Use data to promote a holistic approach | Behavioral Healthcare Executive Skip to content Skip to navigation

Use data to promote a holistic approach

April 1, 2009
by Ray Dunn
| Reprints
Organizations that effectively use data will help consumers live healthier

Population health management (PHM) is the concept of predicting, maintaining, and coordinating the health of an individual within an at-risk population. An effective PHM program strives to achieve the optimum treatment cost and clinical outcomes for a population. Within the behavioral healthcare community, PHM is not necessarily just about mental health or substance abuse issues but includes their relationship to physical health conditions in order to treat the whole person.

A clear connection exists between physical and behavioral health, as chronic conditions such as obesity, diabetes, and cancer often co-occur with mental health issues. According to the National Cancer Institute, about 15 to 25% of cancer patients develop depression and anxiety. Depression can lead to a much lower quality of life and trigger substance abuse, other mental health problems, or behaviors that can worsen one's physical condition and prevent a positive outcome to medical treatment. Another example: A study in Pediatrics found that the longer a child is overweight, the more he/she is at risk for depression and other mental health disorders.1

Yet many U.S. disease management programs have not been successful because (1) they are focused only on a condition and not on a behavior that caused or is exacerbating the condition, and (2) they are not doing much to coordinate behavioral healthcare and primary care. Additionally, few disease management organizations proactively examine behaviors that may develop due to an existing condition. But when behaviors are separated from physical conditions, PHM's premise falters as physical and behavioral conditions worsen and the person is not treated as a whole.

To make PHM's ideal of whole-person care reality, organizations must develop a practical, yet proactive, care-coordination approach. This must focus on the individual's behavior and be coupled with a fully integrated care management program that sits atop a foundation of data analytics and business intelligence. This foundation consists of four basic steps: (1) identify, (2) engage or notify, (3) track, and (4) measure. Each step is equally important and requires its own unique business intelligence and data analytics.

Types of business intelligence

Each of the four steps relies on data analytics and business intelligence that must be strategic, analytical, and operational.

Strategic business intelligence is used to drive the performance of an organization as a whole, as well as the individual departments that execute PHM services. While strategic intelligence sets the foundation in the form of key performance metrics, analytical intelligence identifies the source of an issue once it has been uncovered. Tools such as dashboards, online analytical processing (OLAP), predictive analytics, and improvised (ad hoc) queries are used to determine the location or cause of an issue or to identify opportunities for strategic adjustment. Rounding out the cycle, the results obtained from analytical intelligence are used to drive operational initiatives. Operational business intelligence facilitates the kind of day-to-day decision making that happens at an organization's lower levels and enables the attainment of strategic goals.2 All three business intelligence levels must work together and within each step of a successful PHM program.

Step one: Identifying

The first step in establishing a PHM program is to identify the high-risk members (i.e., clients, consumers, etc.) within an organization or population. Approximately 10% of members account for more than 35% of a PHM organization's cost.3 These “toughest 10%” typically have avoidable or inappropriate service utilization because of behavioral or social conditions or issues that aggravate high-expense medical conditions.

Analytics employed in this step must go further than merely identifying the population based on current behavioral patterns, conditions, or factors. They should extend into a predictive model that identifies populations that may need attention in the future.

Step two: Engaging

The second-and most difficult-step is engaging members of these at-risk populations. People typically are difficult to motivate and hard to engage, so this step requires a multitiered approach. Successfully accomplishing this step is based not only in constant, repetitious communication strategies but on advanced technology platforms that produce predictive metrics and indicators that accurately show what these members are experiencing.4

Step three: Tracking

The organization must be prepared to deliver strong interventions and treatments for the engaged members and have the ability to track their progress and coordinate care. This step presents special challenges. If interventions can coherently address the variety of issues that affect health status and ultimately service utilization, they will represent a substantial opportunity for improving clinical outcomes and lowering the cost of care.