We have heard from a biostatistician, an epidemiologist, and a mathematical modeller. This is the combination that is necessary to advance our enterprise of analyzing the reality that transmits infections through populations. It was great to have them on the same podium. It is great to have them meeting in discussions over tea. It would be even greater to have them working together on joint projects.

Julian Peto has presented us with an example of a chronic infectious disease where the need to combine the principles of infectious and chronic disease analysis is clear. HPV is a chronic infection where the course and outcome of the infection are dependent upon numerous host and agent specific factors. Even before the agent of cervical cancer was determined, the sexual mode of transmission was made clear as the number of partners of the victim and of the victim's spouse were both related to the risk of disease. Now that the agent has been determined, we can proceed to work out the determinants of transmission of that agent in ways that should help better to orient prevention programmes. We can begin to assess what aspects of sexual contact patterns and of host immune responses affect the level of disease in a population. We can then begin to plan both behavioural and biological interventions to prevent disease.

To work out transmission dynamics, we also need to understand the natural history of infection and how infection gets translated into disease and into contagiousness. As Julian Peto suggests, the natural history of HPV is complex and heterogenous.