This article is more than 5 years old. Information may no longer be current.
Evaluating EHRs may identify missed type 2 diabetes diagnoses
Evaluation of electronic health records may reveal health conditions and behaviors that could be useful in identifying patients with undiagnosed type 2 diabetes, according to researchers.
“With widespread implementation, these discoveries have the potential to dramatically decrease the number of undetected cases of type 2 diabetes, prevent complications from the disease and save lives,” Ariana E. Anderson, PhD, assistant research professor and statistician at the Semel Institute for Neuroscience and Human Behavior at the University of California, Los Angeles, said in a press release.
Anderson and colleagues evaluated data from electronic health records (EHRs) on 9,948 U.S. patients from 2009 to 2012 to determine whether EHR phenotyping could improve type 2 diabetes screening compared with conventional models. Three models were evaluated: full EHR model including prescribed medications, diagnosis and conventional predictors; restricted EHR model excluding medications; and a conventional model including basic predictors and interactions.
Compared with the conventional model, both EHR models provided better prediction (P < .001). The area under the curves (AUCs) were 84.9% for the full EHR model, 83.2% for the restricted EHR model and 75% for the conventional model. At a threshold set in which 18.1% of patients would be diagnosed with type 2 diabetes, both EHR models revealed better accuracy, sensitivity, positive predictive values and negative predictive values compared with the conventional model.
Age, hypertensive status and the interaction between age and hypertension were predictors of type 2 diabetes in the conventional model (P < .05).
“Our full EHR approach identified unexpected factors that were associated significantly with current [type 2 diabetes], including ICD-9 302.X (sexual and gender identity disorders), ICD-9 477.X (allergic rhinitis) and the use of depot medroxyprogesterone contraceptives,” the researchers wrote.
“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,” study researcher Mark S. Cohen, PhD, a professor in residence at the Semel Institute, said in the release. “This is a treasure trove of information that has not yet begun to be exploited to the full extent possible.” – by Amber Cox
Disclosure: The researchers report no relevant financial disclosures.