This abstract was presented as a poster at the 11th Annual Canadian Conference on HIV/AIDS Research in Winnipeg in April 2002, and will be presented as a poster at the XIV International AIDS Conference to be held in Barcelona, Spain, in July 2002.

The impact of STARHS "detuned assay" results on HIV incidence calculations in an ongoing cohort of men who have sex with men (MSM) in Vancouver

Steve Martindale, Darrel Cook*, Amy E. Weber, Mary Lou Miller, Keith Chan, Kevin JP Craib, Robert S. Hogg

The Vanguard Project, BC Centre for Excellence in HIV/AIDS, Vancouver, BC;
and (*) the BC Centre for Disease Control, Vancouver, BC.

BACKGROUND:

Serological Testing Algorithm for Recent HIV Seroconversion (STARHS) ­ a.k.a. the "detuned assay" ­ can be used to calculate the probability that seroconversion has occurred within a defined period of time. Potential applications are to narrow down the window period of infection for newly positive patients to assist with contact tracing; and to estimate HIV incidence density in a population when only cross-sectional results are available.

OBJECTIVES:

To compare STARHS estimates of recent HIV seroconversions with the established window periods for seroconversions in an ongoing cohort of young MSM; and to determine the impact on HIV prevalence and incidence calculations if STARHS results are used to designate baseline reactives as recent seroconverters.

METHOD:

Using STAHRS we tested 61 stored blood samples from 20 of 33 seroconverters and 20 of 28 baseline reactives. Results were compared with the dates of existing HIV results and incorporated into Vanguard Project prevalence and incidence calculations.

RESULTS:

Of 20 seroconverters, STARHS results indicated that 17 had seroconverted within 170 days. Of 31 available results, 26 (83.9%) were congruent with existing data, two (6.5%) were incongruent, and three (9.7%) couldn't be validated, although they were consistent with existing data. Of 20 baseline reactives, STARHS results indicated that nine had seroconverted within 170 days. Of 30 available results, only one (3.3%) could be validated, as prior HIV-negative results were unavailable for most baseline reactives. Designating these nine participants as seroconverters reduced the baseline HIV prevalence from 2.48% to 1.70% and marginally increased the overall HIV incidence from 1.50% to 1.52% when the observation period was increased by six months per participant.

CONCLUSIONS:

The STARHS detuned assay appears to be a useful method of estimating how recently an infected individual has seroconverted, however, incorporating STARHS data had no significant impact on the overall HIV incidence rate in our cohort.