Grants

Improving Public Health and Medicine with Reality Mining
Funded by the Robert Wood Johnson Foundation, 9/2008-3/2009
Alex Pentland, Ph.D., Principal Investigator, Massachusetts Institute of Technology
Devon D. Brewer, Ph.D., Co-Investigator

The purpose of this project is to produce a whitepaper on potential applications of reality mining--the process of inferring human activity patterns from the digital traces left by individuals' communication and other electronic devices--in clinical medicine and public health.  A conference of professionals engaged in relevant work will also be convened to discuss topics related to this broad venture and the results of the meeting will be summarized in a report. 

Daily Reports of Injection Drug Use and Risk
Funded by the National Institute on Drug Abuse, 4/2006-3/2008
Barbara Leigh, Ph.D., Principal Investigator, University of Washington
Devon D. Brewer, Ph.D., Co-Principal Investigator

Injection drug users (IDUs) face high morbidity and mortality as a consequence of injection drug use. In particular, IDUs are at high risk for blood-borne infections, including HIV, hepatitis B (HBV), and hepatitis C (HCV), due to the sharing of drug injection equipment and other blood exposures. As in other areas of behavioral epidemiology, measurement issues in this domain are crucial but are difficult to address adequately. In the absence of noninvasive and ecologically valid biological or observational techniques, comprehensive measurement of drug use and risk behavior depends largely on self-reports, which are generally retrospective. Although researchers have attempted to increase the accuracy of these self-reports, a “gold standard” of measurement for these behaviors does not exist.

One approach to more valid measurement of behavior is the use of daily reports, which are assumed to be more accurate than retrospective reports covering longer time periods. Daily measurements are used in several disciplines to collect information about people’s experiences, behaviors, and life events. In research on alcohol and tobacco use, daily measurements have been used to validate retrospective measures of consumption. The purpose of this project is to conduct a pilot study in which IDUs report daily on their drug use and injection risk behavior, using an innovative technique of computerized telephone interviewing (Interactive Voice Response, or IVR). These daily reports will then be compared with retrospective measures of risk similar to those used in most epidemiologic studies of IDUs.

The goals of this exploratory project are to evaluate IVR as a tool for collecting daily reports of injection drug use and risk and to examine the quality of retrospective measurements compared to daily reports. Our aims in this project are both methodological and substantive:

Although daily reports do not yield perfect measures of behavior, they are not subject to some of the errors inherent in retrospective reports. By comparing the two kinds of measures, we can examine some of the ways in which retrospective reports may or may not capture injection drug use and risk behaviors. If we demonstrate that the use of IVR is feasible with this population, future studies can make use of a cost-effective means of measuring injection drug use and risk in IDUs.

Clients of Prostitute Women: Deterrence, Prevalence, Characteristics, and Violence
Funded by the National Institute of Justice, 9/2003-5/2006
Devon D. Brewer, Ph.D., Principal Investigator

Prostitute women are victims of physical and sexual violence at exceedingly high rates. Recent research indicates that prostitute women have the highest murder rate of any population of women ever studied. The available evidence suggests that male clients of prostitute women make up the bulk of the perpetrators. Prostitute murders are among the most difficult to solve, and in recent decades many local jurisdictions have mounted expensive and lengthy investigations of such homicides, often without resolution. In addition, vice operations against clients are more resource intensive than those against prostitutes, and these activities represent nontrivial resource commitments for many local law enforcement agencies across the country. In this project, we will address large gaps in knowledge about clients to enhance efforts to curb prostitution and develop profiles of violent clients.

The goals of this project are to:

1) Assess the specific deterrent effect of arrest for patronizing a prostitute;

2) Estimate the prevalence of clients overall and the subset of clients who are violent toward prostitutes;

3) Compare clients with the general population of men in terms of demographics and geography; and

4) Compare clients who are violent toward prostitute women with clients overall in terms of demographics, geography, and criminal history.

To achieve these goals, we will conduct secondary analysis of criminal history records, homicide files, field observations, survey data, census figures, and related information from many states and metropolitan areas in the US.  We have already obtained much of these data and during this project we will acquire the rest of the data needed.

To assess the specific deterrent effect of arrest, we will analyze a unique data set that permits calculation of the arrest rate for clients first detected through disease control activities with the rate of rearrest for those first identified by arrest.  We will estimate the prevalence of clients and violent clients for different geographic areas across the US with arrest data, field observations of violent clients, and capture-recapture statistical methods. Our comparisons of clients with the general population of men and of violent clients with clients overall will involve univariate analyses of data obtained from all sources.

