Medication Management: Can IT Systems Minimize Human Error?
 |
IT-enabled medication management strategies tap computerized provider order entry (CPOE), bar-coded electronic medication administration records and clinical decision support modules in electronic health records to reduce medication errors. They have their work cut out for them: The Institute of Medicine estimates that 1.5 million preventable medication errors occur in the U.S. every year, costing some $8 billion.
Reducing these errors is a major focus of the federal push for electronic medical records with CPOE. But even the technologies’ biggest advocates say IT systems might address one type of error only to introduce a new one. For example, studies have shown that electronic medication ordering can reduce transcription errors by 55 to 85 percent, but side effects might include workflow disruption and silos of inaccessible patient information.
“I don’t think too many people would argue that the need for doing things electronically isn’t important, but we are examining these systems closely to better learn how to prevent these errors,” says Gordon D. Schiff, MD, associate director of Patient Safety Research and Practice at Brigham and Women’s Hospital in Boston.
Fighting errors with data
Clinical decision support (CDS) features built into the home-grown EHR at Brigham and Women’s enable staff to “ensure information doesn’t fall through the cracks,” says Schiff. “If you have abnormal test results when you’re ordering a drug, the system checks for these so the order doesn’t conflict with a patient’s allergy information.” If the system detects an allergy or potential interaction, a warning pops up on the screen, alerting the physician to possible contraindications and whether medications may not be handled well by a patient, he says.
The University of Maryland Medical Center (UMMC) in Baltimore adopted a Cerner Millennium Pharmnet CPOE system in 2007, along with tailored CDS add-on tools in the Cerner EHR. The system enables medication to be ordered and dispensed electronically by UMMC’s Omnicell automated dispensing cabinets system.
UMMC found that standard out-of-the-box EHR tools for medication management were “too conservative,” so the facility customized tools to scan for allergies, drug-to-drug and drug-to-food interactions, and medication duplications, says Mark Kelemen, MD, CMIO at UMMC.
“Customization of these tools is time-consuming, but is essential to developing a used and useful program, accepted by providers and effective clinically,” says Agnes Ann Feemster, PharmD, assistant director of Pharmacy at UMMC.
CPOE has provided UMMC with faster order turnaround times, while CDS tools have helped the facility meet core measures and The Joint Commission’s National Patient Safety Goals related to medication reconciliation, Feemster says. In addition, UMMC’s use of diagnosis-based order sets and restrictive formularies have helped decrease contraindications and therapeutic duplication, Feemster and Kelemen say.
| Syncing medication data: The next great challenge |
| Synchronizing data from various entities—the hospital pharmacy, the primary care physician’s office and insurance claim databases, to name three—will help researchers better understand systemic medication discrepancies and trends. However, integrating clinical systems data and updating patient information across systems “will be one of the next decade’s great challenges,” says Mark Kelemen, MD, CMIO at University of Maryland Medical Center in Baltimore. To study potential health IT-induced errors, a research team at Brigham and Women’s Hospital in Boston is analyzing nationally collected errors reports in the MEDMARX Medication Error reporting system, with the goal of clarifying where these errors occur and how IT systems can be improved to avert them, says Gordon D. Schiff, MD, associate director of Patient Safety Research and Practice at Brigham and Women’s. “We are trying to find if there are any recurrent patterns, such as staff putting in the wrong drug, or confusion between pharmacy systems and ordering and dispensing systems,” he says. Similarly, the Medication Management Research Network (MMRN) at the University at Buffalo (UB) in Western New York is conducting research projects to evaluate how health IT solutions can streamline patient data and analyze medication duplications and incorrect medication doses, says Gene D. Morse, PharmD, associate director of the New York State Center of Excellence in Bioinformatics and Life Sciences at UB. The MMRN is partnering with area health systems and community pharmacies to evaluate data in their systems and insurance claims systems to better monitor refills, medication adherence and use formularies to find less-expensive, generic forms of medication. UB also is looking at how personal health record (PHR) and patient smart cards could help retain patient information and transmit data among physicians, the hospital and the pharmacy, to root out incongruities, says Morse. Like UB, the University of California San Francisco analyzes patient data and medication use for future treatments and research projects, says Nancy Nkansah, PharmD, assistant professor of Clinical Pharmacy and director of UCSF’s Medication Management Services, in Fresno, Calif. Doing so “allows us to really drill down to ensure that we are identifying potential drug-related problems and ensure that the meds patients are on are meeting the outcomes and aren’t causing any safety issues,” she says. |
Tracking progress
Clinical Support Systems’ web-based MTMPath software has enabled the University of California San Francisco’s Medication Management Services to expedite patient medication tracking and consistency of orders, says Nancy T. Nkansah, PharmD, assistant professor of Clinical Pharmacy and director of UCSF’s Medication Management Services, in Fresno, Calif. MTMPath includes risk-assessment tools that check patient history and medication data, and Nkansah uses Micromedex and Lexi-Comp CDS tools to run checks for drug interactions, she says.
MTMPath uses a systematic approach to specify drug-related problems and can indicate reasons why a patient is on a medication and for how long, says Nkansah. With this information available, UCSF’s medication management facility has performed 41 medication regimen reviews using MTMPath, says Nkansah. The results were sobering: All but two patient records showed potential medication-related problems.
The facility’s systemwide checks on medications have helped identify drug reactions and better manage anticoagulation and diabetes medication while also increasing patients’ adherence to their medication regimen, says Nkansah.
Barring the door
Using bar-coded electronic medication administration record (EMAR) reduces transcription and medication administration errors, as well as potential drug-related adverse events, according to a 2010 study funded by the Agency for Healthcare Research and Quality and published in the May 6 issue of the New England Journal of Medicine.
Individual medication doses are bar-coded in the pharmacy. Using bar-code scanning technology similar to what’s found in retail stores, a caregiver scans the bar code on the medication package to ensure it corresponds to the code on a patient’s wristband. The EMAR tracks the medications throughout the process.
For bar-coded EMAR e-prescribing to work, pharmacy and hospital ordering systems must be integrated to “speak the same language” and work on a transparent, automatic system, Schiff says. At UMMC, bar coding helps prevent errors during administration of medications, according to Feemster and Kelemen.
Finally, the very human habit of relying on technology to solve problems can be hard to break: “Certain critical thinking skills lapse after the implementation of new technology,” says UMMC’s Kelemen, although improved staff training and re-training could prevent this.
Nevertheless, “I can’t imagine being able to safely prescribe drugs without using the computer,” says Schiff. “I’m a great believer in using these tools to help patients and to help physicians and pharmacists, but we also have to be learning from them and continually re-assessing and revising them to work as good as possible.”