SAN FRANCISCO, Oct. 2, 2012 (GLOBE NEWSWIRE) -- Practice Fusion, the nation's largest doctor-patient community, has named Datalanche the winner of its Analyze This! data challenge hosted on Kaggle, a platform for predictive modeling competitions. Using a HIPAA-compliant sample dataset of 10,000 Practice Fusion records – including lab results, diagnoses, medications, allergies, immunizations, vital signs and other key markers of health behavior – Datalanche developed an intuitive application to inform patients about medical conditions, common medications for those conditions and additional insights into treatments.
"Datalanche allows patients and physicians to find tailored answers to complex health questions quickly and intuitively," said Ryan Pedela, Datalanche co-founder and developer, "We built it using de-identified patient data because this is the most accurate, real-time source for health information on a national scale—from the cause of certain symptoms to the effectiveness of specific treatments."
Pedela realized the power of information-driven technology to fix major gaps in healthcare after working as a software engineer in Silicon Valley for several years. Large, de-identified datasets from real-world electronic records and other medical sources could help to identify lifesaving solutions to these problems, he recognized, as long as that data was made accessible and meaningful. Datalanche aims to provide this meaningful and intuitive structure using data made available by organizations and companies like Practice Fusion.
"Data holds profound possibilities for researchers, doctors and patients to improve the way they approach complex health problems," said Matthew Douglass, Practice Fusion's Vice President of Engineering. "Datalanche helps prove that this data can actively inform health studies, identify health trends and enhance health care as it is practiced in real-time."
Analyze This! entrants were asked to creatively combine Practice Fusion's clinical dataset, posted on Kaggle, with one or more public datasets available at Data.gov, in a way that addresses an important public health topic. Datalanche will receive $7,000 for submitting the winning entry—a video demonstration of their application can be viewed here. Other finalists in the Health 2.0 Analyze This! competition included:
- Whatkindofpersongoestothedoctor?: A comparison revealing EHR patients to be more female, more diabetic, more obese and less likely to smoke than the general population. (Yasmin Lucero, Pasadena, CA)
- WordTornado: A health data visualization designed to help EHR users identify important patient information quickly and intuitively. (Shea Parkes, Indianapolis, IN)
"We were thrilled to open up one of the world's richest de-identified medical datasets with Practice Fusion in their Analyze This! Challenge," said Anthony Goldbloom, founder and CEO of Kaggle. "Datalanche and the other impressive entrants prove what we see consistently at Kaggle— crowdsourcing is a powerful way to bring diverse insights to valuable data in order to address tough problems."
In a second challenge, Practice Fusion solicited ideas for prediction problems based on the dataset. This competition first solicited ideas for prediction problems on Kaggle's new Prospect platform for data exploration and analysis, visualization and problem identification. The top-voted submission – predicting type 2 diabetes – was the basis for a competition that attracted 170 participants and 2,200 entries. The top three prize winners were:
- Jose A. Guerrero (Sevilla, Spain)
- Matthew Berseth (Jacksonville, FL)
- Shashishekhar Godbole (Mumbai, India)
With a total of $20,000 in prize money, these Analyze This! data challenges comprised Practice Fusion's biggest competition to date. The company has recently become a hub of disruptive health care innovation, incubating startups such as 100Plus, whose CEO Chris Hogg is also a former data challenge winner. Practice Fusion will honor Datalanche during Health Innovation Week in San Francisco October 3-12. Reporters interested in meeting Datalanche or attending the event should signup online or email sschlegel@practicefusion.com.
About Practice Fusion's Research Division
Practice Fusion's Research Division provides information from one of the largest, longitudinal clinical databases in the US for the purposes of clinical research and public health analysis. With a focus on real-time insight into adverse drug reactions, disease outbreaks and other national health issues, Practice Fusion's Research Division powers studies that are providing new insight into healthcare and how it is delivered. All research information utilized is fully de-identified and compliant with HIPAA cfr 164.514. For more information, please visit http://www.practicefusion.com/research.