Brief description: This is a graduate level course in the conceptual and analytic tools used to understand how spatial distributions of exposure impact on processes and patterns of disease, introducing students to the special design, measurement, and analysis issues associated with spatial patterns of diseases. We will address contemporary diseases of public health importance and present the statistical and inferential skills that can be used in understanding how spatial patterns arise and what they imply for intervention.

Target Audience: Ph.D. and second-year Masters students in epidemiology, environmental health, ecology and various aspects of community health.

Pre-requisites: Knowledge of epidemiology and biostatistics, including probability theory, is essential. Familiarity with GIS and the analysis of spatial data is helpful but not required. Course size is limited to 20 students.

Course Objectives: Provide students with the knowledge, theory, and methodological skills for analyzing and interpreting the spatial patterns of various diseases in order to elucidate underlying exposure processes giving rise to the observed patterns.

Course Description: The course will begin with a description of spatial components of human health data, and the characteristics and covariates of such data. The objectives of spatial analysis are then presented, with an emphasis on (1) quantification of spatial pattern, and (2) mechanisms for inferring past space-time processes from spatial pattern. The CDC guidelines for investigating health event clusters will be reviewed. Spatial statistical methods for quantifying spatial pattern will then be presented, including spatial autocorrelation statistics (both local and general), disease cluster tests (both focused and general) and methods for disease surveillance through both space and time. The framework for using these techniques will be Exploratory Spatial Data Analysis (ESDA), whose objective is the quantification of spatial pattern in order to generate testable hypotheses. Laboratory exercises will use appropriate software (e.g. ArcView, Stat!, GeoMed, Gamma) to analyze example data sets, which will include concrete examples of infectious, toxic-exposure, and other non-infectious diseases. Students will be expected to complete a research project using their own or supplied data, and to produce a manuscript-style report as well as a web-ready presentation (e.g. powerpoint presentation) which will be given in class.

Course Functions: The course is broken up into discussions/lectures and labs:

1. Discussions are on Tuesday (10-12, SPH-I 3000), Labs are on Thursday (10-12, computer center, basement of SPH II - Computing Classroom A).
2. Readings (which students should read before class) raise issues which are discussed in class and addressed in lab. Readings may come from traditional print media, or the web. In fact, each discussion session makes use of a web-based "module", accessible from our course site -
   http://www.sph.umich.edu/geomed/course/

   The appropriate module should be investigated prior to class, and links to each one are available from the course details page, or the Course Calendar. They will constitute part of the resource material under discussion. Of course, it is not possible to follow every link, but it will generally be clear what aspects should be investigated prior to class.

Readings: Coursepack: There will be an assemblage of readings photocopied from various sources available for purchase at Accu-Copy (518 E. William). Other readings will come from the web, or will be provided ahead of time by the instructors.

Class Participation: This is a new, experimental course that we plan to offer regularly in the future. Therefore, your feedback and input are essential. In order to regularly evaluate our performance and keep in touch with your concerns, we ask you provide us with regular feedback. We seek your candid observations and concerns. Your constructive feedback will not be taken personally; rather, it will be used to improve the course as we progress through the term.

Labs: Problems that involve use of computer and statistical methods to analyze data sets will be done during labs. Dr. Long will be primarily responsible for these sessions, and will be present to help during the regular lab meeting time. We consider the lab exercises to be important to learning how to apply many of the techniques of the course. The problems and software will be available through the web site and on the machines in the SPH computer room.

Final Project: Each student is expected to undertake a research project using their own data or data that we will supply, and to analyze these data using tools taught and discussed in the course. Students are encouraged to discuss this project with the instructors early in the course. The final product will be written in the form of a manuscript. In addition each project will be presented orally to the group on either the 11th, 13th, or 17th of April during our regular meeting times or final exam hours. The presentation will be short (15 minutes, with 5 minutes for questions). The written report is to be submitted at the time of the presentation. The presentation will be made public via the course web site, so it is important that it be in a web-ready form (e.g. powerpoint, html).

Students will submit a short, preliminary description of projects by February 10th, including

• a description of the data set to be used,
• questions to address, and
• expected analysis approach.

Forms of Evaluation: This course will be graded, with the grades determined (on the usual 90/80/70/60 scale) as follows: