Professor of Sociology and Statistics
University of Washington
The parallel epidemics of HIV, STDs and blood-borne infections (BBIs) associated with drug use and sexual activity continue to raise major scientific and public policy issues. Recent social science and epidemiologic research on HIV transmission among drug users and others has begun to shift attention from individual behavior to the pattern of connections among individuals. The fundamental network hypothesis states that risk for disease propagation is not fully accounted for by personal behavior, but is independently influenced by social network structure. Developmental work in this field suggests that public policy for the control of STDs and BBIs can be guided by mathematical models that simulate transmission networks, and that such models are most likely to be useful when they are grounded in empirical data. We propose to assemble a number of data sets previously funded by NIDA and other agencies (see section D. 1) that focus on infectious diseases that spread via well-defined contact networks, such as HIV, HBV, and HCV in networks of injecting drug users or STDs and HIV in sexual networks. These will provide the empirical base for integrating networks into traditional epidemiologic analysis (aim 1.0), developing statistical methods for modeling and simulating networks (aim 2.0), and analyzing the network effects on disease transmission dynamics (aim 3.0). Specifically, we propose to: (1) Perform epidemiological analysis of existing network data sets to characterize the association of social network structure with the spread of sexually transmitted diseases (STDs) and blood borne diseases (BBIs). We will incorporate network measures in traditional epidemiologic analyses of these data; analyze the relationship of network measures to basic epidemiologic measures such as the basic reproduction ratio; and develop a public use repository for the data. (2) Develop statistical methods for estimating population level network parameters from network survey data, and simulating dynamically evolving networks with similar properties. Both the estimation and simulation methods will be based on a common Markov Chain Monte Carlo (MCMC) algorithm, which will enable researchers for the first time to simulate networks that have the same statistical properties as those observed in real data. (3) Use these methods to examine the impact of networks on transmission of STDs and BBIs. In particular, we will examine the independent and joint effects of needle sharing and sexual transmission via their respective networks. The results of this project will help to provide a systematic empirical basis for identifying how networks determine the risk of exposure at the individual level, and how they influence the population dynamics of disease transmission. This will support prevention efforts at both levels. It will provide the scientific basis for individual-level prevention strategies that focus on partnership interventions. It will enable public health professionals to identify population-level prevention strategies that make a network less vulnerable to spread. And it will identify the type of network data needed to inform such efforts.