by Paddy Padmanabhan

IBM’s insatiable appetite for healthcare data

Aug 13, 2015
AnalyticsBig DataHealthcare Industry

Thereu2019s obviously a great sense of urgency to realize the full potential of Watson, but for that to happen, IBM needs data u2013 and lots of it. Can Watson 'see' far enough into the future to tell if itu2019s going out play out as expected?

IBM dropped another cool $1 billion for imaging software company Merge Healthcare last week, continuing a string of acquisitions from the beginning of the year. Last quarter, IBM announced the acquisitions of Cleveland Clinic spinoff Explorys and population health analytics company Phytel on the same day, rocking the annual HIMSS conference with the announcements.

Around the same time, IBM also announced the formation of the Watson Health unit and teamed up with Apple, Medtronic, and Johnson & Johnson. The intent was clear. Gain access to as much medical data, as quickly as possible, to feed the Watson engine. The acquisition of Merge Healthcare would appear to be just another tidbit in the grand scheme of things.

Can you see?

The Merge Healthcare acquisition would help Watson “see” better, according to IBM’s announcement. And so it would seem. Images are the next frontier in healthcare data analytics, part of the broader expansion of data sources from structured to unstructured and semi-structured sources. Combined with Watson’s cognitive sciences and machine learning capability, the ability to analyze mountains of patient medical information to improve predictive ability got even better for IBM.

One thing is clear. IBM seems to be carrying a shopping basket and a wallet full of money these days for companies that can feed the Watson engine.

Given IBM’s strong focus on healthcare, the next big one would have to be a play for a significant chunk of patient data sitting inside electronic medical record (EMR) systems such as Epic and Cerner. While the Explorys acquisition gave IBM some access to patient medical records through the Cleveland Clinic connection, it’s a drop in the bucket relative to the amount of data sitting inside these big systems.

What next? Or, perhaps, who’s next?

It’s pretty clear that IBM would love to get its hand on a large chunk of EMR data. The only catch is that EMR vendors are notoriously protective about their systems and work hard to keep their proprietary data inside their walled gardens. Interoperability with their systems is one of the biggest issues today in the free exchange of medical information. Privately owned EPIC, the largest among these vendors, has not provided any indications of going public anytime soon, so one wonders if the current owners might sell to a large behemoth such as IBM.

Cerner, riding high on their recent multibillion-dollar Department of Defense (DOD) contract for building a brand new electronic health record (EHR) system, would seem an unlikely seller at this point in time.  

So it would appear that IBM is locked out of maybe a good two-thirds of the patient medical records sitting inside proprietary systems that don’t want to play. At least for now.

A matter of time

The market is likely to take some interesting turns in the near-term, with initiatives involving wearables data gaining ground. Several hospitals including well-known names such as Cedars Sinai, are looking to boost patient engagement by collecting information from activity tracking devices. The IBM partnership with Apple can potentially yield returns through large-scale harvesting of data from the Apple Healthkit.

With the growth of the Internet of Things (IoT), smart medical devices from Medtronic and other device makers are beginning to enable a connected health ecosystem that makes more data accessible now to IBM’s Watson engine.

It’s even possible that relatively smaller players in the EHR market, such as Allscripts, may choose to accept a deal from IBM and exit the market for EMR systems. One or more of these acquisitions will continue to strengthen IBM’s hold on healthcare data in the near term.

There’s obviously a great sense of urgency to realize the full potential of Watson, but for that to happen, IBM needs data – and lots of it. Can Watson “see” far enough into the future to tell if it’s going out play out as expected?