The Promise of Health IT to Reduce Adverse Drug Events
Communication errors between providers are common in the care of older patients. Often a covering physician is not familiar with the patient’s history and drug-drug interaction problems can arise. For several years, researchers have been studying how health information technology could reduce the number of preventable adverse drug events in long-term care settings and are now are studying health IT interventions in transitions of care.
In an August 19 webinar sponsored by the Federal Agency for Healthcare Research and Quality, researchers from the Meyers Primary Care Institute in in Worcester, Mass., described some of their research on how CPOE with clinical decision support might cut down on preventable adverse drug events.
Jerry Gurwitz, M.D., executive director of the Meyers Primary Care Institute, noted that in a study of two large long-term care facilities during a nine-month period in 2000-2001, his research team found there were more than 800 adverse drug events, of which more than 40 percent were judged preventable.
"Extrapolated to the total U.S. nursing home population of 1.6 million at that time, that would be 1.9 million adverse drug events per year, 40 percent of which could be preventable,” he said. The researchers found that preventable adverse drug events occurred most often during the ordering and monitoring stages of care, he added.
Gurwitz led an AHRQ-funded intervention involving CPOE with clinical decision support to warn about drug interactions, yet the study found no statistically significant impact of the intervention. Why?
One reason, researchers believed, was the lack of specificity of alerts led to “alert fatigue,” and physicians tended to just click through them. They saw a need to increase the scope of the system to address a broader range of adverse drug events.
A follow-up research project in the long-term care setting focused specifically on renal insufficiency, which is recognized as having a substantial rate of inappropriate dosing, said Terry Field, an epidemiologist who serves as associate director of the Meyers Primary Care Institute.
With alerts on how often to order serum creatinine level tests and recommendations for dosing and for drugs to avoid, this second study found that providers with clinical decision support gave appropriate orders 63 percent of the time compared with 52 percent in the control group.
This study suggests that clinical decision support can improve the quality of prescribing decisions, Field says, in cases where providers recognize the difficulty of prescribing accurately, detailed information is needed, and the alerts are highly specific and always relevant.
The Meyers research team is now working on a study of ambulatory medication reconciliation following hospital discharge. They are looking at the use of health IT to improve transitions of patients from hospitals and skilled nursing facilities to home. Interventions include automating notification to primary care physicians, reminders to schedule visits, lists of any new medications, and alerts with recommendations about dosing issues and lab monitoring. Will that intervention have an impact on adverse drug events in transitions of care? “Stay tuned,” Fields said.
