"Appliances" translates to purpose-built servers that drive analytics speed and efficiency. Those don't come cheap.
Khorey says it's money UPMC would spend anyway, and the system "will drive efficiencies that will pay for it." But the investment is also a mix of outcomes-oriented work and research, because personalized medicine based on analyzing genomic data is still a bit of a science project. "New models of care and good science is the UPMC agenda for personalized medicine," she says.
When UPMC looked around a year ago, no comparable institution had made a similar investment in personalized medicine analytics R&D. "We have friends at Intermountain and Kaiser and other large organizations that have huge analytics shops, but they tend to be focused on one facet of the problem, whereas we're basically saying it's a personalized medicine agenda, and we're going to solve the information management problem for all," Khorey says.
Of course the money can't just sit there for five years with nothing to show for it. "The goal is to do something quickly," Khorey says. She rattled off a punch list of early efforts: patient outcomes related to cardiothoracic patients; a small, intersecting set of breast cancer genomic and phenotypic data; some CMS quality compliance reporting. If any of them click, they can be scaled up quickly, she says.