Predictive Analytics Identify High Risk Hepatitis C Patients Posted on : May 27 - 2015

Researchers from the University of Michigan have developed a predictive analytics algorithm that uses basic EHR data to flag patients at high risk of developing complications from the hepatitis C virus (HCV), according to a study published in the most recent issue of Hepatology.  As payers seek to control the costs of extremely expensive but highly effective new HCV treatments like Sovaldi, the use of clinical analytics to pinpoint the most meaningful course of treatment for patients may help to avoid unnecessary spending.

More than three million Americans are living with HCV, the Office of the National Coordinator said recently, and the virus presents a major challenge when it comes to reducing the costs of chronic disease management.  Patients with the liver-damaging virus incur five times the average annual healthcare costs, yet only half of patient suspected of having the liver-damaging virus have been diagnosed with the condition.  Just one tenth of those patients ever receive treatment, and only half of those are ever cured.

Predictive analytics for chronic disease management

Approximately one-third of HCV patients are at high risk of developing complications from the virus, says the research team, but providers aren’t always aware of who those patients may be.  “Offering immediate treatment to patients identified as high risk for poor health outcomes would allow these patients to benefit from highly effective treatments as other patients continue to be monitored and their risk assessment updated at each clinic visit,” says lead study author Monica Konerman, MD, MSc, a fellow in gastroenterology at the University of Michigan Health System.

“Ideally we would treat all patients,” she added. “Until logistic and financial barriers are solved, clinicians and policy makers are faced with trying to target these therapies to patients with the most urgent need.  The model allows us to identify these patients with greater accuracy.”

The predictive analytics model was developed using a dataset derived from the Hepatitis C Antiviral Long-term Treatment Against Cirrhosis (HALT-C) trial conducted by the National Institutes of Health.  The team used machine learning techniques to process clinical information such as lab results, age, body mass index, and details of the virus type to create a risk score for patients.  The score is more accurate than previous attempts because the algorithm uses more lab values than other models and analyzes how the values change over time.  View more