Behavioral Informatics

New computer technologies and advances in modern measurement theory make it possible to develop and maintain area-specific item banks, conduct statistical modeling of responses, and create computerized adaptive testing that allows item subsets to be tailored to the individual without loss of scale precision or content validity. In 2004, RFA-RM-04-011 targeted the dynamic assessment of patient-reported outcomes and established a new network – the Patient-Reported Outcomes Measurement Information System (PROMIS) to collect, manage, and evaluate the data, and to develop a computerized system that administers dynamic questionnaires (i.e., computerized adaptive tests based on item response theory). Not surprisingly, the multi-million dollar award went to an institution that already had in place information infrastructure capable of supporting such a demanding task. The conclusion is that investing now in a comprehensive computing environment will not only put our department on the cutting edge of basic and clinical research, but will also attract government funding and industrial clinical trials. We have made the first step in this direction by creating several flexible and comprehensive database systems with integrated project-specific computational tools. While the complete description of these systems are out of the scope of this overview, the systems allows for:

1. Creating custom study-oriented data collection environment including questionnaires and other sources of information (e.g. EEG files, driving assessments, etc.);
2. Prospective data collection through web interface that allows remote access from anywhere and by anyone (researchers, administrators, subjects, patients);
3. Event tracking through e-mail and telephone-call interaction;
4. Creating custom study-oriented data processing tools and including these tools within the database system.
5. Comprehensive data tracking of research studies for project coordinators to know what needs to be completed on a daily basis as well as the status of all subjects in a given study.

A behavioral informatics core facility will ultimately include critical shared resources, which can be used by all research projects in our department. It will provide computational and statistical support and could establish a shared database server system to house data across projects. The idea is that we will have integrated research databases that will be organized hierarchically to encompass the consecutive levels of biomedical data acquisition from molecular and genetic, through physiological networks, to human behavior. These structures will facilitate discovery of connections between the bio-system levels, such as understanding of genetic mechanisms driving physiology, or interactions between physiology and behavior. In addition, it should also promote collaborations across sub-specialties, and demonstrates our commitment to using new and important technologies to improve health care.