More hospitals are choosing to adopt secure clinical text-messaging systems, rather than paging, for more efficient communication between nurses and physicians, but this can cause a higher volume of texts that interrupt practice.
“Residents don't want to hear their team smartphone go off when they're in the middle of a procedure, talking to a patient, or in a teaching session, and wonder whether it's just about a diet order or whether it's something important,” said Arielle Mendel, MD, an MSc candidate in the quality improvement and patient safety program at the Institute for Health Policy Management and Evaluation at the University of Toronto. “We wanted to reduce nonurgent interruptions, and in particular, we wanted to focus on interruptions occurring during resident educational hours.”
At the hospitals, clinicians helped design the communication system and so are actively involved in changing the platform over time, she said. To reduce interruptions during teaching time, they launched a quality improvement (QI) project in 2015 targeting eight teaching teams at the two hospitals.
How it works
The QI team worked with nursing leadership to define for nurses which events would necessitate an urgent versus nonurgent interruption message. If a nurse was about to send a nonurgent message during educational hours, he or she would see an automatic alert message that it was protected educational time, as well as an option to send it later. “We also modified the existing platform so that nonurgent messages would not ‘expire’ during educational hours and start causing noisy alerts,” said Dr. Mendel. “It sounds very simple, but before, these messages were expiring at the wrong time, so they were starting to alert residents while they were in teaching.”
Nurses could override the alert and send the message anyway. Over time, the team used data from the system to measure the number of interruptions happening during educational hours, she said. To make sure nurses were still reaching residents in urgent situations, three independent reviewers audited consecutive messages from one month in 2014 and one month in 2016. “We blinded them to the year that they were sent. That way, we could grade the urgency of these messages and come up with a consensus of what was urgent or not,” Dr. Mendel said.
After the intervention, the volume of nonurgent messages, based on the audited samples, decreased by about one-third, she said. From January 2014 to December 2016, the overall interruption frequency decreased from an average of 0.92 to 0.59 message per team per educational hour, according to results published online in April by the Journal of Hospital Medicine. In addition, the proportion of nonurgent messages sent during educational hours decreased from 82% at baseline to 68% in 2016, while the number of urgent messages remained constant.
There was difficulty with retrospectively judging the urgency of messages. “Sometimes without the clinical context at that moment when it was sent, you can't really tell,” said Dr. Mendel. In addition, a challenge with any alert-based intervention is the risk of alert fatigue; nurses may start ignoring the warnings and reduce the effectiveness of the intervention, she noted. “We haven't looked at this and measured that. Going forward, it would be important to look to see, is the effect being sustained?” Dr. Mendel said.
The QI initiative is still in effect and may evolve to reduce nonurgent interruptions, including phone call interruptions, at all times of the day in addition to during educational sessions, she said. “This has been a long saga of iterative change, and it's definitely going to continue over time. . . . There may even be unintended consequences from our intervention that we haven't measured yet,” Dr. Mendel said.
Words of wisdom
Even when teaching hospitals adopt clinical text messaging versus paging, interruptions during educational time are still going to be a widespread problem, she said. “If other institutions develop the ability to modify the electronic platform [and] to customize it . . . I think that [this intervention] could be something generalizable,” Dr. Mendel said.