Background:  Most U.S. studies of HIV drug resistance in drug naïve persons have used special blood drawn from consenting persons receiving care in academic or specialist HIV centers after HIV diagnosis. Studies conducted during 2002 through 2004 have reported prevalence of resistance to 1 or more drug classes of > 20%; prevalence of resistance to non-nucleoside reverse transcriptase inhibitors (NNRTI) in recent studies is higher than prevalence of resistance to nucleoside reverse transcriptase inhibitors (NRTI), whereas the reverse was true in studies from the late 1990s. CDC is collaborating with public health departments to support drug resistance surveillance using HIV diagnostic sera, beginning in publicly funded counseling and testing sites and in publicly supported clinical sites.

Methods:  In 2003–2004, drug resistance surveillance began in 65 sites in 5 states. A total of 595 residual HIV diagnostic sera from drug-naïve persons newly diagnosed with HIV were processed in the public health laboratories where routine HIV testing takes place. Sequencing was performed at Stanford University, Maryland State, and University of Washington laboratories. Results were available to providers within 30 days.

Results:  Of the total, 56 (9.4%) specimens could not be amplified for genotyping. Results for routinely handled serum specimens from 539 drug-naïve persons newly diagnosed with HIV were available for 2003–2004. The prevalence of mutations associated with HIV drug resistance was 15.2% (7.1% to NRTI, 9.1% to NNRTI, 3.2% to protease inhibitors, 3.2% to ≥ 2 drug classes). Drug resistance prevalence in participating states ranged from 11.4% to 19.4% with a median of 14.3%.

Conclusions:  In participating surveillance sites, as in other recent U.S. studies, resistance to NNRTI was more prevalent than resistance to NRTI, although overall HIV drug resistance prevalence was lower than prevalence reported in other studies. Persons diagnosed in publicly funded sites may have a lower prevalence of drug resistance than those in specialists’ care, possibly because persons who transmit HIV to persons diagnosed in specialist centers may, as a group, have more access to HIV treatment and therefore to antiretroviral drugs. Representative drug resistance surveillance methods and sufficiently large numbers are needed to obtain accurate national estimates. The use of residual HIV diagnostic sera from counseling and testing sites as well as clinical sites can provide a representative sample of persons newly diagnosed with HIV for surveillance purposes.

Keywords: HIV Drug Resistance; Surveillance; Epidemiology