The Cleveland Clinic recently announced plans for a Neurological Institute and how it would attack several diseases. Guess what will be among the first technologies the new Institute expects to put in place? Labs equipped with high-tech electronics? Advanced MRls? Nope. It is data management software. That's not as odd as it sounds.

“The conundrum of medicine is that it's both art and science,” says Robert Cecil, one of the key contributors to the Knowledge Project founded by Chair of the Neurological Institute Michael Modic and its CIO C. Martin Harris. “For instance, a treatment might work well for some patients, but not for others. A well-stocked database of medical knowledge might let us find 100 patients that benefitted from a particular treatment, 100 that did not, and then let us find reasons for the difference.” This may provide solid evidence to improve the care of Clinic's patients.

To that end, Cecil and his team are writing the software that will transform a garden-variety SQL relational database into a medical research tool. He has about five programmers working on recommendations coming from several staff level doctors including Irene Katzan and Erica Schneider. The database captures a range of information such as X-rays, text, and physical data so others can check the integrity of conclusions.

Electronic medical records (EMRs) are a critical first step in capturing patient specific medical data. The analysis of this electronic data is the job envisioned by the Clinic's Knowledge Project. EMRs revolve around the patient while the Knowledge Project database will focus on diseases.

There is plenty of medical information in the world, but it's not as useful as it could be because it's scattered all over the place. “In fact, we don't know what we know,” said Eli Lilly CEO Sidney Taurel at this year's Innovation Summit sponsored by the Clinic. But suppose that data were in an electronic format and searchable by device makers and pharmaceutical researchers. The result could be safer drugs and those developed for uncommon diseases. Taurel suggests Federal restrictions are making data mining less useful than it could be. There's little doubt, he said, that life-saving information is already gathering in EMRs.

Smaller research facilities might find it difficult to do what the Clinic is undertaking. But what if there was a national database of human conditions, one that let you donate data about yourself for the common good? A national medical database begun today would pay off for generations because health data does not depreciate. It becomes more useful as more of it accrues.

Of course contributions to such a database would be voluntary and anonymous. Taurel says three out of four patients are already willing to share information for research. That should be no surprise. After all, many of us will share organs after we die, so what's a little medical data?

Such a database will only be practical if rule writers at the Health and Human Services Administration get out of the way and let medical researchers decide what a database should hold. A spec could turn into a working database within a few months. Funding would come from research institutions and access to it could be licensed.

What's ironic is that databases are the responsibility of IT departments and in most companies that refers to people who keep networks working and Websites up-to-date, the geeks. They're often ridiculed and frequently the butt of jokes. But these code pounders could be heroes if they make it possible to find information needles in the enormous haystack of a national medical database. All of a sudden, they're not so geeky.