Population health management (PHM) solutions represent a broad set of technologies and processes that have been designed with specific end users and applications in mind, including: data collection, management and analysis, care management and coordination, and patient engagement. Software solutions can be implemented as standalone applications or as an integrated suite that provides client organizations with a more comprehensive if not total end-to-end solution. Demand is driven by a set of business models that are increasingly defined by an organization’s exposure to risk within a value-based system of care.
Electronic health records (EHR) have provided the technological foundation to support the shift toward PHM strategies and interventions. Although EHRs are critical for realizing PHM in practice, as standalone solutions they are viewed as limited in value when it comes to the available data to inform PHM interventions and the quality of data they provide. As a result, PHM requires a much broader range of functionality than what EHRs currently provide in order to achieve targeted population health outcomes.
PHM software solutions typically sit above EHRs to enable the collection, aggregation, and transformation of data from multiple, disparate sources in near-real time. Software solutions can provide a longitudinal record for each individual within a defined population. This enables the identification and stratification of patients; the highlighting of gaps in care; the delivery of actionable decision support to clinicians to provide the right care to the appropriate patient at the right time; and the ability to support communication and coordination of services along the care continuum.
The Challenges of Handling Data
The total end-to-end solution for PHM faces a number of challenges with data collection, analysis, tracking, and monitoring. First is the knowledge of which data elements are key for the effective care management of targeted populations. Second is the standardization of data metrics and management systems at both the individual and community levels to support their integration with more traditional data sources to develop holistic care plans. Lastly is the ability to continuously update and analyze a patient’s information as a patient’s health status and their utilization of healthcare resources is dynamic.
Data sources for PHM typically consist of a mix of administrative, clinical, and operational information from systems both inside and outside the organization (i.e., claims, clinical, financial). However, PHM also needs to include additional data sources, such as patient-centered and community-based data, to be complete, and data warehouses that can accept, store, normalize, and integrate data from multiple systems are a common requirement for supporting advanced data collection, aggregation, transformation, analysis, and predictive modeling in PHM.
The quest for data quality is ongoing, as the effectiveness of any PHM software solution relies on the availability and quality of underlying data for analysis. As a result, data collection, aggregation, and analytical capabilities are far from a commoditized practice. Statistical forecasting techniques and modeling methods are commonly used approaches to assess risk and inform targeted interventions and the monitoring of care once delivered. Machine learning and natural language processing (NLP) are emerging data mining approaches.
Facilitating Patient Engagement for Improved Patient Health Management
Accountability for patient care outcomes outside of traditional care settings has increased the importance of patient activation and engagement in care. Technology solutions that facilitate patient engagement strategies include those that promote greater connectedness with care teams, such as patient portals and remote patient monitoring applications, as well as patient activation in care, such as mobile health apps, wearable monitoring devices, and personal health records. Such solutions, which can facilitate the scaling of interventions compared with in-person encounters and provide patient-generated data that, when integrated with more traditional data sources, can support improved PHM outcomes.