Health IT

How are hospitals responding to startup queries to use de-identified patient data? Cautiously

question-markHospitals generate a lot of data and startups know it. Entrepreneurs want access to that data to field test their own technology platforms. But hospitals are being very cautious with even de-identified data and are choosing who gets to see that info very carefully.

A check of three major health systems generated a few different approaches to managing data requests.

Penn Medicine set up a protocol for reviewing and prioritizing external requests for its data and integration with its technology suite. Brian Wells, associate CIO of health technology and research computing, mentioned it at a Philadelphia-area HIMSS chapter seminar. He joked that it sounded a bit like Get Smart in his use of the word “Control” to describe the group that reviews these queries.

Penn Medicine is a partner in the DreamIt Health accelerator, along with Independence Blue Cross, so it stands to reason that Penn would be getting more queries from startups.


Although he declined to say how many queries it is fielding, Wells responded in an email to the follow-up question: “It is an increasing amount as we continue to create a rich data repository and a more integrated and automated health care information technology environment.”

A spokeswoman from Johns Hopkins Medicine said in an emailed statement that it only allows Johns Hopkins University faculty trained in research to access these data and share “where appropriate”.

“This is for the protection of those patients in the database and to make certain the results of these queries are interpreted correctly. Clinical data are complex and difficult to interpret without thorough knowledge of how the data were collected.”

Other health systems see this information as way too valuable to pass on and instead are spinning out companies to best utilize these data. Case in point, Geisinger Health System. It was way ahead when it started MedMining through its venture arm in 2006. The idea for the company started because it was helping its own researchers answer questions such as how patient care can be enhanced and evaluating the cost effectiveness of certain drugs. Its core customers are medical device and pharmaceutical companies who want to better understand the impact of their drugs and devices on people. It also gets patient safety and epidemiology queries.

Since it had begun using electronic medical records in 1996, it was in a position to take full advantage of its data including having what it believes is the health care industry’s largest reportable database of lab results.

Julia Rockey, the senior director of operations, was the second employee hired after founder and CEO Jim Peters. Now the company is expanding to 15. an emerging trend is it is increasingly getting queries from marketers. In a a recent phone interview she described the meticulous process it goes through in screening queries from these companies and scrubbing the data it packages. And yet the level of depth of the data it provides is what draws customers to them. IT can link clinical data with claims data. They can specify whether patients are still alive or when they died through a monthly feed from the National Death Index.

It is also looking into developing a sensible interface for these data sets to make it easier to organize the information they are seeking.

We’re bringing active investors and corporate development execs together with promising startups at MedCity News INVEST. Join us April 12-13, 2016, in Chicago for the premier national investing conference based in the Midwest.

No comments