EHR comorbidity information lacking in accuracy

Electronic health record (EHR) problem lists do not accurately portray a patient's comorbidities, according to a study published in The American Journal of Managed Care.

Patients comorbidity information within EHRs is crucial in providing high-quality care for chronic diseases, but information regarding the accuracy of this information is lacking. In this study, researchers evaluated the EHRs of 1,596 men diagnosed with prostate cancer to identify the accuracy of comorbidity documentation.

Researchers compared EHR problem list-based comorbidity assessments with a manual review of EHR notes. Results showed EHR problem list–based comorbidity assessment had low sensitivity in identifying comorbidities. Specifically, the EHR identified only 8 percent of myocardial infarctions, 32 percent of cerebrovascular disease, 46 percent of diabetes, 42 percent of chronic obstructive pulmonary disease, 31 percent of peripheral vascular disease, 1 percent of liver disease and 23 percent of congestive heart failure.

“Inaccuracies in EHR problem list–based comorbidity data can lead to incorrect determinations of case mix,” concluded first author Timothy J. Daskivich, MD, MSHPM, and colleagues. “Such data should be validated prior to application to risk adjustment.”