There are great hopes but more than a little uncertainty around whether the pharmaceutical industry will ever have the technical means or the political latitude to mine troves of data in electronic health records.

Some people point to pilot projects in this arena and exult, “See, it already works!” Others believe that the same pilot efforts are indicative of the Herculean labor awaiting those who seek to link the pharmaceutical research world and that of ordinary clinical care. Their taunt: “Okay, but let’s see if they can connect anything else to their pilot project.”

The recent DIA Clinical Data Management conference had an interesting twist on this debate. It was not about using live patient data at all. Rather it demonstrated what (we can only conjecture) is C-level interest in the topic at GlaxoSmithKline. We heard two separate speakers from Glaxo at the meeting. Both were excellent.

A Priority At GSK

In the first talk, Michele Thomas, manager of the company’s business planning and development, within the drug development science operations center, said her company is studying electronic health records closely. “GlaxoSmithKline has initiated an electronic health care information initiative,” Thomas said. “What we’re looking to do is have a clear understanding of how health care records are developing and participate in helping to shape that change.”

As the company learns, she said, that knowledge is ploughed back into the scientific process. But as a result of the project, senior managers at the company are starting to rally behind data standards (not just CDISC but HL-7) to allow different clinical and hospital systems to communicate, after a fashion.

Another presentation reviewed some of the commercial databases that GlaxoSmithKline uses. Like his colleague, GSK’s Doug Clark, manager of worldwide epidemiology, gave a brisk, illuminating presentation.

Not Just For Trials

He listed about a hundred (okay, we exaggerate) databases to which GSK subscribes. Some of its data providers include: ICHIS, LifeLink, IMS, GE, IMS, Premier, GPRD, Pharmetrics and Medstat. “Why do we have all this data? Do we just get data for data’s sake?” Clark asked. “Actually, we don’t. There is not much duplication. Each database has its own strengths and weaknesses. We use databases to complement clinical trial data, not to replace it.”

Some of the databases, however, are used in the planning of clinical trials. “We can test inclusion and exclusion criteria for clinical trials to see where the bottlenecks might be, and predict recruitment rates to some extent,” Clark noted.

Digital Ore

The company is always evaluating new data sources and tools with which to analyze that data, Clark said. In all, collectively, the company can sift through 120 million patients in North America and Europe. That’s not one big pile, as best we could tell. The company picks and chooses which database to use based on the needs of the project at hand.

It’s clear that his group assists other teams throughout the company, and by that we mean not only scientifically-minded folks but commercial types who are trying to assess the optimal markets that GSK should be entering.

To hear Clark tell it, some of his group’s work is not only in the public domain but actively pushed in that direction via scientific publication. “We are strongly encouraged to publish our results. We do about 200 studies a year internally and we have over 300 publications [to our credit] over the past few years,” he said.