The results in this report rely on parameter values that are estimated from data on past pandemic influenza or using data from studies of currently circulating strains of influenza. Characteristics, such as the reproduction number, the mean duration of the incubation period and the effectiveness of AVs, may be quite different for a newly emerged strain of influenza. Some of our results are sensitive to these parameter values. For example the total effort required to eliminate the infection would be much greater, and more difficult to achieve, if R0 is much larger than the values used in our calculations. It is therefore necessary to estimate these parameters afresh using data from the early stages of a local epidemic. Estimates of the effectiveness of AVs for treatment and prophylaxis are particularly crucial because of the limited supply of these drugs.
Collecting data during the early stages of a local epidemic requires careful planning because the response teams will be extremely busy trying to contain the outbreak. Data collection must be made simple, kept to a minimum and seen to be vital. The aims are to estimate parameters needed for strategic planning, to assess the effectiveness of interventions against the new virus and to determine if the initial liberal use of AVs for treatment and targeted prophylaxis is sustainable.
A full analysis of data requirements is yet to be undertaken. However, with the goal of
i. estimating central transmission characteristics of the newly emerged pandemic influenza,
ii. estimating effectiveness measures for antiviral drugs,
iii. assessing the effectiveness of some interventions, and
iv. determining how sustainable the initial use of AVs is,
collecting data on the first 500 affected households seems an excellent target. Households seem the ideal unit for data collection, as such data are required to estimate the effectiveness of household quarantining and the effectiveness of AVs to reduce infectivity. The number of courses of AVs used for treatment, targeted prophylaxis of plausible contacts and prophylaxis of health care and other essential workers should also be monitored over this time. Models can then be used to project the course of the epidemic and the corresponding use of AVs, and provide an objective basis for decisions about the management of the AV stockpile.