Healthcare and other companies will have to push more development of the technology they'll need for the challenges ahead.
7 Big Data Solutions Try To Reshape Healthcare
(click image for larger view and for slideshow)
Allscripts' $235 million acquisition this week of Israeli software company dbMotion is interesting on its own. But even more interesting is the story behind dbMotion's development, a story that healthcare and other industries will need to replicate more often to solve some of their knottiest problems.
dbMotion lets healthcare companies take data from two (or more) different electronic records systems and normalize it so the different sources can be used together. That data normalization helps hospitals analyze the quality of their care versus the cost, and it lets patients access their data from the many places they've been treated. Allscripts, a top electronic medical record vendor, bought dbMotion because that kind of data analytics and sharing is where the value and growth will be in healthcare tech, now that most hospitals and big physician practices have a core electronic record system in place.
The story behind dbMotion's involves University of Pittsburgh Medical Center. UPMC was dbMotion's biggest shareholder, having invested $35 million in dbMotion in 2006. Why did UPMC invest?
One, UPMC itself had a big problem with data interoperability, so its technical teams poured a lot of their knowledge and time into making dbMotion better to solve the hospital group's problem. And two, it knew many other healthcare providers would have the same problems, making dbMotion a solid business investment if the team could make the software work.
Strategic investments such as this one are nothing new. Companies in a range of industries -- retail, transportation, logistics, financial services, consumer goods -- regularly invest in technology companies to boost their core businesses and reap financial returns. Often they join in alongside pure financial investors, as was the case with dbMotion.
UPMC CIO Dan Drawbaugh sees more interest than ever among big healthcare providers such as UPMC, Kaiser Permanente and Mayo Clinic to do these kinds of co-development projects. "Across the industry we're seeing more [providers] pushing the IT envelope," says Drawbaugh, who was CIO back when UPMC decided to invest in dbMotion. "The discussions around partnering are greater than I've ever seen."
One factor is the economics. Healthcare providers of all sizes are feeling the squeeze on profit margins, so big research institutions in particular are looking at strategic investing as another profit source.
Institutions also know they need breakthrough technologies to meet the demands of healthcare reform, which will push providers for better care at a lower cost. These institutions have an intimate knowledge of the problems and could help speed up new technology development.
And third is the opportunity. Health records are (or soon will be) mostly digitized, which provides the foundation to do new things in analytics and data sharing. There's a surge of investment in health IT, and these providers hope to cash in for contributing their know-how to new products. UPMC, IBM and Oracle have pledged $100 million to an ambitious healthcare analytics initiative, for example.
Other industries are also pursuing new models for faster, smarter tech development. Procter & Gamble CEO Filippo Passerini recently brought together a roster of marquee companies, including Boeing, FedEx, GE and Goldman Sachs, to brainstorm on ways they could speed up vendors' development of analytics software. P&G last month entered a co-development deal with a small startup, Verix, that does data integration for consumer goods and life sciences companies, similar to what dbMotion does for healthcare.
Formalizing The Co-Development Model
In 2010, UPMC assembled a team of people focused on this kind of development work, called the Technology Development Center. It now employs about 90 people, ranging from financial analysts to engineers and designers.
UPMC could just give vendors closer access to its clinicians. But TDC director Rebecca Kaul makes the case that its team can bring a more nuanced understanding of UPMC's needs and influence development to meet them. "It always starts with a problem we have to solve at UPMC," Kaul says. It's currently doing projects with Nuance and Optum related to natural language processing to understand unstructured data, for example.
TDC is working on three main problem areas:
1. Visualization: As data volumes grow, physicians need tools that hone in on the data that matters most to their decision-making.
2. Collaboration: UPMC is focusing here on providing the right data to an entire group of people involved in care -- to understand a patient's total health and relevant history, not just his knee if you're the doctor doing a knee replacement. (It recently completed a development effort with Alcatel-Lucent on this front.)
3. Data transformation: This is where it's working with Nuance and Optum, trying to pull insights from verbal notes to contribute to electronic record data.
A fourth area it would like to work on is decision support, giving clinicians the right data and decision frameworks at a point of care to help them make tough calls on diagnoses and treatments.
We're always going to need disruptive tech startups to rattle industries from the outside. And entrepreneurial leaders continue to be the power behind companies such as dbMotion, so I don't mean to overstate the role of established companies. But industry knowledge can be vital in driving a startup's product and speed of development. Particularly in areas such as improving analytics and in gathering data through an "Internet of Things" of connected devices, co-development projects will be indispensible.
InformationWeek Tech Digest, Nov. 10, 2014Just 30% of respondents to our new survey say their companies are very or extremely effective at identifying critical data and analyzing it to make decisions, down from 42% in 2013. What gives?