IHIE will initially collaborate with Predixion on an application that can help hospitals predict the likelihood of a given patient being readmitted within 30 days after discharge. Healthcare organizations are particularly interested in this kind of tool because Medicare penalizes them financially for preventable readmissions.
Harold Apple, CEO of IHIE, told InformationWeek Healthcare that his organization plans to validate Predixion's risk stratification tool with the help of two of its customers. The software's algorithm uses the characteristics of patients who have been readmitted to predict which discharged patients are likely to be rehospitalized. By stratifying patients according to their risk of being readmitted, the tool enables providers to intervene with the highest risk patients.
[ Will mobile health apps be ruled medical devices? FDA Warns Mobile Health Apps Makers. ]
IHIE plans to work with Predixion to create other kinds of analytic tools that can be applied to its clinical data repository, which encompasses 20 terabytes of data and 10 million patient records.
"In conjunction with Predixion Software, we look forward to leveraging the [data warehouse] to develop and deploy predictive applications that enable physicians to reduce hospital readmissions, choose optimal therapies, reduce redundant testing and avoid drug-drug interactions, among other life-saving and efficiency-generating efforts," said Curt Sellke, IHIE's VP of analytics, in a news release.
The analytics jointly developed by Predixion and IHIE will be marketed exclusively in Indiana, IHIE's Apple said. However, IHIE has formed a subsidiary to sell other services outside of the state, starting in the last quarter of this year.
Historically, Apple said, IHIE has used its clinical data repository for measuring physician quality as part of IHIE's Quality Health First program, which also does public reporting of quality data. Some 2,500 practices participate in this program, he said.
"That was a start" toward leveraging IHIE's immense data resources, he said. "But long term, our value is going to be helping identify, on a predictive basis, activities that will allow providers to be proactive in patient care."
This is an increasingly common strategy among HIEs that view the emergence of accountable care organizations (ACOs) and other value-based payment arrangements as a growth opportunity. According to a recent study in Health Affairs, 87% of operational public HIEs plan to get involved in these kinds of arrangements. IHIE, similarly, is beginning to work with Indiana ACOs.
One advantage that HIEs offer in the ACO environment, Apple said, is that they can enable hospitals caring for particular patients to view another hospital's data on those patients that would otherwise be invisible to them. This value can be "tightly woven into their cost-benefit analysis regarding what they pay for HIE," he added. "You couldn't say that three or four years ago.
IHIE recently launched a service called ADT Alerts to help ACOs and other healthcare organizations that take financial risk track where their patients receive care. Using real-time feeds from ADT systems in hospitals, Apple said, "We know within minutes when a patient has been admitted to an ER or a health facility" in a different enterprise. This ability to detect when patients go "out of network" enables enterprises to "be proactive in care management of those patients" so that they seek more of their care from network providers, Apple noted. This is a priority for organizations that have fixed budgets and need to control use of services.
Although IHIE has long delivered lab results from hospitals to community physicians by fax and other electronic means, it has very limited connectivity with doctors who are outside of the enterprises that are IHIE customers. So when it provides alerts to a hospital about a patient being admitted to a non-affiliated hospital, it depends on the first hospital to notify the patient's primary care doctor, he said.