Risk Stratification in Practice

Published Online:April 25, 2014
Takaji Kittaka Jr, MD, is the chief transformational officer at Adena Health System. Adena sits in a unique position, geographically based in an impoverished section of rural southeastern Ohio that has exceptionally high diabetes rates, approaching 30% in some counties. Dr Kittaka has coordinated the integration of 3 hospitals, 2800 employees, and disparate physicians, while helping to create nurse navigator positions and population managers. Uniting all of these specialties in the common goal of achieving greater alignment was the topic of Dr Kittaka’s presentation at the National Association of Managed Care Physicians’ Spring Managed Care Forum 2014 in Orlando, entitled “Risk Stratification: The Practical Implementation of a Powerful Tool.” Dr. Kittaka went on to describe the many successes of this integration.
Their analysis of patients from the 3 hospitals in the region yielded a risk stratification of approximately 5% in the “high risk” category, with this population accounting for approximately 25% of the total cost of care. Other patients were categorized into either the “rising risk” or “low risk” categories. Most significantly, a graphical analysis with margin on the Y-axis revealed margin erosion if no care management strategy was enacted. An effective care management strategy would result in a positive margin slope, indicative of a financially sustainable business model.
Initially, Dr Kittaka sought something that resembled the software application Turbo Tax. Surprisingly, he said, it was nowhere to be found. So they began to develop a simple, user-friendly, open-box platform. It is an ongoing process that requires continuous interaction with physicians and nurses on the front lines, and it was developed starting with the Charleston Comorbidity Index in an attempt to properly assessing risk.
They had to add diabetes-specific risk stratification along with diabetes care guidelines that would be widely agreed-upon. They would also give consideration to evidence-based guidelines, presenting this as well to individual physicians for consideration. In a pilot program, nurse navigator and population health manager positions were created to identify high risk populations and to educate patients.
Dr Kittaka presented a case study of a patient who benefited from this integration. The individual had normal cholesterol and renal function, but their glycated hemoglobin (HbA1C) was very high (13.1). The 68-year-old woman who worked at Wal-Mart agreed to see a doctor, who asked her how she felt. She responded, “Okay, I guess.” The patient began glucophage treatment, and this dialogue was repeated several times; but her HbA1C levels were not going down. Enter the nurse navigator, who determined that the patient did not have a regular work schedule, sometimes working extremely early mornings in the bakery before getting shifted to later hours. So the nurse navigator, together with the patient, talked to the manager at Wal-Mart. After getting more regular hours and thus a more consistent sleep schedule, the patient was able to get her HbA1C levels to below 7—a stunning achievement. The nurse navigator was able to identify the patient as someone who could benefit from some patient engagement and ultimately, measurable success was realized.
Still today, Dr Kittaka wishes there was a single comprehensive platform, but we’re not quite there yet, he lamented. Nonetheless, the population managers help to sound the alert in identifying patients that are most likely to benefit from greater attention.