It’s been over a decade since Dr. Jeffrey Schnipper first realized the importance of reducing the number of unintentional medication changes that occur when patients enter and leave the hospital. Since then, he’s conducted several studies aimed at figuring out what tools and processes can help minimize these potentially harmful medication discrepancies.

Dr. Schnipper, director of clinical research for the hospitalist service at Brigham and Women’s Hospital in Boston, is the principal investigator for the Multi-Center Medication Reconciliation Quality Improvement Study (MARQUIS). The 3-year study seeks to improve the way medications are prescribed, documented, and reconciled at admission and discharge to the hospital. MARQUIS, which was launched in 2010, is sponsored by the Society of Hospital Medicine and is funded by a $1.5-million grant from the Agency for Healthcare Research and Quality.

In an interview with Hospitalist News, Dr. Schnipper explained what has been learned so far from the MARQUIS experience and what that means for medication reconciliation efforts around the country.

Hospitalist News: What is the status of MARQUIS and what are some of the preliminary results out of the six sites in the study?

Dr. Schnipper: The main outcome that we are looking at is the number of unintentional discrepancies in either admission orders or discharge orders. These are unintended differences in medication regimens across different sites of care, which are not done for medical reasons. Since we often don’t know what patients were taking when they came into the hospital, we sometimes write wrong orders for that reason. We call these "history errors." Sometimes we know what medications a patient was taking when he or she came into the hospital, but maybe we decide to hold the medication on admission for a good clinical reason and then forget to restart it at discharge. We call these types of errors "reconciliation errors."

What we’re finding is that the number of unintentional medication discrepancies for patients varies anywhere from 2.4 to 4.7 per patient, almost a twofold difference between our best performing and our worst performing hospitals across the six sites. We’re also finding, consistent with past studies, that history errors are more often the cause of the discrepancies than reconciliation errors.

HN: Is there a lot of variability in how hospitals perform when it comes to medication reconciliation?

Dr. Schnipper: In MARQUIS, we found a twofold difference between our best and worst hospitals, so I think there really is variability. Who performs each component of medication reconciliation is actually really variable. Where it’s done is also really different between our sites. It’s everywhere from pharmacy technicians doing it in the emergency department to nurses doing it on the floors to residents or attendings doing it on the floors. And in some hospitals it’s a combination. One thing that we’ve found is that at some hospitals there’s not a lot of clarity in terms of roles and responsibilities.

HN: What’s the best role for the hospitalist to improve this process?

Dr. Schnipper: I think there are a lot of people out there, including hospitalists, who feel like medication reconciliation is just a regulatory issue. It’s a box to check. It’s not really my problem. What I would say to them is that medication reconciliation is the process of making sure that, at the end of the day, your patients are being ordered the right medications.

As hospitalists, we are responsible for those orders being right. We are therefore responsible for the medication reconciliation process being done well. That does not mean we have to do the whole process ourselves and we shouldn’t, but we need to make sure that there are systems in place to do it well and we need to help our hospitals improve their systems. It’s a professionalism issue.

HN: Are there types of technology or systems that a hospital should have to perform medication reconciliation properly?

Dr. Schnipper: There are a few key features that good health information technology (HIT) systems should have. One feature is the ability to draw on electronic sources of information from the preadmission world, such as outpatient electronic medication records, pharmacy information, or a recent discharge from the hospital. The best systems take that information in and synthesize it in some way so that you’re not just getting this big jumble of medications from different sources. Second, taking the history and then deciding what you’re going to do with those medications should help you write the admission orders. If, for example, you’ve got a medication from the history and you decide that you want to continue it on admission, the HIT system should help you actually write that order. It’s very easy to do on paper; it’s actually been surprisingly hard to do with electronic systems. At discharge, the system should clearly lay out what the patient came in on, what the patient is currently on, and help to create the discharge order set. It should be able to produce a really coherent set of instructions for both the patients and the providers, showing how the discharge medication list is different from what the patient was taking prior to admission.

–Mary Ellen Schneider

Take us to your leader. Nominate a hospitalist whose work inspires you. E-mail suggestions to m.schneider@elsevier.com.

Reducing harm through medication reconciliation