EHRs effective at flagging undiagnosed diabetes, new disease risk factors
Electronic health records are better at flagging patients who may have undiagnosed Type 2 diabetes than conventional screening--and at identifying several previously unknown risk factors for the illness, according to a new study in the Journal of Biomedical Informatics.
One-fourth of Type 2 diabetes mellitus patients in the U.S. are undiagnosed due to inadequate screening. The researchers, from the University of California Los Angeles, developed a screening algorithm using EHR data from 9,948 patients; they wanted to see if such screening could be more accurate and less expensive than current lab testing and fasting methods.
They compared a full EHR model with commonly prescribed medications, diagnoses and conventional predictors; a restricted EHR model that excluded medications; and a conventional non-EHR model with the generally recognized predictors of diabetes, such as age and body mass index.
The researchers found that the two EHR models were better able to increase the number of correct diagnoses by refining the pool of patients that should be screened. For instance, the tool was 2.5 percent better at identifying people with diabetes, and 14 percent better at identifying patients who did not have diabetes than the standard, non-EHR model.
The algorithm also uncovered several previously unknown risk factors for diabetes, such as intestinal infections, certain illnesses (like Herpes zoster and chlamydia) and sexual and gender identity disorders. For example, a history of gender identity disorders increased the risk of diabetes by about 130 percent--the same as high blood pressure, a leading well known risk factor. The algorithm also found some factors related to lower risk of diabetes, such as being prone to migraines.
"EHR phenotyping resulted in markedly superior detection of DM2, even in the face of missing and unsystematically recorded data," the researchers said. "EHR phenotypes could more efficiently identify which patients do require, and don't require, further laboratory screening. When applied to the current number of undiagnosed individuals in the United States, we predict that incorporating EHR phenotype screening would identify an additional 400,000 patients with active, untreated diabetes compared to the conventional pre-screening models."
EHRs are proving to be a major resource in identifying patients at risk for certain conditions. They have been found to improve diabetes treatment and have identified new subgroups of Type 2 diabetes.
EHRs, genomic data help ID new subgroups of Type 2 diabetes
EHRs, portals have positive effect on diabetes care
EHRs can help improve patient medication adherence
Electronic 'trigger' can identify cancer red flags, reduce diagnostic delays