BioBehavioral Models for HIV Prevention
Abstract
<italic>Background</italic>: In this dissertation, I model HIV transmission in two different populations: circular migrants in South Africa, and men who have sex with men (MSM) in Southern California. <italic>Objectives</italic>: One objective of the circular migrations (Chapter 2) project is to compare compartmental and network models in how they explain the dynamics of a system in which some individuals engage in circular migrations. Another goal is to understand the role such movement patterns play in HIV transmission, given the critical interaction between acute infection and rate of movement. In the population of MSM in Southern California, I model postdiagnosis behavior change (PDBC). The objective in Chapter 3 is to quantify the populationlevel effects of individuallevel PDBC. A second objective (Chapter 4) is to compare various strategies for early diagnosis in terms of how many new infections they produce over a 10year period. <italic>Methods</italic>: The compartmental models in Chapter \ref{chap:migrations} are developed using ordinary differential equations (ODE's) and the network models are derived from the exponential random graph models (ERGM's). Epidemiological studies on circular migrations in the KwaZuluNatal province conducted over the past decade provide the necessary motivation. The models on PDBC in Chapters 3 and 4 are based on ERGMs. The behavioral components of these models are parameterized primarily using the Acute Infection and Early Disease Research Program (AIEDRP). Several other studies are used to model the various biological and demographic components of these models (details in Appendix F). <italic>Results</italic>: In Chapter 2, I find that compartmental models do not show any impact of the frequency of migrations on HIV prevalence. Under suitable assumptions, network models show that ``frequent'' migrations (that take place at intervals shorter than the window of acute infection) produce a larger epidemic than infrequent migrations. In Chapter 3, I find that if there were no PDBC, HIV prevalence in Southern California MSM would be higher by about a third. In Chapter 4, I find that individual testing strategies where people test based on the number of partners or time since last test work better than other strategies I consider. <italic>Discussion</italic>: Chapter 2 is a largely theoretical study, parameterized by qualitative descriptions of data. While the question of the precise impact of circular migrations on HIV transmission is still an open one, and likely contingent upon specific behavioral components in particular populations where circular migrations are practiced. However, our %%my? work demonstrates that we need to network models to understand this connection. We also demonstrate the potential impact of acute infection and its interaction with frequency of migration, and the effect of delayed pathacceleration. In Chapters 3 and 4, I focus on behavioral interventions to prevent HIV infections. I do not consider ``treatment as prevention'' type strategies; though such considerations are suitable for future work and may act in synergy with the behavioral interventions I consider here.
Collections
Related items
Showing items related by title, author, creator and subject.

Uncertainty Quantification Problems in Tsunami Modeling and Reduced Order Models for Hyperbolic Partial Differential Equations
Rim, DonsubIn this thesis, we consider an uncertainty quantification (UQ) problem that arises from tsunami modeling, namely the probabilistic tsunami hazard assessment (PTHA) problem. The goal of PTHA is to compute the probability ... 
A flexible class of models for regression modelling of multivariate failure time data
Lynch, James Charles (1996)The ClaytonOakes model (Clayton 1978, Oakes 1989) has become a popular semiparametric model for bivariate failure time data. The model is attractive because the marginal distributions are left unspecified and can be ... 
Statistical Hurdle Models for Single Cell Gene Expression: Differential Expression and Graphical Modeling
McDavid, AndrewThis dissertation describes a set of statistical methods developed for analysis of single cell gene expression. A characteristic of single cell expression is bimodal expression, in which two clusters of expression are ...