The researchers at Montefiore Medical Center in New York wanted to find a way to detect wrong-patient orders in their CPOE system. Since clinician reports were unreliable, they looked at a marker for these errors: the retraction of orders within 10 minutes of placement, followed by reorders 10 minutes later.
The investigators hypothesized that doctors had placed many of these orders on wrong patients, and it turned out they were right: Interviews with 233 physicians over a four-month period showed that 76% of the original orders had been for the wrong patients. In other words, they were near misses that could have turned into actual medical errors involving orders for imaging or lab tests or medications.
While few of the near misses would have caused serious harm according to physicians who assessed them, the number of these errors was unexpectedly high. Analyzing data on 9 million patients treated at Montefiore's four hospitals in 2009, the researchers found 6,885 "retract-and-reorder" events. A statistical analysis that applied their measuring tool to these incidents concluded that about 5,246 orders were placed on wrong patients in 2009.
That translates to wrong-patient electronic orders being entered by one in six clinicians (including physicians, midlevel practitioners, and nurses) for one in 37 patients admitted to the hospital. Put another way, practitioners at Montefiore's facilities entered an average of 14 such orders per day.
That doesn't reveal how many orders for the wrong patients went through and were carried out, said Jason Adelman, MD, a coauthor of the report and Montefiore's patient safety officer, in an interview with InformationWeek Healthcare. But based on other patient safety research and anecdotal evidence, he said, there's a strong likelihood that interventions to reduce retract-and-reorder events would also prevent actual wrong-patient errors.
Phase 2 of the Montefiore study, which lasted from December 2010 to June 2011, showed that two different interventions could decrease the number of wrong-patient orders that were later retracted. In this randomized controlled trial, about 4,000 providers were assigned to groups that were prompted to verify patient identifiers before entering orders, reenter patient identifiers, or neither. The results were encouraging: compared to the control group, ID verification reduced the odds of retract-and-reorder events by 16%; ID reentry decreased them by 41%.
It took clinicians only half a second on average to verify a patient's ID, versus 6.6 seconds for reentering identifiers. The latter could add up to a fair amount of time over many orders, but Adelman doesn't believe this explains why only half of the wrong-patient orders were prevented even with ID reentry. The physicians, all of whom agreed to participate in the trial, could not override the prompts, but many could have entered the data incorrectly. If vendors included photos of patients on the order screen, Adelman suggested, errors might be reduced further. Also, he said, additional staff might be required to do double or triple checks.
While it would be easy for vendors to program the kind of prompts used in the study in their CPOE systems, Adelman was not aware of any commercial applications that included them. However, he said, Johns Hopkins Hospital in Baltimore has begun requiring clinicians to re-enter patient identifiers before submitting orders.
In a 2011 report commissioned by the Office of the National Coordinator of Health IT, the Institute of Medicine (IOM) found that health IT can lead to errors that cause serious patient harm. Other studies have reached similar conclusions, although CPOE is believed to prevent far more mistakes than it causes.
IOM recommended that the Department of Health and Human Services (HHS) create a new watchdog agency to oversee the safety of health IT. However, Adelman pointed out, there still are not many studies showing the quantitative effects of information systems on patient safety.
Get the new, all-digital Healthcare CIO 25 issue of InformationWeek Healthcare. It's our second annual honor roll of the health IT leaders driving healthcare's transformation. (Free registration required.)
InformationWeek Must Reads Oct. 21, 2014InformationWeek's new Must Reads is a compendium of our best recent coverage of digital strategy. Learn why you should learn to embrace DevOps, how to avoid roadblocks for digital projects, what the five steps to API management are, and more.