Why various data feeds are needed to improve diabetes management
Optimizing treatment of this condition must factor in and find relationships between various types of information about the person.
Diabetes is one of the few chronic conditions that can be very effectively managed by an individual. So why are there so many people with poorly managed diabetes?
Most people aren’t willfully noncompliant with their care. Instead, many have extenuating socioeconomic or other circumstances that can lead to drifting away from their care plans.
Insights that surface actionable recommendations and explanations by bridging clinical, financial, social determinants of health (SDoH), and buying behavior data with patients’ self-reported information are essential elements of chronic condition management. Aggregating these disparate data sources and applying advanced analytics to deliver insights to patients and their care teams can positively transform the health and wellness of millions of people.
Technology’s potential role
Technology-assisted disease management makes the patient experience less arduous. Devices and digital tools such as continuous glucose monitors (CGMs), insulin pumps, chatbots and smartphones have not only made the experience easier for patients in recent years, but they have also generated an enormous amount of new and unique data. This information can be used to derive valuable insights into both the individual and population levels. Visibility into metabolic trends is just the first step to successfully managing a person’s chronic condition.
These new data sets must be appropriately extracted, aggregated and analyzed in conjunction with clinical, financial, socioeconomic and buying behavior data to develop personalized risk-based patient profiles, which are crucial for effectively predicting outcomes and guiding interventions.
Finally, these insights must then be embedded into automated, precise clinical workflows to drive action. Providing a view into patient activity that isn’t just clinical but also consumption-focused (like supplies ordering as one important example) provides a platform to take intervening action.
Arming physicians and an extended care team with these data-driven insights to optimize outcomes is essential for building and maintaining trusted patient-provider relationships, maximizing resources and offering individualized, unique patient-centered care to people living with diabetes.
Data as an enabler for care
Signals on the state of a patient can come from an array of different sources, including admission, discharge and transfer (ADT) feeds; CGM feeds; timely claims; supplier or pharmacy data; and social determinants data.
By integrating these and other data sources into a single population health management platform, care teams can better identify rising risk factors, predict population trends or individuals’ actions based on clinical and non-clinical factors in their lives, and intervene appropriately before these issues become full-blown, costly crisis-level events.
This profile-based strategy takes traditional patient care to the next level by enabling care teams to efficiently deploy personalized interventions in a proactive and holistic manner. As individuals respond to interventions and potentially move between higher- or lower-risk categories, it creates a feedback loop in which care teams can learn more about how people navigate their condition, iterate on outreach plans, and start predicting future behavior and outcomes more accurately.
The need to aggregate data sources
Because diabetes is a round-the-clock condition that requires constant management, it produces multiple streams that can be used to develop and enhance care.
For example, CGM device data is produced constantly and in real time, making it incredibly valuable for understanding how people manage their condition day to day, and respond to alerts on approaching highs and lows.
Less frequent but no less valuable are EHR and claims data, which can provide a broad view of a patient’s disease progression and resource utilization, as well as ADT data feeds that can alert on emergency department admissions, and social determinants data that may indicate barriers to care access.
Individually, these data streams provide significant standalone value. But when combined and supplemented with additional quantitative and qualitative patient-reported data, they become even more powerful by creating comprehensive patient visibility.
But care teams can go even further by leveraging novel data elements, such as product orders or claims, to better understand adherence to care plans and provide triggers for interventions.
For example, diabetes supply ordering patterns can be an early indicator that there’s a problem with a person’s self-care routines. Someone who stops ordering supplies on a regular basis may be struggling with costs, may not have a stable home in which to receive deliveries or has fallen away from their recommended care plan for other reasons. These types of data, which are traditionally only used as a business metric for healthcare supply chain companies, could have significant value for clinicians if successfully integrated with other data sets into one population health platform.
Trusted relationships and analytics
Even the most sophisticated data-driven intervention runs the risk of failure if it is not backed by an established, trusted bond between patients and a provider that is well positioned to deliver personalized interventions. Ideal care team members can include physicians, care managers, payers and even supply distributors with whom patients interact regularly.
These team members must be willing and ableto go beyond the “what” and dig down to the “why” to understand a patient’s preferences or actions. Again, most people aren’t willfully noncompliant with their care. Instead, many have extenuating socioeconomic or other circumstances that can lead to falling away from their care plans.
To understand and address these challenges, care teams need access to meaningful, actionable recommendations and explanations built on historical patterns of healthcare engagement and detailed SDoH data.
Caring for people with diabetes truly is a team effort. Patients must be engaged and proactive in their care, but they also need the support of a surrounding ecosystem of physicians, nurses, care managers, and suppliers. Access to the right data is critical to the team’s success. Data must be integrated into intelligent platforms that can extract insights and automate workflow processes as much as possible to drive precise and relevant action.
By merging rich, non-traditional data sources such as purchase trends with foundational elements like claims and clinical services, trusted care team members can develop one-of-a-kind insights into individuals’ risks and behaviors. Translating broad, extensive multiple data sets into actionable information specific to an individual holds the potential to better manage populations while simultaneously changing the trajectory for each patient living with a chronic disease.
Richard Mackey is chief technology officer at CCS and Jean-Claude Saghbini is chief technology officer at Lumeris