Georgia State Economists Using Big Data To Help Doctors Improve Hospital Discharge Decisions
ATLANTA – Medical doctors will soon be able to use the combined wisdom of hundreds of other doctors and thousands of patient variables to help them determine the safest, healthiest and most cost-effective day on which to discharge each of their patients.
Experimental economists at Georgia State University and doctors at Emory University used tens of thousands of carefully de-identified patient records, econometrics and health information technology to design Clinical Decision Support System software that will help doctors shorten the length of hospital stays and reduce the number of unplanned readmissions that occur within 30 days of discharge.
“No one is approaching discharge decision-making the same way we are,” said Professor James C. Cox of the Andrew Young School of Policy Studies (AYSPS) at Georgia State, who leads the project with Emory surgeon John F. Sweeney.
They and their collaborators, Vjollca Sadiraj (AYSPS) and Kurt E. Schnier (formerly of AYSPS) formed an HIT company, 4C Health Analytics Inc., in August with the help of a Georgia Research Alliance VentureLab Grant, and a patent is pending on their decision-making software.
“Discharge decisions are largely subjective and based on a doctor’s medical education and personal practice experience, a very small information base” said Cox. “Our system will enable doctors to make these decisions with the accumulated wisdom that can be obtained from a large database.”
Their goal is to assist physicians in making timely discharge decisions that decrease length of stay and improve the quality of medical care by decreasing the likelihood of unplanned re-admissions.
“If average length of stay were reduced by 10 percent, the saving to the U.S. healthcare system would exceed $128 billion per year,” he said.
And readmissions are estimated to cost Medicare $26 billion per year, $17 billion of which is potentially avoidable, according to the Center for Medicare and Medicaid Services. Medicare recently fined more than 2,600 hospitals for too-high readmission rates under its Readmissions Reduction Program. These penalties will grow for hospitals that cannot improve their discharge decision-making.
Cox and his team conducted three waves of experiments with medical students and resident physicians to test the efficacy of the software. The research was funded with grants from the National Institutes of Health.
“We’ve tapped into the information and experience available from a large population of discharges at a major university hospital – about 25,000 a year. We’re making the precision that can be gained from all of this information available to be brought to bear at the point of care where the discharge decision is made,” he said.
To learn more about this project, go to http://excen.gsu.edu/workingpapers/GSU_EXCEN_WP_2014-07.pdf.