Circulatory shock.

n engl j med 370;6 nejm.org february 6, 2014 582 are highlighted. First, we concentrate on serodiscordant couples in long-term partnerships (the target population of the HIV Prevention Trials Network 052 study); Hontelez and de Vlas simulate a generalized epidemic. Second, we focus on patients who are already identified and in care (treatment as prevention). In contrast, Hontelez and de Vlas describe an analysis of a test-andtreat strategy in which a patient’s choice not to undergo HIV testing, an inadequate linkage to care, and a lack of initiation of ART would all attenuate the intervention effect. We respectfully disagree with the suggestion that by restricting attention to firstand secondorder transmissions we failed to account for a substantial number of downstream infections. As noted in Table 2 of our article and in Figure S2 in the Supplementary Appendix (available with the full text of our article at NEJM.org), even second-order infections had no material effect on our findings over a period of 5 years. Projections beyond second-order infections in serodiscordant couples require assumptions about the future (e.g., ART efficacy, loss to follow-up, and sexual behavior) that are well beyond the trial data and unnecessary to reach our critical policy conclusions.