Notwithstanding the focus on COVID-19, chronic disease is the leading burden of our healthcare system.
Six in ten American adults have a chronic disease such as type 2 diabetes, asthma and heart disease. Four in ten have two or more. Seventy percent of deaths are caused by chronic diseases and associated complications. Three quarters of our healthcare spending is on chronic diseases and associated complications.
The Geneia Data Intelligence Lab
That’s why the Geneia Data Intelligence Lab (GDI Lab) has prioritized analytic models that:
- Predict the onset of chronic disease for those who don’t have it yet
- Predict complications of chronic disease for those who have a diagnosis
At its heart, the GDI Lab is comprised of innately curious PhD- and masters-level data scientists who use leading-edge data science to drive lower healthcare costs and improve health outcomes. Our lab prioritizes projects that address major cost drivers and enable our clients – health plans, hospitals, physicians and their value-based partners – to identify, stratify and predict high-cost patients and conditions such as heart failure and diabetes. Our models help clients to intervene earlier with patients whose risk is expected to rise and/or who are at risk for major, expensive conditions, allowing healthcare organizations to mitigate future costs while improving the health of those they serve.
Deep clinical experience enhances analytics
Geneia is fortunate to have population health experts with deep clinical experience on our team. As such, we acutely understand the costs and complications of chronic disease, leading the GDI Lab to prioritize chronic illness models. Clinical practicality is always front and center as our GDI Lab creates, tests and perfects data models.
Predicting type 2 diabetes-related complications
As many know, diabetes statistics are staggering.
- More than one in 10 adults has diabetes.
- Approximately one in three have prediabetes, and 90 percent don’t know it.
- Total estimated costs associated with diabetes in 2017 were $327 billion.
- In 2012, complications accounted for 43 percent of the money spent on diabetes care.
Diabetic complications include cerebrovascular disease, heart disease, eye damage and more, all of which have serious implications for the affected patient.
Type 2 diabetes complication model
Predicts for those with type 2 diabetes who is likely to experience a serious diabetes-related complication in the next 12 months.
The model yields binary classification with probability and risk buckets (dictated by probability) for the next year.
There are a number of important variables used in this model, including:
- Charlson Comorbidity Index (CCI)
- Prescriptions for managing chronic disease(s)
Given the number of complication events per person per year, a health plan that uses this model to target the top 100 riskiest members can expect annual savings of approximately $500,000, and those members can potentially avoid unnecessary hospitalizations.
Predicting complications of heart failure
Heart failure is defined as the inability of the heart to pump enough blood to meet the body’s demand. Approximately two percent of adults have heart failure. That statistic jumps to six to 10 percent for those 65 years or older for whom heart failure is a leading cause for hospitalization. Historically it’s been one of the top 10 most expensive inpatient conditions. It’s also a leading cause of death, with 35 percent dying in the first year after diagnosis. However, long term prognosis improves after that first year, making early interventions and management critical.
Early signs and symptoms of heart failure complications are numerous, including dizziness, shortness of breath, acute and chronic pulmonary edema, pleural effusion and respiratory failure. The ability to forecast these preventable exacerbations before the disease progresses to more severe complications is an important challenge the GDI Lab explored.
Heart failure complications model
Predicts for those with heart failure who is likely to experience a preventable exacerbation in the next 12 months.
The model output is binary classification with probability and risk buckets for five future time windows:
- 0-3 months
- 3-6 months
- 6-9 months
- 9-12 months
- 0-12 months
Among the important variables in this model are:
- Heart failure visits
- Inpatient costs
- Outpatient costs
- Longer time with heart failure diagnosis
Variations of this model predicting likelihood of future heart failure-related hospitalizations are currently being utilized to identify people who would be good candidates to target for enrollment in a remote monitoring program.
Predicting hypertension-related complications
Hypertension is the leading preventable risk factor for global morbidity and mortality. This is due to:
- Its high prevalence, affecting nearly one in three adults and nearly 60 percent of people ages 45-64, and
- Because when not well-managed, can progress through a range of stages, ultimately contributing to devastating complications, including stroke, end-stage renal disease, and heart attack.
Because one of the aims of care management is to intervene before chronic illness complications can progress that far, the GDI Lab designed models that can predict various stages of hypertension complication severity, including earlier signs and symptoms such as proteinuria and retinopathy.
Hypertension complications model
Predicts for those with hypertension who is likely to experience a complication from three complication stages in the next 12 months.
The model yields binary classification with probability and risk buckets for each of the three complication stages.
Some of the important variables used in this model are:
- Total medical costs
- Outpatient visits
- Emergency department visits
I expect we’ll soon be discussing chronic disease complication models in conjunction with COVID-19. Early research suggests people with comorbidities like diabetes, hypertension and heart disease are at higher risk for COVID-19 complications and hospitalizations. Companies like Geneia are creating analytics models to determine the population of patients at highest risk. Physician practices, hospitals and health plans can then use this kind of information to triage and prioritize patient outreach and engagement.
To learn more about the Geneia Data Intelligence Lab, click here.