A recent survey from Health Research & Education Trust, Public Health Institute, and Association for Community Health Improvement uncovered that hospitals define populations for the purposes of health management strategies based on several criteria (listed below). However, the results show that most hospitals have taken a very broad approach to identifying at-risk populations, which could lead to valuable resources and time being spent in an inefficient manner.
- Users of a hospital or health system 70%
- Residents of specified geographic area or community 69%
- Individuals experiencing a certain disease or condition 59%
- People for whom the hospital has financial risk 47%
- Other 11%
Successful population health initiatives require an organization to drill down to the finest points of their patient populations to identify strategic areas for improvements. While these broad risk categories (as shown in the survey results) may provide a foundational understanding of where risk is located, they are not specific enough to impact why these populations are experiencing sub-optimal health. Resources and interventions must be allocated in a patient-specific fashion to insure relevance, adoption and adherence.
In my past life as the Integrated Care Manager at non-profit, safety-net, community hospital, population-focused improvements were a large part of the organization’s Delivery System Transformation Initiatives. As much as improving the access and delivery of quality care to all hospital users was the ultimate goal, we needed to identify where our limited resources would best be utilized to make the greatest impact.
As a hospital whose patient population was predominately low-income, uninsured, and had multiple health, social and behavioral concerns, it was necessary to identify a core population that would benefit most in our long-term goal of continued patient engagement and care coordination. This was not easy. When using basic demographics (age, gender) and clinical information (diagnosis history, utilization patterns), many patients appeared to “look” the same. It was very difficult to develop patient-specific intervention strategies for a group of, for example, high-utilizing 35-55 year old men with diabetes. While all of these patients may be high-risk, each has specific needs and factors that are contributing to their overall health.
As one of the leaders of a multi-million dollar initiative to transform care delivery, this was an eye-opening challenge. How do we extract the highest risk patients for readmissions or post-discharge failure from a group that looks medically similar? As a social worker, I knew that a patient’s barriers to achieving and maintaining health had much to do with their social and behavioral wellness, which differs from patient-to-patient in these medical cohorts. Patients must have the right resources such as transportation, food outlets and housing to adhere to their individualized treatment plans.
At Connance, I have the opportunity to implement predictive analytics that provide patient-specific insight on socio-demographic and behavioral factors. This “Whole Patient Insight” can truly drive an organization’s efforts towards defining increased high-risk populations, and bringing to the forefront the best interventions in preventing readmissions and unnecessary utilization.
The ability to approach population health initiatives requires a holistic view of individual patients, the context of their community environment and the barriers that prevent access to resources. Continued, “siloed” use of claims and clinical data without socio-demographic and behavioral insight to define targeted patient populations will only provide a superficial view of those in need of strategic interventions.