"This is not for the weak-kneed.”

Jim Routh, chief information security officer at Aetna, is describing the latest trend in healthcare product development: partnering with startups to build custom solutions. It’s an approach that has clear advantages but also opens IT departments to some risk.

The reality today is that many IT platforms are not as mature as they need to be for enterprise-class use – and those that are mission-ready carry a hefty price tag. Plus they often don’t provide all of the functionality healthcare organizations actually need, let alone the flexibility to add new features as needs change.

[Read more Innovation Pulse columns]

Routh and others in the industry, including Mony Weschler, chief application strategist and architect at Montefiore Health System, are shattering the old manner of IT acquisition and picking startups as partners for internal product development.

The result: getting pretty close to exactly what they need and often at a much lower price.

‘Let’s build our own’

We all know the established way of acquiring IT products and services. Write a big check to a technology consultancy, pore through their research reports, and ultimately plunk down top dollar for one of three vendors qualifying as market share leaders.

That’s expensive. And it tends to be safe. But it can leave an organization with a solution that doesn’t meet everyone’s needs.

By custom building applications, the IT staff can deliver software more likely to meet with widespread acceptance.

When Montefiore wanted to improve bar code scanning for medication adherence, for instance, it joined forces with the Menon Group to co-develop software that runs on an iPod, because that form factor, rather than a smartphone or slate, was what Montefiore’s nurses preferred.

That’s just one example. The health system actually has several, including a pilot project with AirStrip, wherein an EKG talks to AirStrip’s methodology for accessing patient data, to essentially deliver that information to clinicians; another involves working with a New York City accelerator on bi-directional SMS communications.

Montefiore even took a co-developed clinical population health analytics platform to market last year with Streamline Health Solutions. Weschler wants to do that with other products.

“We look at 200 companies, pick the 10 we like, partner with one or two,” Weschler said. The effort requires a fairly large investment of time by Montefiore’s IT staff. Much of that involves helping startups understand the demands that are unique to healthcare and the regulatory landscape. 

[See also: Innovation Pulse: Can platforms foster an ecosystem of IT advancement?]

To fast track the non-trivial aspect of finding these startups, Weschler is in the process of building an innovation center, taking a cue from the likes of Cleveland Clinic and Mayo. He has plans to keep the process going with an official Venture Capital fund soon. By taking an equity stake in some of these startups, Montefiore will have a chance to benefit from the work its staff puts into the projects.  “We’re waking up to the fact that we help move the forklift. We do all this work … we want a piece of the profit, too.”

Reward not without risk

Good opportunities don’t just drop from the sky. The best, in fact, require active engagement.

Routh tells a story about how he became involved with Patrick Peterson, a former Cisco engineer who developed IronPort. Peterson tracked Routh down on the streets of New York to introduce himself – after which the two men walked together from Columbus Circle to Penn Station, Peterson dragging luggage to keep up with the fast moving Routh. Peterson used that trip to sell Routh on email authentication idea that became Agari.

“It wasn’t a product when we started. It was an idea and a process with some software around it,” Routh explained. “In cases where there’s game-changing technology I’ll purchase license agreements often before they have a product. I’ll help them develop the product, so it meets our needs and we meet theirs.”

In that particular instance, Routh bought Agari for a song, he said, and it’s not the only time Aetna has taken that approach. Indeed, within a portfolio approximately 150 products strong, about 30 are early-stage startups.

Another category with tremendous promise for Aetna is microvirtualization software. Virtualizing desktops and servers provides a shield for data, because it virtualizes an instance of every object created – be that a Microsoft Word doc, PDF, browser session or others – and each has its own container.

“It basically neuters malware at the immediate endpoint,” Routh explained. “If malware is delivered it cannot do anything because everything is its own session.”

That promise, of course, depends on getting the technology right, which Routh acknowledged hasn't happened just yet. If it works it will be a game-changer and if not, well, then there go the financial, personnel, technological and other resources.

And that’s the risk with managing a product portfolio this way: Substantial savings can be reaped but there are risks which Routh and Weschler believe are worth taking.

“IT products that are mature and command high-dollar value typically work at scale across an enterprise and that’s why they have market share,” Routh said. “On the flip side, for the startups that don't have that pedigree and haven't learned the challenges of market share and scale, there’s engineering work that has to be done that’s part and parcel of implementation.”