A recent press release from the University of Washington Tacoma provides a great example of how Big Data and Advanced Analytics are having a direct impact on the Healthcare industry. Readmission rates are a hot topic for hospital and health care systems across the country, so the application of cutting edge data science to the task of reducing readmissions could be a literal life-saver.
This is exactly what the team at MultiCare Health System in Washington found when they asked for assitance from the Center fro Data Science at UW Tacoma. With the help of an Azure for Research grant from Microsoft, the team was able to leverage Azure to accelerate the data analysis and model building process to develop their “Risk-O-Meter” predictive analytics tool. This tool is able to predict 30-day readmissions with a walloping 82 percent accuracy rate. Ccontrast this against a standard accuracy rate of predictive models of around 60 percent.
While initial use of the tool was granted to clinicians, the team was able to expand its availability to patients as well — putting in their hands the ability to predict how likely they are to be readmitted within 30 days given biometric and behavioral inputs.
The collective UW Tacoma and MultiCare teams are now looking towards commercialization of the tool as a “readmission score as a service” offering. This could help put a serious dent in the estimated $26B annual national cost of 30-day readmissions. Yes — that’s “Billion” with a B. To healthcare companies, this is obviously an area with serious potential for cost savings. And that doesn’t even take into account that the Centers for Medicare and Medicaid Services (CMS) will soon start levying penalties on companies exceeding threshold readmission rates.
And of course, this is obviously of great benefit to patients, who can see a direct link between following their courses of care and avoiding readmission. The coming together of Big Data technology and Healthcare know-how will continue to provide advances in quality of care and cost savings. To find out more about how Microsoft and Azure technology play a role, follow these links:
http://research.microsoft.com/apps/video/default.aspx?id=217993&r=1
http://research.microsoft.com/apps/video/default.aspx?id=159290
http://azure.microsoft.com/en-us/services/machine-learning/