Like many terms, Population Health means different things to different people. For example, in the context of the CDC it means contagious diseases that can impact a broad population, while in a medical trial it may refer to a specific cohort being studied. For the context of this article, however, we will focus on a broader definition. Population Health is the consolidation of individual patient records such as Electronic Health or Medical Records (EHR or EMR), Encounters and Interactions, and Vitals into an aggregated data store that can be used to visualize and analyze the entire population in order to gain insights. These insights may be correlations, propensities, and potential causalities related to common patient markers. From this “macro” view of the population overall health outcomes can be studies within a population to ascertain personal, social, economic, care protocols, and environmental factors that influence the health outcomes. The value of Population Health is derived through the analysis of “like” individuals, and the targeted analysis related to specific health outcomes based on all the information known about the individuals within the population.
While Population Health is often tracked through clinical research, there is a wealth of data available on nearly every individual that can provide tremendous insights to effective and ineffective care.
Most care today is based on the “Episodic” model. The Episodic model focuses on a specific health episode, and then attempts to address the health outcome through intervention. While chronic conditions are less episodic in nature in that they continue over time, they are still primarily based on response to some adverse medical event, albeit a long-running response.
The goals of Population Health are two-fold:
Larger populations with greater data sources lead to greater statistical relevance. For example, enriching existing data sources to include EHR, claims, appointments, vitals, and family history allows for a more complete picture. Further augmenting this with vitals data obtained from Remote Patient Monitoring (RPM) and Medical IoT’s (Internet of Things devices), the information can not only extend in breadth but in frequency. With this timely and extensive view of the individual and the overall population, a number of additional benefits can be derived including threshold-based alerts, reminders, and even machine learning algorithms that predict the likelihood of adverse events proactively. The population view also allows for unique tools that allow for rapid experiments in the care protocol to validate changes applied to a study group to statistically validate effectiveness prior to a broader adoption. Believe that a change in a meal plan will reduce Type II diabetes? Run a rapid, low cost experiment against a study group to prove the plan works. Think a dog walker will reduce falls? Test the hypothesis to understand the cost/benefit. Since the test and control populations are contained within the population health platform, experiment results can be monitored in near real time.
With the cost of adverse events running in the tens of thousands of dollars, and having long-term adverse effects, it is easy to see why we must shift from episodic care to a proactive model, and how Population Health is a critical component in this change.