Speed and strength of an epidemic intervention

Jonathan Dushoff; Sang Woo Park

We compare two different possible intervention strategies to shed light on the speed and strength decompositions. First, we consider a condom intervention that reduces HIV transmission by approximately 75% at the population level. Assuming that condoms act as a physical barrier, and that condom use will, on average, remain roughly constant through time, it is reasonable to model the proportional reduction in transmission due to condom use as constant across the course of infection: L condom = 1/(1 − 0.75) = 4 (Fig. 3A) . The estimated strength of such an intervention is simply the average of L condom , i.e., θ = 4, whereas the estimated strength of the epidemic decreases as the proportion of early transmission increases (Fig. 3B) . Thus, the predicted effectiveness of the condom intervention will depend strongly on our estimate of the importance of early transmission: if early transmission is low, we expect disease spread to be too strong to be controlled completely by our intervention.