The etiology and prevention of blood-borne viral infections in injection drug users (IDUs) have not been fully characterized, and many questions remain regarding which injection practices may result in infection.  Viral hepatitis infections in IDUs are among the most frequently occurring blood-borne infections in humans; in low HIV-prevalence settings, morbidity and mortality in IDUs attributable to hepatitis virus infections may exceed that for HIV. Hepatitis C virus (HCV) is highly prevalent in IDU-populations.  Because sexual HCV transmission is relatively rare, HCV infection may serve as a highly sensitive biologic marker of direct percutaneous exposure in drug injectors.  Increased knowledge of HCV transmission in drug injectors may contribute to understanding the mechanisms of transmission of other infections via injection practices.  The long-term goal of our research is to advance knowledge of the epidemiology, etiology and prevention of HIV and hepatitis infections in IDUs. 

We propose studies that will make new contributions toward our long-term goal: Aim 1. Examine the extent to which HCV prevention education at the Seattle needle exchange program has reduced the risk of HCV infection. Aim 2. Measure the risk of HCV seroconversion associated with specific injection risk behaviors, and calculate the risk of HCV attributable to these practices in the IDU-population. Aim 3. Compare reporting of socially-undesirable or stigmatized injection and sexual risk behavior collected by A-CASI to interviewer-administered data collection methods. Aim 4. Assess the feasibility and disease control benefits of HBV and HCV partner notification for IDUs. Aim 5. Study whether changes in hepatitis C reporting laws are associated with increased reporting in IDUs.  The significance of this research is the potential contribution to our understanding of the etiology of these infections, and examination of many practical questions related to the effectiveness of public health surveillance and prevention programs.

Interdisciplinary Scientific Research's primary responsibilities for this project are leading the work for aim 4 and assisting with the scientific aspects of the other project components.

 
Interdisciplinary Scientific Research's primary responsibilities for this project have been providing advice about questionnaire development, interviewing techniques, and research design, performing data analysis, and preparing manuscripts.

Contracts/consulting

Event-level Analysis of Expectancies, Alcohol Use, and Consequences
Funded by the National Institute on Alcohol Abuse and Alcoholism, 5/2010-4/2014
Christine Lee, Ph.D., Principal Investigator, University of Washington

The goals of this project are to: (1) examine the daily relationships between alcohol expectancies and alcohol use; (2) study bidirectional influences of expectancies, consequences, and drinking over time; (3) examine whether the immediate consequences of drinking change subsequent expectancies and drinking over time; and (4) examine the time-varying and time-in varying moderators of the within-person relationships between expectancies, drinking, and consequences. 

Interdisciplinary Scientific Research advises on questionnaire and study design, and programs, deploys, and monitors an interactive voice response interview system for daily data collection.

Development and Reliability Study of a Daily Alcohol Expectancy Questionnaire, 2008-2009
Christine Lee, Ph.D., Principal Investigator, University of Washington

The goal of this project was to develop and pilot test an interactive voice response daily interview focused on drinking expectancies, behavior, and outcomes. 

Interdisciplinary Scientific Research advised on questionnaire design, and programmed, deployed, and monitored the interactive voice response interview system for data collection.

Random Digit Dial Survey of Men Who Have Sex With Men
Funded by the National Institute of Allergy and Infectious Diseases, 5/2006-12/2006
Matthew R. Golden, MD, MPH, Principal Investigator, University of Washington

The goal of this project is to assess various aspects of health and behavior in men who have sex with men and reside in Seattle.  This cross-sectional, probability sample survey will be used to compare with a similar survey conducted in 2003 and assess health risks not measured previously. 

Interdisciplinary Scientific Research's primary responsibilities for this project include questionnaire development, coordinating and monitoring the implementation of the survey, and contributing to manuscripts. 

Organizational Learning from Error
Funded by the Agency for Healthcare Research and Quality, 2006
Michal Tamuz, Ph.D., Principal Investigator, University of Tennessee Health Science Center

Interdisciplinary Scientific Research's primary responsibility for this project was to advise investigators on questionnaire and study design for a project on medication errors and the social networks among hospital personnel. 

Adoption of BlackBerries in Law Enforcement Teams
2004-2006
Edward Balkovich and Susan Straus, co-Principal Investogators, RAND

Interdisciplinary Scientific Research's primary responsibilities for this project were to advise investigators on study and questionnaire design and perform data analysis.