Current or former injecting drug use is not related to earlier switch or discontinuation of HAART compared to non-IDU patients since 1999

Simon Collins, HIV i-Base

A combined analysis from three prospective US cohorts, published in 6 June issue of AIDS Research Therapy reported that injecting drug use was not related to earlier changing, reducing or switching treatment – discussed as a marker for poorer long-term treatment success – after adjusting for other factors.

The three cohorts – AIDS Link to IntraVenous Exposure (ALIVE), Women’s Interagency HIV Study (WIHS) and Multicentre AIDS Cohort Study (MACS) – were used to select approximately 1400 patients with no history of injecting drug use and compare treatment outcome to just under 850 former or current IDUs. These 1,588 patients contributed 2,358 patient-years with 713 events.

The IDU group had a lower nadir CD4 count and higher proportion of patients who were unemployed, on low income, had lower educational level and a higher proportion of Black, non Hispanic patients. Use of treatment and choice of drugs was similar between the two groups.

All three cohorts collect similar follow-up data, and reported similar trends in ARV prescribing (generally with a similar shift from PI- to NNRTI-based therapy over the time of the study (April 1996-April 2004).

The median time to a first report of discontinuation was 1.1 years vs 2.5 years for people without vs with a history of IDU, and overall the relative hazard (RH) of HAART discontinuation was higher for any IDU use when looking at the whole time period (pre- and post-1999) ([HR1.24 (1.03-148)], However, when looking at the pos-1999 period alone (852 people contributing 382 events over 1,396 person years) this association disappeared in the multivariate analysis [HR = 1.05 (0.81-1.36), after adjusting for previous health, race, income and employment. For patients switching treatment, HR was 0.96 (0.82-1.14) and 1.09 (0.89-1.34) in the pre- and post-1999 periods respectively.

Over time, the proportion of patients using the same HAART regimen increased in both group: form 55% in 1997 to 70% by 2004 (in the non-IDU group) vs increasing from 35% to 65% at the same time points in the IDU group. Similar results were seen when looking at current vs former IDU: in the post-1999 analysis: HR = 1.32 (0.90-1.94) vs RH = 1.00 (0.77-1.31).


These results are particularly useful to challenge the common assumption that drug users are not able to be adherent.


Morris JD, Golub ET, Shruti H et al. Injection drug use and patterns of highly active antiretroviral therapy use: an analysis of ALIVE, WIHS, and MACS cohorts. AIDS Research and Therapy 2007, 4:12 doi:10.1186/1742-6405-4-12.

Links to other websites are current at date of posting but not maintained.