Do Electronic Health Records Cause Patient Mixups?
Each year, there are approximately 600,000 wrong-patient orders—when an order intended for one patient is written for another patient. Such errors can lead to serious harm and even death.
To reduce the risk of wrong-patient errors, patient safety experts have recommended that electronic health records (EHRs) allow only one patient record open at a time. But according to a previous study, many doctors believe that this approach impedes their ability to care for multiple patients in busy health care settings.
A new study, published in JAMA, put the theory to the test. The study, led by researchers at Columbia University Vagelos College of Physicians and Surgeons, found that the risk of wrong-patient orders was similar whether providers were allowed to have just one patient record open or multiple patient records open at the same time.
For Jason Adelman, MD, assistant professor of medicine and lead author of the paper, the finding is reassuring from a patient-safety standpoint, though the reasons for these results are more complex.
Why did you perform this study?
The underlying assumption that having multiple records open could lead to an increase in wrong-patient errors is based on the opinion of safety experts rather than clinical trial or observational evidence. We aren’t sure where this thinking originated. It could be that when individual providers investigated their own wrong-patient orders, they noticed they had two or more records open while placing the orders. But there was no proof that this was the cause of the errors.
In our previous research, we found that health information technology officers were divided about whether to restrict or allow access to multiple records, so it’s been an open question.
What did your new study find?
Our study included 3,356 health care providers treating patients in the emergency department, hospital, or outpatient setting. Half were randomly assigned to an unrestricted group, which allowed them to open up to four patient records at a time. The other half were randomized to a group that limited them to opening only one record at a time.
We tracked the number of wrong-patient order sessions (multiple orders can be placed during one session) in each group using an electronic algorithm I developed called the Retract-and-Reorder measure. The measure identifies orders that were placed for a patient, retracted, and reordered by the same clinician for a different patient. The error rates were similar, with 88 errors per 100,000 order sessions in the restricted group and 91 errors per 100,000 order sessions in the unrestricted group.
However, providers in the unrestricted group placed most orders when a single record was open, which limited our ability to draw definitive conclusions.
We did find that within the unrestricted group, wrong-patient orders increased with the number of open records. But the association was largely eliminated when we looked at the clinical setting—clinicians in outpatient settings tended to open one record at a time and made fewer errors, while providers in the hospital and emergency department used multiple records and made more errors.
We suspect that opening multiple records is a marker for working in a busy health care environment, where clinicians are faced with interruptions and care for multiple patients simultaneously. It may be the nature of the environment, not the number of open records, that is causing the errors.
Can electronic health records improve patient safety?
Before I developed the Retract-and-Reorder measure, we relied on self-reporting to identify wrong-patient errors, which are known to be underestimated. In one study, the average number of wrong-patient order errors voluntarily reported in one year was nine per hospital, compared with over 5,000 instances identified by the Retract-and-Reorder measure in a large medical center in a single year. These are near-miss errors. Because they occur frequently, and follow the same pathway as errors that cause harm, we are able to test strategies for prevention.
As a patient safety officer and health information technology researcher, I believe that computerizing health records and patient orders does more than optimize work flow and communications. Ultimately, electronic health records will make health care much safer by offering a more accurate way to document patient care, guarding against diagnostic errors, and supporting clinical decision-making with evidence-based guidelines.
But the EHR also has the potential to introduce risk in ways we hadn’t anticipated, and we need to figure out how to guard against such risk. Some methods, such as patient ID verification alerts and patient photographs, have shown promise in reducing wrong-patient orders and are already being implemented.
For our next study, we will look at provider satisfaction—whether the technology helps providers do their jobs effectively—in the two groups from our randomized trial. We hope this will offer insight into how to support providers while preventing potentially harmful errors in different clinical settings.
Jason Adelman, MD, MS, is chief patient safety officer, associate chief quality officer, and executive director of patient safety research at Columbia University Irving Medical Center/NewYork-Presbyterian.
The paper is titled “The Effect of Restriction of the Number of Concurrently Open Records in an Electronic Health Record on Wrong-Patient Order Errors: A Randomized Clinical Trial.”
The other authors are Jo Applebaum (NewYork-Presbyterian, New York, NY), Clyde Schechter (Albert Einstein College of Medicine, Bronx, NY), Matthew Berger (Albert Einstein and Montefiore Health System, Bronx, NY), Stan Reissman (Montefiore), Raja Thota (Montefiore), Andrew Racine (Albert Einstein and Montefiore), David Vawdrey (CUIMC and NYP), Robert Green (CUIMC and NYP), Hojjat Salmasian (Harvard Medical School, Boston, MA and Brigham and Women’s), Gordon Schiff (Harvard and Brigham and Women’s), Adam Wright (Harvard and Brigham and Women’s), Adam Landman (Harvard and Brigham and Women’s), David Bates (Harvard and Brigham and Women’s), Ross Koppel (University of Pennsylvania, Philadelphia, PA, and University at Buffalo, Buffalo, NY), William Galanter (University of Illinois at Chicago, Chicago, IL), Bruce Lambert (Northwestern University, Evanston, IL), Susan Paparella (Institute for Safe Medication Practices, Horsham, PA), and William Southern (Albert Einstein and Montefiore).
The study was supported by the Agency for Healthcare Research and Quality (R01HS023704).
The authors have no financial or other conflicts of interest.