EXOME

all the information, none of the junk | biotech • healthcare • life sciences

Microsoft’s Life Sciences Game Plan: Use IT to Usher in the World of Predictive, Personalized Medicine

Nobody ever accused Bruce Montgomery of being a mealy-mouthed, politically-correct businessman. On Microsoft’s own turf, he offered some free advice last night about what to do with an extremely messed-up healthcare IT industry.

“If Microsoft really wants to own the world, create a standardized electronic medical records system and give it away for free the first five years. Then start charging,” said Montgomery, a senior vice president at Gilead Sciences, the world’s largest maker of AIDS medicines, in an event hosted by the Washington Biotechnology & Biomedical Association last night at the Microsoft campus in Redmond.

The problem is that every hospital is using different programs that don’t talk well to each other. Some of these, including one called SAS that’s been favored by the FDA (and universally loathed by pharma companies, from what I gather), deserve to be put out of their misery, Montgomery said. “Put SAS out of business,” Montgomery said. Get some “code monkeys” to work on a modern user interface, he added, and, “You could put them out of business in a week.”

This naturally generated a few laughs, and probably a few nervous ones, at the software company with a troubled history of anti-competitive behavior. Microsoft didn’t make any bellicose statements about world domination last night, but the company clearly has its sights on a big opportunity to make hospitals, pharmaceutical companies, and drug regulators drop their old habits of using paper records and clunky old proprietary programs, in favor of something easier to use and more efficient that captures a patient’s medical records in the doctor’s office, or in clinical trials of experimental drugs.

Zachary Hector, Microsoft’s worldwide pharma industry solutions manager, laid out the broad vision with a slick TV commercial at the beginning. It depicted a young woman who’s pre-diabetic, going for a run, carrying around a PDA-type device that monitored her blood sugar in real-time. The information would be relayed to her doctor, and automatically channeled into a database along with records from other patients in a clinical trial. Later, while going about her everyday chores, the woman would carry a credit card with a fingerprint scanner, which she could stick into an ATM like device to order a diabetes medication.

“Patients are taking more control over their health because of the Internet,” Hector said.

Patients want this kind of predictive information, he said. Market research shows patients care about wellness, prevention, fitness, early detection, disease avoidance, and preventive medicine, Hector says. The age of one-size-fits-all blockbuster drugs like Pfizer’s atorvastatin (Lipitor), which generated more than $12 billion in sales last year, is coming to a close as its patent protection will be dead in 2011. In a world where entire human genomes can be sequenced for as little as $5,000 next spring, a deluge of information is coming about tiny variations in the 6-billion letter string of DNA. Enormous research is being done on how those variations might affect the way certain individuals respond to certain drugs.

When the talk turned to Microsoft’s strategy to capture this emerging information technology opportunity, things started sounding vague and fuzzy for me (although to be fair, Hector ripped through a one-hour presentation in 15 minutes). The company has about 700 people working on its life sciences strategy, Hector said. They are focused on how to speed up innovation by helping researchers and drugmakers bring their treatments to the market faster, and to enable doctors and patients to share health data more efficiently through programs like HealthVault.

Montgomery, who has developed five FDA-approved drugs, didn’t sound like an eager buyer. He told an anecdote about when, as a young doctor in San Francisco in the 1980s on the front lines of the AIDS epidemic, he drafted a 5-page clinical trial design in one afternoon, enrolled more than 500 patients in three weeks, had a successful result three months later, and whipped it up in an FDA application that was approved.

Now, it takes two months to write up a clinical trial plan, he says. Instead of two or three weeks to get an ethics review board to approve the study, it now takes four to six months because of all the cautious lawyering involved, Montgomery says. Then clinical trial sites usually aren’t motivated to recruit patients, because they complain they aren’t making enough money off them from the sponsors, usually a pharmaceutical company. These cultural problems aren’t something that a new software program can fix, he said. “The problem is not really operational systems. We are interacting with a healthcare system that makes it nearly impossible to do clinical studies.”

Still, the information technology sounds like a big part of the problem to this reporter’s ear. Many researchers cling to old Word and Excel programs to manage reams of genomic data, which those programs were never intended to handle, said Bob Webber, CEO of TranSenda International, a Bellevue, WA-based maker of custom life sciences software.

Small software companies are making customized versions of SharePoint so that far-flung collaborators on a global clinical trial can work together more efficiently with patient records. But change isn’t coming easy. One questioner from the audience said his medical center just switched recently from paper records for patients to an electronic system. But people are still making paper copies from the electronic record, he said.

Before ripping too hard into the healthcare Luddites, WBBA president Jack Faris asked the audience whether they voted via electronic machines, or chose the comfort of a paper ballot. Only a couple hands went up. Sounds like the pre-diabetic woman from the TV commercial, who’s willing to have her health records zip around the healthcare system at electron speed, might still be pretty far off in the future.