Death by spreadsheet. Those three words pretty well sum up the plight of health plan managers and care coordinators today. With multiple information technology systems spewing out a vast array of spreadsheets, it’s a herculean task to pull all those data together so that populations can be analyzed, segmented, and targeted appropriately.
Candace Saldarini, MD
Monica E. Oss
The key is integrating data from disparate sources—including physical, behavioral, and social determinants of health—and using it to gain insights and drive action. That means segmenting the population into clinically meaningful groups so that health plans can understand the health status, needs, and associated costs of each segment.
For example, a health plan might identify four segments, ranging from most healthy to least healthy. For the healthiest group, the plan might focus on providing reminders of immunizations, screenings, and other preventive services. The next tier—individuals with a chronic condition like depression or asthma that is well managed—should be regularly monitored for early signs of symptom exacerbation. After that, there are those with a chronic condition that isn’t well managed—say, people with asthma who don’t use an inhaler, resulting in frequent emergency room visits.
The last group are people with multiple, poorly managed comorbid conditions who are at highest risk for unplanned emergency care and hospitalization. Managing so-called superutilizers who have multiple chronic conditions and complex support needs is a particularly vexing challenge. They are the 5% of Americans who consume roughly half of all health care resources.
As we shift to a value-driven population health model, the incentives to maximize volume that were built into the traditional fee-for-service environment will disappear and be replaced with a focus on consumer health outcomes and preventing acute episodes of care. People with multiple chronic conditions and disproportionate utilization volume and costs will require much closer management. Within this population, mental health conditions are a key driver of costs. More than 80% of Medicaid superutilizers have a comorbid mental illness. Consumers with behavioral health disorders and comorbid medical conditions, such as diabetes and asthma, incur higher average costs. Indeed, people diagnosed with a comorbid behavioral health and chronic physical health condition cost approximately 300% more than those having only a chronic physical health condition.
Health plan executives, clinicians, patient advocates, and policymakers have long pursued the holy grail of integrated behavioral health and physical health care. Fortunately, the use of emerging enabling technology and data analytics is putting this previously elusive goal within reach. Enhanced analytic tools, which are designed to assess behavioral health risk, can be used to address the effects of behavioral health conditions on individual health outcomes and spending. By leveraging advanced analytics, one Medicaid health plan has been able to integrate multiple consumer data sets—including medical, pharmacy, and, importantly, behavioral—to segment their population. A risk-scoring algorithm identifies high-risk members. The analytic platform builds profiles of each member segment and identifies members where care coordination support can positively affect outcomes and costs.
Armed with up-to-date information about a member’s medical and behavioral diagnoses, prescription medications, physician office visits, emergency room visits, and hospitalizations, care managers can quickly assess clinical complexity and monitor members climbing the cost and risk curve.
In the long run, the power of integrated data and applied analytics promotes better care coordination and enhances members’ ability to manage their own care. It is a plus for payers, health plans, and consumers alike.