Epidemic Models: Their Structure and Relation to Data

Surveys the state of epidemic modelling, resulting from the NATO Advanced Workshop at the Newton Institute in
1993.

Norma Kanarek

A NATO Advanced Research Workshop Conference held in 1993 yielded this collection of papers about epidemic modeling updated by a renewed interest in methods to predict, manage, and understand current public health problems of infectious disease. The editor has amassed presentations that explicate the background theory of infectious diseases and their control; the notions of time, space, and nonlinearity in an epidemic; heterogeneity in human populations; and the fit between data and the models to be tested. The text is a compendium of mathematical epidemiologic process issues that initiated a six-month Newton Institute study section. Although highly technical and quite comprehensive, it is a readable volume and irresistible to the epidemiologist in pursuit of understanding disease processes. A conceptual framework for the study of infectious disease determinants is offered, examined, and modified. Papers deal with the particular issues of irregularities of place, time, host characteristics, population size, and data needs for estimation and prediction. Each chapter includes examples, a summary of the topic, and applications to the practice of public health. The Appendix captures edited symposium comments about issues remaining. It is a textbook only for those sufficiently oriented to the statistical and epidemiologic issues. Terminology is neither indexed, defined in the order used (e.g., the definition of SIR is first found on page 53,), nor used consistently. Those with curiosity about both the basic elements of an epidemic process or a more complex view ( for example, deterministic versus stochastic processes) will be rewarded. Those with interests outside the infectious diseaserealm will find applications to other areas (e.g., diffusion of innovation). Use of these models to reason deductively or inductively is discussed and highlights the need for an interface between those who dwell on these models (theoreticians) and those who dwell in them (practitioners).

Mollison, Denis (Heriot-Watt Univ)

The contributors represent the specialties of statistics, mathematics, epidemiology, zoology, biostatistics, ecology, and evolutionary biology. Most are from universities and research institutions in the U.K., the Netherlands, Australia, Scotland, and Switzerland. Institutions prominently represented include Cambridge Univ, Imperial Coll, London, Heriot-Watt Univ in Scotland, Emory, and Univ of Georgia.