas between 1998-2003 it was 0.028 (95% CI, 0.016-0.045), or almost 3% per year. For IDUs who had been injecting for more than a year, the force of infection was 0.13 lower across all time periods. Since age and length of injecting habit were found to be highly correlated (p=0.001), this suggests younger IDUs in London are acquiring HIV more rapidly than older IDUs in London or elswhere.
This increase in new HIV infections is similar to the 3.4% rate found in a recent London-based cohort study.
Awareness of HIV infection
Overall, 54% of the total cohort had ever taken an HIV antibody test outside of the surveys.
Of those testing HIV-positive, 81%( 371/461) reported ever having taken an HIV antibody test. Of those who reported the results of their last HIV test, 75% (193/259) were aware of their infection.
In 2002-2003, however, only 69% (25/36) of those who were HIV-positive and who reported the results of their HIV antibody test were aware of their infection.
Is UK policy to blame?
Although the combined surveys found that reported needle- and syringe-sharing in the previous month remained uniformly high both in London (31%) and outside London (29%) in 2002, the higher force of infection in London may reflect higher HIV prevalence amongst IDUs in London compared with those outside London, as well as an increased prevalence of injecting drugs, crack cocaine in particular.
However, the authors point out that in 1998, the UK's national drug strategy changed its focus from harm-reduction and the reduction of blood-borne viruses to "wider social harms, in particular drug-related crime." They suggest that this "simultaneous shift in the focus of policy and service provisioning for drug users in England and Wales" may have "unintentionally hindered the development and re-invigoration of harm reduction measures in response to evolving patterns of drug use and risk behaviours."
In addition, younger IDUs would not have been exposed to either national or targeted HIV prevention campaigns that took place earlier in the HIV epidemic.
It also appears that many of the recently-infected IDUs are foreign nationals. "Data on country of birth from clinicians' reports of newly diagnosed HIV infections indicate that two-thirds of HIV-infected IDUs diagnosed in the UK in 2003 were born in another country," the authors write. Thus the recent increase in HIV prevalence in London may reflect recent patterns of emigration to London, particularly from south-western and eastern Europe where the prevalence of HIV is higher among IDUs than in other risk groups.
Reference
Hope VD et al. HIV prevalence among injecting drug users in England and Wales 1990 to 2003: evidence for increased transmission in recent years. AIDS 19:1207-14, 2005.