IBM Makes New Health Care Push with Predictive Analytics, Process Management

by   |   October 23, 2012 10:51 pm   |   0 Comments

LAS VEGAS—There’s a lot of room for improving the health care system, from patient outcomes to inconsistent business processes and cost structures. IBM introduced an offering to address all of these issues using process management expertise, predictive analytics and natural language processing technology from the Watson system made famous on the Jeopardy TV game show.

Related Stories

A prescription for better analytics at Atrius Health.

       Read more»

Business problems suited to big data analytics.

       Read more»

The Patient Care and Insights offering, which IBM unveiled Oct. 23 to coincide with the company’s’ Information on Demand 2012 conference here, is designed to provide doctors with personalized care recommendations for patients, and to support tracking of follow-up care programs and health outcomes. Analytics guide the process by giving insights into how an individual’s health conditions compare to his own history and to the histories and outcomes of other patients served by the health care provider.

The offering represents an important use case in the big data field because it enables users to collect and analyze large volumes of disparate records in a variety of formats, including electronic records and text documents, said Craig R. Rhinehart, director of IBM enterprise content management strategy. Evaluating a patient’s current condition becomes a deeper exercise when caregivers also can examine that patient’s symptoms in the context of an analysis of longitudinal data—both on the individual and others in a statistically significant cohort.

When implemented, IBM believes the system will allow health care providers to be more effective and efficient and lead to more precise and personalized care, in which doctors and patients can make more informed choices about courses of treatment using both analytics readouts and clinical care guidelines, Rhinehart said.

Other potential benefits cited include: predictive modeling to enable caregivers to help manage a patient’s health care needs in a forward-looking fashion. And, with data providing insights about outcomes for patients, a new capability for matching patients with doctors who have proven to be effective treating certain kinds of patients and conditions.

Several pieces make up Patient Care and Insights. Content analytics enable the collection and analysis of the structured, and especially, unstructured data types. (“The gold is in the conversations with the doctor, the captured notes, the text of lab results,” Rhinehart said.) What IBM calls “similarity analytics” uses natural language processing and machine learning technologies from the Watson project to analyze thousands of variables for a patient’s condition, medical history, comparison to others with similar conditions and potential outcomes, said Shahram Ebadollahi, senior manager, healthcare systems and analytics research.

SPSS software provides predictive analytics capabilities. The company also includes a care management process to track outcomes. The whole package is part of an effort to take existing pieces of IBM’s portfolio and extend them into a health care-specific offering, Rhinehart said.

IBM researchers, who have been developing the similarity analytics technology over the past three years, can use the experiences of its customers to further refine the system over time, Ebadollahi said. And it could be tailored for other industries. “The underlying methodology is very generic,” he said.

The new offering will have a global rollout, but its relevance to the United States health care reform efforts earned a mention in a company statement.

“It supports emerging health care business models such as accountable care, complex care management, and the patient-centered medical home, which aligns with our core vision,” said Dr. David Hanekom, chief medical officer at MDdatacor, an IBM partner. “The end result will be better care for patients and lower costs for health plans and providers.”

Michael Goldberg is editor of Data Informed. Email him at

Tags: , , ,

Post a Comment

Your email is never published nor shared. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>