Computerized Physician Order Entry: Accomplishments And Remaining Challenges
April 6th, 2010
 by
Leah Binder
|
A few weeks ago I toured a hospital that had recently adopted computerized physician order entry (CPOE). We visited a patient room on the Med-Surge floor, where a nurse explained what happens in an emergency. Before they adopted CPOE, nurses would run 300 feet down the hall to find the patient’s paper record during a code, and race back, flipping frantically through the chart to find what medications the patient was on. Now that they had adopted CPOE, one touch of a keyboard at the bedside was all it took for clinicians to immediately access the patient’s current medications and vital signs. According to the nurse leading the tour, the CPOE system reduced errors and saved time and lives.
This is the kind of triumphant advancement in quality, efficiency, and safety The Leapfrog Group founders pictured when they insisted on adoption of CPOE as one of the first standards on the Leapfrog Hospital Survey. When Leapfrog first announced the standard back in 2000, conventional wisdom called us quixotic at best; only about 2 percent of hospitals reported adopting CPOE, and most said they were years away from even considering such a large and expensive initiative. Fast forward nearly a decade, and today nearly 12 percent of hospitals meet the Leapfrog standard, and CPOE adoption is a centerpiece of national health policy, with federal dollars available to hospitals that move forward in the next two to three years.
Yet progress is not without pain. During my tour of that same hospital, another nurse pulled out an index card covered with tiny, painstaking handwriting: this was her day’s plan for the shift. Because of the way the CPOE system was formulated to communicate and process medication orders, she needed to write a new index card every day or else she would find herself willy-nilly up and down the hall not accomplishing what patients needed. Obviously, this workaround was not intended when CPOE adoption went forward: this daily index card is an ideal platform for medical errors, and hardly an enhancement to clinical efficiency.
Such unintended consequences are predicted by Jane Metzger and her colleagues, and articulated convincingly, in their study “Mixed Results in the Safety Performance of Computerized Physician Order Entry.” Writing in the newly released April issue of Health Affairs, the researchers demonstrate that CPOE adoption per se does not guarantee optimal patient safety outcomes—systems must be tested and perfected over time to ensure best performance. But they go beyond arguing for good tests to monitor technology: Metzger and her colleagues pioneered development of Leapfrog’s groundbreaking CPOE Evaluation Tool, the only such tool we know of in the public domain. Their study today reports early results from 62 hospitals that used the CPOE Evaluation Tool on Leapfrog’s Hospital Survey in 2008, testing whether CPOE systems would flag orders that would result in adverse events or even death. According to the study, systems did not flag half of the fatal orders, and missed disturbing numbers of orders that would have resulted in serious adverse events.
What Should Be Done Now?
Leapfrog offers several recommendations from the findings of the CPOE Evaluation Tool. First, alongside federal investment in technology, and as part of the definition of meaningful use, there needs to be a testing and monitoring component for all technology adoption in hospitals. It seems obvious that systems as complex as CPOE need to be monitored, but until Jane Metzger and her colleagues developed Leapfrog’s CPOE Evaluation Tool there was nothing we know of in the public domain to test any form of health information technology.
Second, we must find a way to share information transparently about best practices for adoption of health information technology in hospitals. Technology systems are not plug and play; they require thoughtful engagement of all stakeholders in the hospital system. Currently hospitals aiming to invest in CPOE or other HIT systems rely on their vendor to map a process for adoption, or invent their own process. This is not always efficient and results in performance variation that the study by Metzger and coauthors describes. Leapfrog is collaborating with HIMSS to formulate a set of best practices and disseminate them publicly.
Third and most importantly, Leapfrog and our purchaser members continue to hold hospitals accountable for adopting CPOE. The evidence that CPOE saves lives and prevents the most common adverse event in hospitals—a medication error—remains abundant and urgent. CPOE systems can reduce the number of adverse drug events by up to 88 percent, preventing three million serious medication errors in the U.S. each year, saving billions of dollars and alleviating significant human suffering.
I saw that on my hospital tour. On the one hand, nurses’ index cards masked fundamental problems in the adoption of CPOE, and Leapfrog test results are discouraging. But on the other hand, when there is a code at that hospital, clinicians use technology that any patient would hope for and expect when his or her life is at stake. Over time, the hospital will figure out how to eliminate the index cards and improve its performance on Leapfrog’s test. The lifesaving potential is worth the trouble.
As medicine grows ever more complex, it will not be adequate to rely on the individual memories of each and every clinician to ensure that a plethora of medication errors are avoided. We will need to rely on advancing technology to support clinicians, and we will need to improve on the performance of that technology over time. Leapfrog thanks the developers of the CPOE Evaluation Tool, Jane Metzger, Emily Welebob, David Bates, Stuart Lipsitz, and David Classen, for their vision and excellent research. Thanks to their pioneering work, we can pursue the best possible use of emerging health information.
Email This Post
Print This Post
