Jim has been oriented to advancing public health since he entered the University of Michigan medical school in 1963 after two years of undergraduate work. He benefited from a joint studies program that gave him six years to complete medical school while simultaneously pursuing public health, social, political, and historical issues. He deepened his interest in Latin America during six months of medical training in Chile. After completing a pediatrics residency at Harbor General Hospital in 1972 he spent two years as an EIS officer and then acting state epidemiologist in the state of Washington. Upon completing his EIS stint he worked for WHO in smallpox eradication in India. He saw the last cases of smallpox ever in Azamgarh district of Uttar Pradesh and also saw the devastating effects of the vaccine that was given to contacts of cases in a population that was mostly unimmunized.
He then worked in Colombia, South America establishing new surveillance approaches and training programs in the Cali health department and chasing diarrheal disease and malnutrition. In Cali he developed theories about infection transmission systems and studied the philosophy and methods of mathematically modeling infection transmission. He returned to Michigan as assistant professor of epidemiology in 1978, largely pursuing his practical field approach to epidemiology both in the state of Michigan and South America. After getting tenure, he took leave from the University for two years to work with CDC to establish a national field epidemiology training and investigation service in Mexico that subsequently has had great success.
Upon returning to Michigan in 1986, Jim changed careers from field epidemiologist to modeler and theorist. His colleagues John Jacquez and Carl Simon taught him the basics. Initially he worked on HIV transmission systems and with his colleagues demonstrated the great potential of the early stages of HIV infection to accelerate the population dissemination of infection. HIV was not the ideal agent for establishing a science of infection transmission system analysis so lately he has focused on common throat infections and waterborne infections. His major focus, however, is not on agents but on new methods for a science of infection transmission. He is advancing two cornerstones of that new science. The first is the integration of diverse modeling methods into a single coherent analytic approach called Model Transition Sensitivity Analysis. The second involves using the nucleotide sequences of infectious agents isolated from individuals with documented exposure points in the transmission system to make inferences about transmission systems. He helped to initiate Michigan's Center for the Study of Complex Systems and continues to be an active investigator in that center.
EPID524: AIDS: A Public Health Challenge
EPID606: Advanced Infectious Disease Epidemiology
EPID802: Computer Simulation of Epidemiologic Processes
MD, University of Michigan, 1968
M.P.H., University of Washington, 1975
Research Interests & Projects
Jim's research focus is on promoting a science of infection transmission systems. His work together with that of many other researchers has demonstrated that the backbone of transmission systems for human adapted infectious agents is the pattern of contacts that direct the flow of infection through a population. Traditional epidemiology has focused on identifying risk factors for infection or disease, modes of transmission, and agent characteristics affecting disease. The biggest determinants of population levels of infection, however, are risk factors for transmission from infected individuals and contact patterns. New epidemiological methods are needed to address these. Jim believes that a cornerstone of those methods must be a new approach to population models of transmission. The core new knowledge that must be uncovered to understand and control the spread of infection makes transmission models complex. These are complex natural histories of infection, infectious agent diversity and evolution, numerous risk factors affecting transmission, multiple modes of transmission, transmission potential via different modes in different social and environmental settings, and complex contact patterns in those settings. Important contact pattern dimensions involve: varying duration of contacts, repeat contact rates, social and geographic locality restriction of contacts, selectivity of encounters in different settings, and temporal variations in contact rates.
The sensitivity of transmission models to realistic complexities requires a new approach to modeling. All model uses should be validated by assessing the robustness of model based inferences to simplifying model assumptions. Such assumptions should be relaxed either by adding realistic model detail or by using modeling approaches with different assumptions. Jim has pointed out how to do this using a Model Transmission Sensitivity Analysis (MTSA) approach. A community of transmission system scientists is needed to identify and execute the needed model analyses. But any such community will not be able to function effectively until software exists to facilitate the modeling tasks, facilitate communication of exactly what was done in a model analysis, and facilitate replication of results by others. Jim and his colleagues have laid out a path for MTSA software development to meet these challenges. Geoff Jacquez has acquired NIAID support for development of MTSA software through his company Biomedware Inc. The development and promotion of this software is one of Jim's a major projects.
Software, however, can only facilitate the development of transmission system science. The establishment and advancement of this science requires addressing the transmission systems of specific infections and their models with data that is up to the task. The specific infections that Jim addresses are Non-Capsulated Haemophilus influenzae (NCHi), Cryptosporidia, gonorrhea, Hepatitis A, and influenza. The advancement of transmission system science for these agents will help face smallpox and other bioterrorist agents and Jim is undertaking projects to insure that this linkage is established.
