How COVID-19 Data Science Models Can Be Used to Reduce Care Costs for High-Risk Diabetics

In Care Coordination, Care Management, Covid-19, Thought Leadership by Marketing

Studies continue to confirm that diabetes is a serious risk factor among patients contracting COVID-19, contributing significantly to mortality and length of hospital stays. Because glucose levels begin to rise with the onset of symptoms and continue to elevate during infection, early COVID-19 diagnosis and treatment is critical to managing insulin protocols. As discussed in our last post, the virus’s 14-day symptom arc requires the ability to consume and make data actionable in as close to real-time as possible. Case report data lags, as new cases are not entered into databases until after the incubation period occurs and the patient seeks testing or treatment as noted by the CDC.

Data Science Enables Rapid Use of New Data Types To Combat COVID-19 Spread

To overcome this data lag, care managers must be able to rapidly turn information about COVID-19 case counts, geographic locations and more into insights.  This is where data science, Medical Directors and case managers can work together and utilize custom-developed COVID-19 data models to ingest the CDC’s case count data by geography and combine this data with internal claims diagnosis codes into a real-time prioritized care queue of risk-stratified members grouped by disease, age or other criteria. This approach enables health plan leaders to consider a wide variety of variables and become more proactive and answer questions like:

  • Which geographies have the highest COVID-19 case count and the most members over 65 with high-risk pre-existing conditions (Diabetes, asthma, COPD, etc.)?
  • Which family members are at risk for contracting the virus?
  • Who are the most vulnerable within our plan’s population based on SDoH criterion?

Using COVID-19 Diagnosis Data With Diabetic Members Improves Care & Reduces Cost

Case Managers focused on members with diabetes, for instance, will be able to connect with members at-risk based on whether or not the member has Type 1 or Type 2 diabetes, rather than just “diabetes.” Cross referencing these risk factors with member characteristics such as age, diabetes type, address, other underlying health conditions, and more allows for more tactical approaches to education and intervention. For instance, studies have indicated that patients with type 2 diabetes may be at greater risk from COVID-19 than type 1 patients.

This insight can ensure the care managers are instrumental in improving care quality by intervening in a timely manner with prevention education, reminders to use perform at-home insulin checks, attending primary care visits, and more. 

Leverage these COVID-19 data science algorithms to capture, incorporate and analyze external data from the CDC and other sources.  Let us know how we can help below.

Keep on Learning

Laura Barnett, BSN, RN, CDE
Vice President Client Partnerships

Laura is the Vice President of Client Partnerships at Vital Data Technology and product owner of Affinitē PlanLink. In this role, Laura oversees all aspects of the customer life-cycle, serving as an ambassador for all clients and partners to ensure a world-class customer experience. With her career spanning over 20 years in healthcare ranging from nursing leadership, medical device sales, and healthcare information technology account development, and partner management, her clinically-founded expertise ensures her astute alignment with health plan goals.

Laura is nearing completion of her Masters of Health System Information Management from Texas Women’s University, and she holds her BSN, Nursing from The University of Texas Health Science Center San Antonio.

Share on