Follow Datanami:
October 1, 2013

Healthcare IT Activist Releases 30 Million Edge Prescription Dataset

The Open Source Health IT movement got boost last week when healthcare data journalist and author, Fred Trotter, released a ~30 million edge dataset on prescription patterns from the Medicare prescription drug benefit (Medicare Part D). The release, which was announced at the Strata Rx conference, follows a previous release made last year – a 50 million row dataset that detailed how physicians worked together within hospital systems, labs, and radiology centers.

The release of the data is a continuing effort by Fred Trotter, author of the book Hacking Healthcare, and co-founder of The DocGraph Journal, a site dedicate to reporting and building community around the datasets that he’s compiling.

Using the data he released at the 2012 Strata Rx conference as a foundation, Trotter is building an open source data resource he refers to as a “Doctor Social Graph,” (a.k.a. Doc Graph). Using data visualization tools, Trotter was able to show a visualization of the DocGraph data, revealing physicians and their relationships within the industry. “You can see how much a doctor works with different providers,” explained Trotter demoing the graph.

Since the release, Trotter says that the data has been integrated in several healthcare related information products. One such company is HealthTap, which has taken the information to create a network called “DOConnect,” a sort of LinkedIn for connecting patients with doctors, complete with doctor and patient recommendations, medals, and “DocScores.”

With this year’s data release, Trotter says he is adding a layer of data on top of the foundational data in the DocGraph, showing the prescribing patterns of doctors interacting with Medicare Part D patients. “What I’m about to do is give you seven figures of data,” hyped Trotter in his address at the Strata Rx conference. “Which is to say, a week ago if you wanted to get this data, you had to have a million dollars.”

The dataset, said Trotter, piggybacked off of a Freedom of Information Act request submitted by ProPublica, a nonprofit journalistic group focused on investigative journalism in the public interest. Trotter explained that his group replicated the ProPublica FOIA request, and merged it into the DocGraph, providing details on how different physicians prescribed medication. “You can see these patterns an can now do an analysis with open data not only on how a physician prescribes, but with these datasets together, you can see how a network of providers is prescribing.”

Trotter said he was worried last year when he first made the initial data release, stating his concerns that the American Medical Association would sue him for releasing the data. As a precaution to get the data out there, he distributed it on USB sticks so that the data couldn’t be taken back. “As you can imagine, these are uncomfortable datasets for people to have out there.”

Trotter says that they are currently working on adding state data as an overlay in the DocGraph, though he says progress has been tedious. “Part of the thing that has held us up is that it’s actually fairly complicated to get these things from a legal perspective because many states say that they will give you the data as long as you don’t resell them,” he explained. “Well, we want to resell them for nothing – we want to give them away.”

While the dataset that was announced last week is being given away freely to those who can prove they were at the Strata Rx conference last week, Trotter says that those who weren’t there can get their hands on it by participating in the crowdfunding for one of his next big data healthcare projects, including a crowdsourced food database that will be used to build hyper-targeted apps for patients and caregivers “in order to make living with food related healthcare conditions easier.”

Related items:

Syapse Wins Startup Showcase at Strata Rx 

Big Data Aiming to Lower Big Prescription Bills 

Healthcare Alliances Moving the Ball on Big Data 

Datanami