Sometimes it’s impossible to know what the patient was or wasn’t taking, but it doesn’t mean you don’t do your best.
–Jeffrey Schnipper, MD, MPH, FHM
What is the best possible medication history? How is it done? Who should do it? When should it be done during a patient’s journey in and out of the hospital? What medication discrepancies—and potential adverse drug events—are most likely?
Those are questions veteran hospitalist Jason Stein, MD, tried to answer during an HM13 breakout session on medication reconciliation at the Gaylord National Resort and Conference Center in National Harbor, Md.
“How do you know as the discharging provider if the medication list you’re looking at is gold or garbage?” said Dr. Stein, associate director for quality improvement (QI) at Emory University in Atlanta and a mentor for SHM’s Multi-Center Medication Reconciliation Quality Improvement Study (MARQUIS) quality-research initiative.
The concept of the “best possible medication history” (BPMH) originated with patient-safety expert Edward Etchells, MD, MSc, at Sunnybrook Health Sciences Centre in Toronto. The concept is outlined on a pocket reminder card for MARQUIS participants, explained co-presenter and principal investigator Jeffrey Schnipper MD, MPH, FHM, a hospitalist at Brigham & Women’s Hospital in Boston.
“Sometimes it’s impossible to know what the patient was or wasn’t taking, but it doesn’t mean you don’t do your best,” he said, adding that hospitalists should attempt to get at least one reliable, corroborating source of information for a patient’s medical history.
Sometimes it is necessary to speak to family members or the community pharmacy, Dr. Schnipper said, because many patients can’t remember all of the drugs they are taking. Trying to do medication reconciliation at the time of discharge when BPMH has not been done can lead to more work for the provider, medication errors, or rehospitalizations. Ideally, knowledge of what the patient was taking before admission, as well as the patient’s health literacy and adherence history, should be gathered and documented once, early, and well during the hospitalization by a trained provider, according to Dr. Schnipper.
An SHM survey, however, showed 50% to 70% percent of front-line providers have never received BPMH training, and 60% say they are not given the time.
“Not knowing means a diligent provider would need to take a BPMH at discharge, which is a waste,” Dr. Stein said. It would be nice to tell from the electronic health record whether a true BPMH had been taken for every hospitalized patient—or at least every high-risk patient—but this goal is not well-supported by current information technology, MARQUIS investigators said they have learned.
The MARQUIS program was launched in 2011 with a grant from the federal Agency for Healthcare Research and Quality. It began with a thorough review of the literature on medication reconciliation and the development of a toolkit of best practices. In 2012, six pilot sites were offered a menu of 11 MARQUIS medication-reconciliation interventions to choose from and help in implementing them from an SHM mentor, with expertise in both QI and medication safety.
Participating sites have mobilized high-level hospital leadership and utilize a local champion, usually a hospitalist, tools for assessing high-risk patients, medication-reconciliation assistants or counselors, and pharmacist involvement. Different sites have employed different professional staff to take medication histories.
Dr. Schnipper said he expects another round of MARQUIS-mentored implementation, probably in 2014, after data from the first round have been analyzed. The program is tracking such outcomes as the number of potentially harmful, unintentional medication discrepancies per patient at participating sites.
The MARQUIS toolkit is available on the SHM website at www.hospitalmedicine.org/marquis.