Computerized Data Mining Yields Better Approach to Diabetes Screening and Diagnosis

Researchers from UCLA have developed an algorithm that they report has the potential to greatly refine the process of screening and diagnosis for type 2 diabetes. Additionally, their work has revealed a range of previously unknown risk factors for the condition, including a history of sexual and gender identity disorders, intestinal infections, and sexually transmitted diseases such as chlamydia. The discoveries are based on a computerized analysis of large quantities of medical records, a technique called big data mining. Mark Cohen, a professor in residence at Semel Institute for Neuroscience and Behavior commented “The overall message is that ordinary record keeping that doctors do is a very, very rich source of information. If you use a computerized approach to studying patterns in that data, you can greatly improve diagnosis and medical care.” The findings appear this week in the Journal of Biomedical Informatics.

About 1 in 4 people with diabetes are unaware that they have the disease that can engender a wide range of complications including painful neuropathies. Diabetes screening has typically been based on a limited set of factors that include blood pressure, BMI, age, gender, and smoking habits. But the UCLA team calculates that medical record analysis using the new algorithm could identify up to 400,000 people who have not yet been diagnosed with diabetes. “There’s so much more information available in the medical record that could be used to determine whether a patient needs to be screened, and this information isn’t currently being used,” said Cohen. Read more about the findings here. The journal abstract may be read here.


Related Content