Blog Post | April 24, 2017

Managing Serious Illnesses: Predictive Analytics for Patient Selection & Population Health Management

Written by: Terri Maxwell Ph.D., APRN, Chief Clinical Officer

Join Terri Maxwell, Ph.D., APRN, chief clinical officer, Turn-Key Health, and Ian Duncan, FSA, FIA, FCIA, FCA, MAAA, Adjunct Professor, University of California Santa Barbara, when they present a population health approach to serious illness care at the National Hospice and Palliative Care Organization’s (NHPCO) Management and Leadership Conference, May 2, 2017, 3:15 p.m.-4:15 p.m., Washington Hilton, Washington, D.C.

The statistics are shocking: patients in their final six months of life account for about one-third of Medicare’s budget, too often as a result of being subjected to non-beneficial treatment (NBT) that is burdensome and costly. Broadly defined, NBT is an aggressive treatment that significantly diminishes the quality of life, yet is unlikely to extend survival.

Terri Maxwell Turn Key Health

Terri Maxwell, Ph.D., APRN
Chief Clinical Officer, Turn-Key Health

Given that the number of people likely to suffer as a result of serious illness in the United States is significant – and growing — payers must find innovative new ways to address the cost and quality challenges inherent in treating this vulnerable population. Incorporating palliative care into a population health strategy is key to improving care and controlling costs for persons with serious illness. However, identifying this unique population and implementing specialized clinical programs that improve outcomes is challenging.

Predictive analytics is a process that uses data mining and probability to forecast outcomes or determine the likelihood of experiencing an event, such as those likely to be high utilizers of NBT at the end of life. A data-driven approach to population health and palliative care successfully identifies individuals such as this who would most likely benefit from palliative interventions and generate cost savings that can be used to support the program.

It’s important to keep in mind that if the proper infrastructure, staffing, and resources to intervene with something predicated to happen are lacking, the potential of utilizing patient data will fall short. We need to move beyond predictive analytics for member identification to prescriptive analytics, which includes evidence, clinical protocols and patient outcomes to achieve goals.

Turn-Key Health has developed a successful population health management (PHM) approach that utilizes predictive analytics to identify members who are at-risk for NBT or hospitalization who would benefit from palliative care. Turn-Key combines this proactive approach to member identification with a network of specially trained clinical teams of nurses and social workers who utilize Turn-Key’s palliative assessments embedded in a web-based platform and follow Turn-Keys care model and pathways to offer a level of a specialized home and phone-based palliative support that has gone missing from traditional care models.

These clinicians are communications specialists, highly experienced and skilled in leading sensitive discussions with patients and their families to establish care plans that meet their personal needs and wishes.  As a result, people are more informed and may elect different options when they better understand their prognosis and values.  This also opens opportunities for making an earlier election of the Medicare hospice benefit.

To test and validate the results of PIM™, a one-year program was conducted for over 200 seniors enrolled in a Medicare Advantage plan:

  • Overall reduction in healthcare expenditures 34 percent
  • Lower hospitalizations 32 percent
  • Reduced ICU days 37 percent
  • Reduced hospital readmissions 61 percent
  • Reduction related to earlier and appropriate election of the hospice Medicare benefit 15 percent
  • Medicare benefit claims down per member per month 19 percent

Turn-Key’s continuous quality oversight combined with real-time dashboards and reporting help to ensure program success. While these results speak volumes on the value of the model, we are continuously exploring additional opportunities to further refine the program so that it meets the needs of both payers and patients. Please be in touch with us!


Subscribe To Our Blog