A new study by Yale School of Medicine nephrology experts shows that electronic alerts for acute kidney injury do not statistically improve patient outcomes.
F. Perry Wilson, a nephrologist at the medical school, and his team conducted the study. Its main objective was to determine whether having electronic health record alerts for acute kidney injury would affect patient outcomes of mortality, their need for dialysis treatment and progression of the condition itself.
“It seemed logical to us that the provision of alerts, just to let the provider know that their patient has acute kidney injury, would make things better,” Wilson said. “But we felt that even these common sense interventions require rigorous evaluation because it can be hard to predict necessarily what the effects are going to be.”
Acute kidney injury, or a sudden decrease in the kidney’s filtration abilities, is associated with a tenfold increase in patient mortality, according to the study.
Despite its risk, the syndrome is often missed in clinical care, as explained in the paper.
“Acute kidney injury is asymptomatic. There is no pain,” Wilson said. “It requires paying careful attention to a set of blood values that one might miss in a busy hospital situation.”
As of now, several health systems around the country have implemented acute kidney injury alarms, according to Wilson. He explained how these alarms notify the provider when signs of acute kidney injury are present to ensure the condition is not missed. However, there has been minimal data on whether the alarms improve patient outcomes.
In order to determine whether the alarm reduces risk, Wilson and his team studied 6,030 patients across six Yale New Haven Health System hospitals over 22 months. They randomly assigned half of these individuals to have the alert system in place and the rest did not. The study ended up finding that there was no significant difference in patient outcomes between the two groups.
The researchers then did a statistical test for heterogeneity of effects to determine whether the alarms had different outcomes within the individual hospitals. The findings, according to Wilson, were the opposite of what was expected.
“In the two non-teaching hospitals where we enrolled patients those who got alerts had worse outcomes — a higher rate of progression of acute kidney injury, dialysis and death,” Wilson said.
The study showed that, in some cases, the alarms were associated with higher risk. According to Wilson, these results could cause hospitals to reevaluate the existing alerting systems for acute kidney injury.
According to Stephen Latham, director of Yale Interdisciplinary Center for Bioethics and the bioethicist involved in the study, the team is having trouble explaining why the alerts in these two hospitals were linked to worse outcomes.
“We didn’t find anything that was obvious about why the outcomes should be worse,” Latham said. “The only thing that we could kind of think was that the alarms could have somehow distracted the medical team from other things that were going on with those patients, but we don’t really have any evidence of that.”
According to Latham, the study shows how important it is to carefully evaluate everything in medicine, even when it feels like common sense. Seemingly obvious benefits, such as alarm systems, should still be tested before they are put into place.
Melissa Martin, the study’s project manager, says the team has plans to perform a second trial, targeting a subset of the patients with alerts that not only notify doctors of acute kidney injury but also give specific recommendations for patient care.
“One of the limitations of the study was that it was kind of a general alert to all patients who got acute kidney injury,” Martin said. “It was pretty much just informational and didn’t really have a specific action providers could take.”
According to Martin, the alerts used in the second trial will give providers sets of actions that are catered toward specific patients.
Acute kidney injury is responsible for 1.7 million deaths globally each year, according to the International Society of Nephrology.
Kaitlin Flores | firstname.lastname@example.org