Becker Niels G, Britton Tom, O'Neill Philip D
National Centre for Epidemiology and Population Health, Australian National University, Canberra ACT 0200, Australia.
Biometrics. 2003 Sep;59(3):467-75. doi: 10.1111/1541-0420.00056.
This article is concerned with a method for making inferences about various measures of vaccine efficacy. These measures describe reductions in susceptibility and in the potential to transmit infection. The method uses data on household outbreaks; it is based on a model that allows for transmission of infection both from within a household and from the outside. The use of household data is motivated by the hope that these are informative about vaccine-induced reduction of the potential to transmit infection, as household outbreaks contain some information about the possible source of infection. For illustration, the method is applied to observed data on household outbreaks of smallpox. These data are of the form needed and the number of households is of a size that can be managed in a vaccine trial. It is found that vaccine effects, such as the mean reduction in susceptibility and the mean reduction in the potential to infect others, per infectious contact, can be estimated with precision. However, a more specific parameter reflecting the reduction in infectivity for individuals partially responding to vaccination is not estimated well in the application. An evaluation of the method using artificial data shows that this parameter can be estimated with greater precision when we have outbreak data on a large number of small households.
本文关注的是一种对疫苗效力的各种衡量指标进行推断的方法。这些指标描述了易感性的降低以及感染传播可能性的降低。该方法使用家庭疫情数据；它基于一个模型，该模型考虑了家庭内部和外部的感染传播。使用家庭数据的动机在于，人们希望这些数据能为疫苗诱导的感染传播可能性降低提供信息，因为家庭疫情包含了一些关于可能感染源的信息。为了说明这一点，该方法应用于天花家庭疫情的观测数据。这些数据是所需的形式，且家庭数量规模在疫苗试验中是可管理的。结果发现，疫苗效果，如每次感染接触时易感性的平均降低以及感染他人可能性的平均降低，能够被精确估计。然而，在应用中，一个反映部分接种疫苗个体传染性降低的更具体参数估计得并不好。使用人工数据对该方法进行的评估表明，当我们有大量小家庭的疫情数据时，这个参数能够被更精确地估计。
