To the Editors: Burman et al1 present an interesting article concerning the transmission risk behavior related to episodic antiretroviral therapy when compared with continuous therapy. However, we would like to raise some points of concern. The authors emphasize in the title the HIV transmission risk (Episodic antiretroviral therapy increases HIV transmission risk compared with continuous therapy: results of a randomized controlled trial) when the main outcome is that the transmission risk behavior shows no difference between the 2 groups. Although the high viral load in this article is associated with episodic antiretroviral therapy, any other situation of increased viral load such as recent HIV infection and treatment failure are accompanied by an increased transmission risk (the possibility has even been mentioned to participants during the informed consent process).2-5 The fact that the levels of viral load could predict the risk of transmission6,7 has been demonstrated in perinatal and in heterosexual transmission as well.8 In our opinion, only a balanced set of prevention actions (including information and education on HIV transmission and prevention, condom use promotion, early HIV infection diagnosis, effective and timely adequate antiretroviral therapy, and male circumcision, in special cases, among others) will be able to comprehensively face the complex HIV and AIDS scenario, and we would prefer to have the association between the increased viral load observed in episodic antiretroviral therapy9 and the greater transmission risk added as an issue of discussion in a more general prevention strategy. Second, the authors found a decrease in transmission risk behavior from 20% at baseline to 10% at 4 months, among participants without antiretroviral therapy at baseline. That finding could have been more explored. Why did the transmission risk behavior decrease? Were the patients more sensitized because they were beginning antiretroviral therapy? Was the topic of adherence approached in a broader context of prevention? Perhaps this was one of the most important study results left underexplored. Finally, at the baseline data on sexual and needle-sharing behavior, only 20 subjects admitted injecting drug use (with 2 assuming needle sharing) and the authors recognize that injecting drug use was infrequently reported (approximately 2%). Even though there was no difference between randomized arms, it is difficult to accept that we have sufficient data to conclude that the episodic use of antiretroviral therapy did not affect needle-sharing behavior in the same way it affected sexual behavior. Carina Guedes Ramos, MD Nemora Tregnago Barcelos, PhD Universidade Federal do Rio Grande do Sul Porto Alegre Brazil