What started as a young Parkland physician trying to solve a problem could blow up into an enormous enterprise that could save hundreds of hospitals money and have a huge impact on patient care.
When he was assistant medical director for medical services at Parkland in 2007, Ruben Amarasingham noticed demographic similarities in patients who were readmitted to the hospital. He believed patients’ medical records could be mined ahead of time for those demographic clues to alert providers to take extra caution with certain patients to prevent unnecessary admissions.
With colleagues, Amarasingham developed a software model that would scan patient data prior to discharge. The model flags factors such as frequent changes of home address in the past year or a low number of emergency contacts or family relationships.
In 2009, Amarasingham launched the Parkland Intelligent e-Coordination and Evaluation System, or PIECES—a software application he likens to a smartphone app that sits atop a health system’s electronic health record to identify high-risk patients. The algorithm digitally analyzes the patient’s clinical and social factors in patients’ records and uses advanced predictive modeling to predict likely outcomes without extra intervention.
The result: Parkland has cut 30-day readmissions for Medicare patients with heart failure, including readmissions to all hospitals, by 31 percent, with an estimated savings to the hospital of $500,000 with no increase in staffing. Parkland’s readmission penalty rates from the Centers for Medicare and Medicaid Services were 10 times below the national average of similar safety-net hospitals. According to a Commonwealth Fund study released Monday, safety-net hospitals are 30 percent more likely to have 30-day hospital readmission rates above the national average, compared with non–safety-net hospitals, and are therefore disproportionately impacted Medicare’ readmission penalties.
Ron Anderson, Parkland’s former CEO and now senior adviser, said of the software: “PIECES has transformed both health care delivery and health outcomes for patients at Parkland, and has implications for reducing costs and improving safety for health systems nationwide.”
October was a banner month for the 39-year-old Amarasingham. The Robert Wood Johnson Foundation named him one of 10 individuals who are 40 years of age and younger and who offer great promise for leading the way to improved U.S. healthcare. He received an individual award of $40,000. Anderson nominated him for the award. He is profiled in the current edition of the journal Health Affairs.
Parkland also announced in October that Amarasingham would be chief executive officer of the newly created Parkland Center for Clinical Innovation (PCCI), an affiliated non-profit research and development corporation that will expand Amarasingham’s work to other health conditions and help other hospital use predictive modeling.
Amarasingham said PCCI has 30 staff members and is split into four groups: advanced analytics, which does mathematical modeling; software technicians, who will develop and support software; and a clinical core of physicians with skill sets similar to Amarasingham’s, and an operations group that will run the business.
In 2010, Amarasingham published an article about his model in the journal Medical Care. A year later, a comprehensive review in the Journal of the American Medical Association, involving more than 7,000 published works, determined Amarasingham’s predictive modeling approach to be among the best at identifying patients at high risk for readmission.
Because of this broad exposure, Amarasingham said Parkland has been contacted by 200 hospitals about PIECES. He said PCCI’s goal is to help hospitals use predictive modeling in whatever method they prefer: by installing PIECES at their facilities; help the hospitals create their own software; or simply help them analyze their data. PIECES can be accessed by cloud computing or it can be installed at a local facility.
PCCI is working on software to flag high-risk patients for other conditions, including diabetes, chronic kidney disease, heart attack, pneumonia and sepsis. Not coincidentally, these conditions include those targeted by Medicare for readmission penalties. Amarasingham said PCCI’s goal is to supply hospitals with affordable software that would predict hospital readmissions regardless of medical condition or patient circumstance.
PCCI revenue will be plowed back into the organization for further software development. So far, the organization has received about $4.5 million in grants and donations from organizations such as the Commonwealth Fund, the Gordon and Betty Moore Foundation, and the National Institutes of Health.
Texas Health Resources (THR) and UT Southwestern have collaborated with Parkland in field-testing PIECES. THR plans to use PIECES at Harris Methodist Hospital Hurst-Euless-Bedford by summer 2013, focusing on heart failure patients, according to Ferdinand Velasco, M.D., THR chief medical information officer.
The son of Sri Lankan immigrants, Amarasingham began computer programming at age 10 and developed an interest in health information technology and predictive modeling as a Robert Wood Johnson Clinical Scholar at the Johns Hopkins Medical Institutions.
“The approach you learn as a computer programmer has served me well in approaching the challenges of medicine. I learned at Johns Hopkins how to combine the different passions of my life,” he said.
Steve Jacob is editor of D Healthcare Daily and author of the new book Health Care in 2020: Where Uncertain Reform, Bad Habits, Too Few Doctors and Skyrocketing Costs Are Taking Us. He can be reached at [email protected].
