Mortality Decreases After Implementation of Computerized Physician Order Entry System

Melissa Lee Phillips

May 03, 2010

May 3, 2010 — For the first time, researchers have found a decrease in hospitalwide mortality after implementation of a commercial computerized physician order entry (CPOE) system. A study published online today in Pediatrics reports a 20% reduction in mortality 18 months after CPOE was installed at a children's hospital.

The result contrasts with previous studies that found that CPOE systems either had no effect or were correlated with an increase in mortality in pediatric populations.

The authors conclude that their data support recommendations made by the Institute of Medicine, the US Department of Health and Human Services, and others that hospitals should "implement CPOE as a strategy for improving patient safety and quality of care at a national level."

The findings should also apply to adult populations and to a wide variety of commercial CPOE systems, lead author Christopher Longhurst, MD, from Lucile Packard Children's Hospital at Stanford University in Palo Alto, California, told Medscape Pediatrics.

Mixed Results

During the past several years, many hospitals have implemented systems for electronic medical records, including CPOE. Studies have shown that such systems can reduce medication errors and adverse drug events while increasing efficiency and turnaround times. However, analyses have failed to find evidence that these benefits translate into a significant reduction in patient mortality rates.

In 2005, the Children's Hospital of Pittsburgh in Pennsylvania discovered that implementation of a commercially available CPOE system correlated with an increase in mortality in a pediatric critical care unit (Pediatrics. 2005;116:1506-1512). That was "a really drastic finding," Dr. Longhurst said. "We're all putting these systems in because we believe they can improve patient safety, and here was something that showed just the opposite."

In a 2006 paper from Seattle Children's Hospital in Washington, researchers reported no difference in mortality after CPOE implementation (Pediatrics. 2006;118:290-295), which was still not a positive result, but it at least told clinicians that "these systems could be safely implemented without causing harm to patients," Dr. Longhurst pointed out.

In the fall of 2007, Dr. Longhurst and colleagues implemented a CPOE system from Cerner Corporation in Kansas City, Missouri (the same vendor used in Pittsburgh and Seattle). They followed published guidelines for CPOE implementation and also talked with physicians and researchers at other hospitals, making alterations in response to what others had learned.

For example, Dr. Longhurst said, Children's Hospital of Pittsburgh initially set up the system so that orders could only be placed for patients once they had physically arrived at the hospital — a restriction that may have been at least partly responsible for the associated increase in mortality, he explained.

The Stanford scientists followed patient outcomes for 18 months (17,432 total patients) and compared them with historical outcomes from 2001 through 2007 (80,063 patients). After CPOE implementation, the mean monthly adjusted mortality rate decreased by 20%, which the researchers estimate translated to 36 lives saved during those 18 months.

Finding Causes

Dr. Longhurst and colleagues speculate that a number of potential factors could have led to a reduction in mortality, including hardwired order sets; standardized patient care orders; remote access of orders, vital signs, and medication information; and eliminating redundant medication transcription by pharmacists.

However, according to Dean Sittig, PhD, from the School of Health Information Sciences in the University of Texas Health Science Center at Houston, "there are a lot of other things that just as easily could have accounted for the change. I wouldn't have thought that those things would lead to that dramatic of a decrease in mortality."

Pinning down why the Stanford hospital found a mortality decrease after CPOE implementation — and why the other hospitals did not — is complicated, Dr. Sittig told Medscape Pediatrics. "It's difficult to assign cause and effect. When we see a negative result, we say they implemented it poorly or didn't implement all the pieces, and when we see a positive result, we say they implemented it correctly."

More work needs to be done to figure out how these differences arise, Dr. Longhurst agreed, although studies suggest that outcomes depend on choices made by people during implementation, rather than on any software differences between systems. "The majority of what we're learning transcends vendor systems," he said. "Any software system can be implemented effectively or poorly, and it really has to do with the decisions during the implementation process."

Which of those decisions are the most crucial for mortality outcome largely remains unclear, but the current study is "another piece in the puzzle," Dr. Sittig said. "We need to build on it and figure out which of these factors are the keys."

The authors and Dr. Sittig have disclosed no relevant financial relationships.

Pediatrics. 2010;126:e1-e8. Published online May 3, 2010.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.