As more providers continue to delve deeper into adopting value-based care models, many are seeing their costs go down as outcomes improve. But they are also finding the areas in which the mathematical realities of an economic model don’t always align with the transformation needed to optimize an often unpredictable clinical setting.

The issue comes to a head when dealing with high-risk patients. These are the 10 percent of the patient population that account for 70 percent of health costs, and they create major challenges for providers and their practices which increasingly are being reimbursed based on outcomes. The challenge becomes even greater when these patients leave the hospital.

Take, for example, an elderly patient with uncontrolled diabetes and coronary heart disease who lives with his daughter in a major city. That patient is going to have a very different post-discharge experience and wildly different set of needs than a patient with the same diagnoses who lives alone in a rural area.

Traditional care management risk stratification models have no real way to systematically account for that distinction, despite the fact that it is widely known that social determinants—factors such as socioeconomic status, education, physical environment and social support networks—have a significant bearing on health outcomes.

The ideal care management solution would not only be able to connect all the dots that constitute those determinants of health, but would also surface personalized insights that could enhance care and help activate individuals to productive self-management.

That vision is clearer today because of advances in cognitive computing, a form of augmented intelligence that can meld and analyze different types of data, producing highly tailored insights about each individual patient across an entire population. Here are some of the areas where cognitive computing could support care management efforts.

Predictive analytics
The first challenge of care management is determining which patients will benefit most. Optimal outcomes do not necessarily result from high expenditures. Thus, individuals who are already generating high costs will be an obvious target to identify unnecessary spending that is not leading to better outcomes. However, predictive analytics can now be deployed to identify people who are at increased risk of becoming high-cost and might be helped most by enrollment in a care management program.

Because cognitive computing solutions can generate insights from searches of large data sets, and learn continually from the feedback it receives about its conclusions, it could be possible for clinicians to base care management enrollment decisions on sophisticated analyses that relate insights from wildly diverse sources such as medical literature, electronic health record (EHR) data, social determinants of health and even individual genetic characteristics.

Intelligent patient engagement
After patients are enrolled in a care management program, the chief imperative for care managers is engagement. They need to foster a quick rapport rooted in trust and informed by data from multiple care settings. Optimally, this information needs to be accessed quickly and efficiently so care managers can see an analysis of the latest information before they contact the individual.

Programs leveraging cognitive computing could help a care manager prepare for an interview with an individual by providing insights gleaned from a variety of sources, including unstructured documents such as clinician notes, nurse intake questions and other records. Access to this aggregated disparate information can help identify unnoticed risks and narrow down possible questions that should be asked, focus the clinician on specific assessments that should be conducted and offer insights based on the experiences of similar individuals and care managers.

Workflow efficiency
For any of this to work, however, scalability is key. Care managers typically oversee hundreds of patients. Cognitive technologies are designed to help care managers assemble a holistic view of a person’s health from across myriad data sets without requiring the care manager to dig for information in charts and other systems. These technologies may even help care managers prioritize the day’s work by focusing on the most critical cases first.

Care management is a critical aspect of value-based care, but it’s historically one of the most difficult to operate. While it’s still early in the implementation of cognitive computing capabilities into the care management setting, the experiences of early adopters will yield a valuable roadmap to scale cost-effective care management as a core component of optimizing care.

Register or login for access to this item and much more

All Health Data Management content is archived after seven days.

Community members receive:
  • All recent and archived articles
  • Conference offers and updates
  • A full menu of enewsletter options
  • Web seminars, white papers, ebooks

Don't have an account? Register for Free Unlimited Access

Anil Jain, MD

Anil Jain, MD

Anil Jain, MD, is vice president and chief health informatics officer at IBM Watson Health.