|Home | About | Journals | Submit | Contact Us | Français|
The comments of Wilson and Hosein1 regarding my article on the elimination of human immunodeficiency virus (HIV) transmission in the United States2 are most welcome and allow for a further dialogue on this important topic.
Wilson and Hosein make two central points. First, they discuss the technical complexities (or lack thereof) in the model employed in my article. They assert that the Anderson-May reproductive rate model I employed ignores key temporal changes in behavioral, clinical, and social parameters, and that deaths due to non-HIV/acquired immunodeficiency syndrome (AIDS) causes are omitted. As noted in the article, all mathematical models have limitations and, therefore, transparency is critical to allow discussion of the pros and cons of various parameter assumptions and particular equations employed in the model. Indeed, model-based policy and programmatic conclusions are stronger when they rest on the output of a variety of different modeling approaches, so the comments of Wilson and Hosein are most welcome. However, I would note that while the model I employed was straightforward in terms of the equations used, a number of complex factors are captured in the empirically based parameter values employed in the model. For instance, the transmission rate [T(x)] of 5.0 discussed in the article was based on the best known estimates for HIV incidence and prevalence in the U.S. This transmission rate of 5.0 is the result of a multitude of behavioral, clinical, social, and population health factors that unfolded in the U.S. from the beginning of the epidemic to recent times. This is also true for deaths from non-HIV/AIDS causes; the prevalence estimate used in the denominator of the best estimate of T(x) is a function of the HIV and non-HIV/AIDS deaths as actually experienced in the U.S., and is therefore incorporated.
Second, Wilson and Hosein argue that elimination of HIV infection in the U.S. is sufficiently distal that we should not focus on it, but rather should focus on setting short-term goals to ensure the best proximal response to the epidemic. I agree that considering the shorter-term goals and actions is absolutely critical; however, I would assert that strategic planning that includes both proximal and distal aspects is important. For instance, in my article I highlighted two sets of congressional testimony that stated it would be possible to reduce the HIV transmission rate by 50% in (variously) five to 10 years in the U.S.3–5 My congressional testimony described for each of the five years the specific programs, intensity of service delivery, and resource levels needed to reach a 50% reduction in T(x). I believe we must consider the short-term action steps, resource needs, and immediate outcomes necessary to achieve longer-term impacts.
My article was designed to determine if the shorter-term programs described in my (and the Centers for Disease Control and Prevention's) congressional testimony were sufficient to cause the HIV reproductive rate (R0) to fall below unity in the long term (and, thereby, indicate that disease elimination was indeed possible even if in the distant future).
A further illustration of this type of use of my original article can be found by considering President Obama's recently released National AIDS Strategy (NAS).6,7 In the NAS, the President sets a goal of reducing the HIV transmission rate by 30% (from 5.0 to 3.5) by 2015. My article considers a T(x) value of 3.5 as a specific case; at this T(x) value, the threshold value of D (duration of infectiousness) at which R0 shifts from less than to greater than unity is 28.58 (a value just less than our base case modeling estimate of D=28.73). Hence, if the national target of T(x)=3.5 is achieved via the shorter-term actions highlighted in the NAS and its accompanying implementation plan, that is sufficient to lead to a long-term R0 less than unity, but only if D≤28.58. The implication is that meeting these conditions would set the stage for an eventual elimination of HIV in the U.S. (though that would be in the somewhat distant future). I believe it is useful to be able to make some estimate, even if a crude estimate, as to how much long-term impact the short-term NAS goals may have in the future.
Again, I welcome the comments of Wilson and Hosein, and appreciate the opportunity to continue this important dialogue.