Clinicians whose experience predates the Internet can attest to the revolution instant literature searches represented for clinical practice. Ida Sim, MD, PhD, foresees as dramatic a revolution in the next generation of accessing medical research - in fact, she's designing it.

Sim and colleagues are joining with the Journal of the American Medical Association and Annals of Internal Medicine to introduce a new standard that would fundamentally change the way clinical research is published. This Trial Bank Project offers a new model that will allow clinicians and researchers to not only access data, but to manipulate it.

Sim aims to do for clinical trials data what GenBank has done for DNA sequencing. GenBank, sponsored by the National Institutes of Health, publishes annotated genetic sequence data that can be searched by organism, sequence length, and other biological features. The aim would be to make clinical trials data equally "computable," and enable clinicians to search a robust "trial bank" of actual data on subjects, intervention, and outcomes - not just medical literature about the data.

For example, a request for "trials with over 100 patients that assessed, as their primary outcome, total mortality in post-heart attack patients who were given aspirin" would retrieve both the narrative of a published article along with the complete, structured information of every relevant trial. (To sample a trial bank entry, see http://rctbank.ucsf.edu/Presenter/.)

With this kind of a reporting system, the synthesis of evidence from clinical trials would be far more comprehensive. In addition, a standardized format would produce more complete descriptions of trial characteristics and results, and could reduce bias in interpretation.

"This can help close the gap between proof and practice," says Sim.

Medical informatics is one part of broad efforts at UCSF to apply computer science and technology to health care research and delivery. New computational methods are contributing to advances in genetic sequencing, molecular modeling, and pharmacogenomics, on up to clinical decision-making and the use of electronic medical records. Biomedical informatics is an increasingly critical discipline for translating the findings from the human genome project into clinical care.