The NCHi project is supported by a grant from NIAID that designs and evaluates vaccine trial designs that can determine how vaccines will affect transmission systems. The NCHi vaccine is still off into the future. That was one reason for choosing NCHi. Another was that the organism is of interest to Jim's colleagues Janet Gilsdorf and Betsy Foxman who do laboratory and field epidemiology investigations of this agent. The project is directed currently mainly by Jim'’s doctoral student Ximin Lin. Another doctoral student, Thomas Riggs, addresses fundamental modeling issues for the project.
The Cryptosporida project is in collaboration with Steve Chick at INSEAD in France and is supported by the EPA. Steve has elaborated new data analysis methods to address the transmission system issues on which Jim focuses. Cryptosporidia is spread rather readily person to person. It is also spread by water, however, and special water treatment currently not in place is necessary to insure that water does not spread this agent. As for all infectious agents with multiple modes of transmission, the role of the different modes of transmission in the transmission system is not precisely known. Water might just be a peripheral part of the system, it might be a core of the transmission system that all alone can keep the agent circulating, it might act jointly with person to person transmission to sustain transmission, or more likely it might play a role in disseminating infection to localities where person to person transmission amplifies the impact of the waterborne transmission. In any case, comparing communities with different water sources or studying families with real and placebo water filters cannot by themselves resolve the issue. Steve's new approaches to analyzing water contamination levels and endemic incidence of infection are a first step toward determining the role of water. A more complete MTSA approach, however, will be required to resolve the issue.
CDC supported the gonorrhea project but funding ran out some time ago and the project is now sustained only by the doctoral student work of Tsung Wen Kuo. The project seeks to model the STD transmission systems and use models to improve surveillance and control. Originally the project was also intended to address HIV. Fundamental to the project was software development pursued in collaboration with Steve Chick. The software approach pursued was called GERMS. It was the difficulties of using this approach that led to the development of the MTSA approach. GERMS represents an advanced type of model that relaxes many of unrealistic assumptions commonly incorporated into STD models. Specifically it models a process of linkage between individuals into complex networks of sexual partnerships that are continually forming and breaking up. One issue addressed by GERMS is the use of information on where gonorrhea partners met their partners. At the Genessee County Health Department such information was collected. Tsung Wen Kuo is analyzing this data and using GERMS models to assess the potential of such information to focus control efforts to where they will do the most to stop transmission. The original objectives of this project, however, cannot be fully realized until MTSA software is developed to the point where it can address complex dynamic network models like GERMS.
The hepatitis A project is supported by a Rackham collaborative grant from the University of Michigan and is being pursued by Kate Jacobsen for her doctoral work. This project will also contribute greatly to the goals of the EPA grant. Kates choice of hepatitis A to address the control of waterborne infections in developing countries was a stroke of genius. The agent is ideal for the development of methods using nucleotide sequences to describe transmission system infection flow patterns and fit transmission system parameters. It has extensive genetic variation on long and short-term scales but amazingly it has a single serotype and infection induces lifelong immunity to all genetic variants. Kate is first developing models of the hepatitis A transmission system that capture both person-to-person and waterborne modes of transmission and reproduce observed patterns of age specific infection rates and antibody levels. She then plans to analyze these models to show the potential for water purification, hygiene improvement, and vaccination to reduce transmission. She then hopes to define the data that can resolve control issues raised by her models. The focus will be on nucleotide sequence data.
The influenza project will hopefully begin in the spring if funding becomes available. It addresses the use of chemoprophylaxis, infection treatment, vaccination, and contact interruption through closure of public encounter venues to stop the transmission of new pandemic influenza strains or bioterrorist spread influenza strains. Louise Herlocher is the laboratory collaborator who will make this project go. Jeff Long and Rod Little will form a statistical genetics core for the project and help develop ways to use the epidemiological and nucleotide sequence data that will be collected. Arnold Monto's vast influenza experience will provide guidance and support for the project. Data and specimens will be collected from volunteer families, randomly selected individuals from the community, and volunteers who agree to connect regularly on the internet to provide exposure and symptom information. Three entire segments of each flu virus isolated or identified by PCR will be sequenced. The MTSA model development and analysis strategy will be used to maximally extract information from this data. If and when funding is attained, a post doc and doctoral students will be needed for this project.
