Article
Author(s):
A recent study shows electronic health records make it easier for health care professionals to report adverse drug events.
Among the millions of ripple effects expected as the nation’s health care system shifts to electronic health records (EHRs), one unexpected result may be safer medications.
According to a study published Monday in the journal Pharmacoepidemiology and Drug Safety, EHRs have the potential to simplify the normally complicated path health care practitioners must follow to report adverse drug events.
By creating an automated reporting tool linked to patients’ EHRs, researchers found they could “greatly increase the number and quality of spontaneous reports submitted to the FDA.” Improvements in the quantity and quality of drug safety information available to the FDA could help the agency refine its regulatory processes.
In a small experimental group of 26 physicians at Massachusetts General Hospital and Brigham and Women’s Hospital in Boston, the tool caused the number of adverse event reports to soar—from nearly 0 in the previous year to a total of 217 reports during the 5-month study period. Of the 217 reports, 23% were defined by clinicians as serious.
The tool shaved time off the cumbersome reporting process by providing clinicians with a pre-filled electronic form. Using the patient’s EHR, the form was automatically populated with data that could be used to classify an adverse drug event, such as comorbid conditions, concurrent medications, weight, and laboratory tests. Once received, clinicians could complete and send the report in less than a minute.
According to a report by the Wall Street Journal, lead study author Jeffrey Linder, MD, of Brigham and Women’s Hospital, said the tool “might need to be tweaked to reduce reporting of nonserious effects.” The system is scalable, however, and could be easily implemented nationwide to improve reporting.
Currently, US physicians report fewer than 1% of adverse drug events to the FDA, the study’s authors noted.
For other articles in this issue, see: