In clinical trials, technology is an elephant. It will do extraordinary work for some masters. And it creates a mess for others.
Sylva Collins is a Ballanchine of elephants, getting them into tutus and ballet slippers and spinning lightly on their big, gray toes. Based in New Jersey, the Armenian Collins is VP of electronic data management at Novartis. Thanks to her, the company has increased its capacity to run clinical trials and done so specifically to handle the upsurge of compounds from drug discovery scientists in the genomic age.
In November, Collins says, Novartis recalculated the savings from its use of electronic data capture (EDC). The company now puts the savings at $100 million annually. Cumulatively, the amount is even larger. "In the last two and a half years, since we have started scaling up EDC," says Collins, "there is an estimate of a quarter of a billion dollars."
Collins herself will not say what each of her laptops dedicated to EDC costs - but the figure, she insists, is far less than $15,000. Collins also declines to discuss one of her biggest supporters at the company. But it's hardly a secret that Novartis CEO Daniel Vasella personally recruited her from Bayer, another EDC leader, where she'd first started running EDC trials in 1989.
A better dollar figure to cite, she believes as a Ph.D statistician, is the cost per page of clinical data: $4.60. That includes hardware and software that support data management, including the company's EDC help desk, software maintenance, hardware management, process and standards development, user training and corporate overhead.
"Perhaps the problem with the other companies is that they view EDC as a technology project," says Collins. "It is a radical change. It is not an IT project. That is the biggest pitfall that other companies are falling into. The key is to keep the EDC system, the software, simple. Just focus. Look at the big picture, not just technology. Technology is a small component of implementing EDC."
Then Collins reels off more numbers - numbers that no American pharmaceutical company may ever be able to match. Percentage of Novartis Phase II and III trials using EDC: nearly 100%. Percentage of Phase I trials using EDC: nearly 60%. Continents on which Novartis is running EDC-based trials: six. Investigative sites: 8,000. Nations using EDC: 50. Total number of trials started or completed using EDC: 290, as of May 2004. Time to database lock: four days, vs. 10 weeks in the paper era. Reduction in number of queries, per 1,000 data points: 92%.
Though U.S. companies often claim to lead the world in using technology, it is the Europeans who dominate in clinical trials. This phenomenon is not new: Europeans also rule in cutting-edge commercial airliners, just as the Japanese do in the auto trade. The European lead is hardly insurmountable: it appears that Merck (with help from Microsoft and DataLabs) is on the verge of using technology to achieve similar results.
But all too often, in discussions of the adoption of clinical technology, the U.S. pharmaceutical industry has become distracted by a parlor game of vendors and sponsors blaming each other. As most eCliniqua readers know, Collins sidestepped that elephant dung by licensing open-source software from another European company - and developing it internally.
Note: Bio-IT World magazine will publish a longer interview with Collins, who will be speaking at next month's Drug Information Association (DIA) conference in Washington, D.C., and at the dinner honoring the winners of our publication's Best Practices Awards.