Another grant has been submitted to model smallpox control in the face of bioterrorist smallpox events. Mass vaccination where people voluntarily come in to vaccination sites will most likely be quite ineffective in stopping smallpox spread. Active pursuit of contacts of smallpox cases will be necessary. Even some quarantine actions involving not only family but also other sorts of contacts may be necessary. The issues are how these activities should be directed, how intense and how extensive they have to be to hold transmission to no more than one generation, and what specific pre-event vaccination activities will maximize the effectiveness of post-event control efforts while not representing such a public risk that they will inhibit voluntary vaccination after an event. There are many different intuitions about the answers to these questions. The intuitions of decision makers are inadequately informed so a good part of this project is directed to developing modeling approaches and policy development strategies that will improve the quality of those intuitions and make the decision maker more comfortable in using model based assessments of infection transmission.
Another general modeling project is being pursued in a broad context that is currently not directed to any specific agent but which will inform the work on all of the agents on which Jim’s research is focused. Collaborators in this project are Carl Simon and Jim's doctoral student Chris Riolo. This project explores how patterns of the number of transmissions that separate two infected individuals from their most recent common ancestor in the chain of infections leading to them are related to risk determinants. It also explores how those patterns are related to the effectiveness of different infection control strategies. A very simple model of endemic transmission in a population with distinct high and low risk segments forms the core of the work. High risk in this model can be generated by increased contact rates, increased susceptibility, increased contagiousness, or increased duration of infection. Chris creatively developed new approaches to analyzing contact network patterns related to factors affecting infection prevalence. These it turns out blended nicely with Jim's ideas about using nucleotide sequences. It is anticipated that in about 8 months there will be sufficient progress on this project to justify seeking funding for it to be more fully developed.
Finally the broad general significance and nature of Tom Riggs's work deserves mention. Tom is exploring a simple model of transmission within transmission units given different risks of infection from outside of the unit. He is working out the full stochastic distributions of the model and how they are affected by various factors including the nature of transmission in the transmission unit. The extent to which contact dilution affects effective contact rates as transmission unit size increases turns out to be a crucial factor. This project was pursued because of its implications for the design and analysis of trials that can assess vaccine effects on transmission. But it has important implications for the design and analysis of all infection transmission models. This work might lead to clearer definition of the data needed to resolve infection transmission issues and it might provide a framework for estimating a fundamental parameter that is currently ignored in almost all transmission models. That parameter is the contact dilution parameter relating transmission unit size to effective contact rates.
Search PubMed for publications by James Koopman >>
Koopman, J.S. (2007). Infection transmisson through networks. In Kepes, F. (Ed.) Chapter 13 of Biological Networks. (1-58). World Scientific, Singapore.
Ness, R.B., Koopman, J.S., Roberts, M.S. (July, 2007). Causal system modeling in chronic disease epidemiology: A proposal Annals of Epidemiology, 17(7):, 564-568.
Lin, X., Koopman, J.S., Chick, S.E. (March, 2007). Mathematical model comparisons of potential nontypeable Haemophilus influenzae vaccine effects. Journal of Theoretical Biology, 245(1), 66-76.
Koopman, J.S. (2005). Mass action and system analysis of infection transmission In Ed. K. Cuddington and B.E. Beisner in the Theoretical Ecology Series (Series Ed. A. Hastings) (Ed.) Ecological Pardigms Lost: Routes to Theory Changes Academic Press.
Koopman, J.S., Simon, C.P., Riolo, C.P. (September, 2005). When to control endemic infections by focusing on high-risk groups. Won the Rothman prize as the best paper published in Epidemiology in 2005. Epidemiology, 16(5), 621-627.
Koopman, J.S. (2004). Modeling infection transmission. Annual Review of Public Health, 25, 303-326.
Koopman, J.S., Chick, S.E., Riolo, C.P., Simon, C.P., Jacquez, G. (2002). Stochastic effects of disseminating versus local infection transmission. Mathematical Biosciences, 180, 49-71.
Koopman, J.S., Jacquez, G., Chick, S.E. (December, 2001). New data and tools for integrating discrete and continuous modeling strategies. Annals of the New York Academy of Sciences, 954, 268-294.
Koopman, J.S., Chick, S.E., Riolo, C.S., Adams, A..L, Wilson, M.L., Becker, M.P. (November, 2000). Modeling contact networks and infection transmission in georgraphic and social space using GERMS. Sexually Transmitted Diseases, 27(10), 617-626.
Koopman, J.S., Simon, C.P., Jacquez, J.A. (1994). Assessing effects of vaccines on contagiousness and risk factors for transmission In EH and Brandeau ML (Eds.) (Ed.) Modeling the AIDS Epidemic: Planning, Policy, and Prediction (439-460). Raven Press, New York.