Stanford Researchers Turn To Big Data To Seek Out Patients With Rare Undiagnosed Conditions
By Lt. Dan
Researchers from Stanford University are turning to big data to help them track down patients at risk for rare but dangerous medical conditions. The project is being spearheaded by Stanford cardiologist John Knowles, MD, who is using machine learning algorithms to pour through EHR data in search of patients that might have familial hypercholesterolemia, an uncommon but potentially deadly heart condition that is inherited and can cause heart attacks at a young age.
Familial hypercholesterolemia affects one out of every 500 people in the US, but many patients go undiagnosed throughout childhood because the condition can remain largely asymptomatic for years. To further complicate matters, primary care providers are largely unfamiliar with the disease, and there are no routine screenings aimed at identifying patients with familial hypercholesterolemia. Knowles explains, “This disorder certainly leads to premature death in thousands of Americans each year … Less than 10 percent of cases are diagnosed, leaving an estimated 600,000 to 1 million people undiagnosed.”
To combat this, Knowles and a team of machine learning programmers have created an algorithm that will look through EHR data in search of patterns in the medical records that can be used to identify familial hypercholesterolemia patients in the future. Once discovered, this pattern will then be programmed into a separate search program that will allow researchers to systematically search for familial hypercholesterolemia patients at a larger scale. Additionally, the program will actively monitor any new patients and alert care providers whenever a patient presents with conditions that match the pattern. The team reports that they will publish the algorithms they create in a way that will allow hospitals across the nation to deploy them within a variety of different EHR systems. “These techniques have not been widely applied in medicine, but we believe that they offer the potential to transform health care, particularly with the increased reliance on electronic health records,” says Knowles.
The research team says that while this project is part of a larger initiative to increase early diagnosis of familial hypercholesterolemia, the technology being deployed has broad implications in other areas. There are a myriad of under-diagnosed conditions in the population that could be very effectively targeted with passive algorithms operating just under the hood of advanced EHR systems.
Article Source: http://histalkmobile.com/stanford-researchers-turn-to-big-data-to-seek-out-patients-with-rare-undiagnosed-conditions/
