By 6 a.m. each morning, an electronic dashboard displaying real-time information about the 8 to 10 patients at highest risk for acute kidney injury (AKI) is available for review by the nephrologists at Phoenix Children’s Hospital.
“We have an entire day’s work cut out for us and see who needs the attention,” said Kanwel Kher, MD, chief of the division of nephrology at the hospital. “We have to evaluate them quickly, we can do that in about 10 to 15 minutes.”
The nephrologists can access patients’ complete charts through links in the dashboard, which is integrated with the hospital’s electronic medical record system (EMR). Using a secure text message system, Kher and his colleagues can send their recommendations for reducing the patient’s risk of AKI directly to the child’s attending physician.
The effort is part of a growing movement in the field of nephrology to use electronic records, e-alerts, and other tools to more proactively prevent AKI and its progression in hospitalized patients. So far, the data on using e-alerts have been mixed. But some initiatives leveraging electronic tools have shown promise at reducing AKI, and experts predict that they will increasingly become a routine part of practice as their use and the technology matures.
“Acute kidney injury is the most common complication of critical illnesses and has a dire impact on patients’ outcomes,” said Kianoush Keshani, MD, MS, a nephrology intensivist at the Mayo Clinic. “Anything that we can do to potentially avoid its development or progression to higher stages would potentially have an impact on mortality.”
E-alerts
Electronic medical records contain important information that can be used to identify patients with AKI or those at risk. So, some hospitals have developed computer programs that can help alert physicians when a patient’s creatinine levels have spiked or their urine production decreases, signaling kidney injury. Or they can warn physicians when a patient is receiving a medication or combination of medications that puts them at risk of such an injury.
“We want the electronic health record to assist clinicians to provide better care to patients,” said Kashani.
Despite the promise of such electronic tools, the effects of these alerts on care processes or patient outcomes have been mixed. Technical, systems, and human level challenges have hampered their effectiveness. In some cases, alerts have come too late to be useful. In others, physicians haven’t been able to easily access all the patient information they need to respond or the information didn’t fit smoothly into their workflow. Many hospitals in the US still have very basic electronic health records and don’t have the capacity to provide such alerts, Keshani said.
Physicians may also develop alert-fatigue when inundated with too many electronic warnings, he noted.
“If you go to the intensive care unit (ICU), there are so many alerts, so many alarms, that after a little while the providers get tired,” he said. “They don’t pay attention to any of those alerts and that has potential risk for the time that patients really need attention.”
Some studies have shown that e-alerts can improve care processes and have identified what works and what doesn’t. For example, Keshani noted physicians tend to ignore warnings not to use vancomycin, which can be nephrotoxic, but they are more receptive to prompts that suggest alternative antibiotics. The alerts also have been more successful in settings with high rates of AKI, such as the ICU, and when they are sent to higher-level physicians.
Next generation
To move beyond the limitations of e-alerts alone, many clinicians and centers have developed more sophisticated systems that leverage computers, pharmacists, and nephrologists’ expertise.
Some hospitals have developed an electronic “sniffer” that ferrets out cases of AKI or those at risk, allowing nephrologists or others to continuously monitor patients. Keshani has one running on a laptop in his office that keeps track of all AKI cases in his center’s ICUs.
“I know who, when, and what stage of acute kidney injury exists across all ICUs,” he said.
At Cincinnati Children’s Hospital Medical Center, the Nephrotoxic Injury Negated by Just-in-time Action (NINJA) project takes its sniffer to the next level. It collects data on every noncritically ill child who is exposed to 3 or more nephrotoxic medications or an intravenous aminoglycoside for 3 or more consecutive days. Each day at 11 a.m. hospital pharmacists and Stuart Goldstein, MD, director of the hospital’s Center for Acute Care Nephrology, receive a secure e-mail detailing the findings. The findings and any relevant recommendations are then discussed with each child’s clinician during rounds. The hospital also adopted a policy of testing creatinine levels daily to assess for acute kidney injury in each of these at-risk children. They’ve also created software that collects all the data, which is also reviewed by the quality improvement and research teams.
In its first 3 years, the program had become “part of the culture” at the hospital and has decreased the number of children exposed to 3 nephrotoxic medications by 38%, protecting 700 children from such exposures. It also reduced AKI rates by 68% averting almost 400 cases, Goldstein said.
“Our vision is that children should get nephrotoxic medications they need only for the time they need them,” Goldstein said. NINJA allows the very precise and reliable collection of information on nephrotoxic drug exposure and disseminates it to health care teams “so they can make a decision near real time at the bedside,” he said.
Twelve other pediatric hospitals have already implemented NINJA, with comparable results. By 2020, Goldstein expects 140 pediatric hospitals in the United States and Canada will be participating.
Preliminary data on the program at Phoenix Children’s shows that there has been a 34% reduction in stage 1 AKI, a 56% reduction in stage 2 AKI, and a 61% reduction in stage 3 in the first 10 months of the program, according to Vinay Vaidya, MD, the hospital’s Vice President and Chief Medical Information Officer and developer of the software for the dashboard. Next, they hope to add automated EMR alerts into the program.
“The alert can stop you just in time from adding that Toradol in a surgical patient who’s already in stage one [AKI],” he said. But the dashboard allows more comprehensive views, so they “complement each other.”
Physicians in other departments have also started leveraging the dashboard to look at their own patients or their departments. For example, a hematology oncologist is using it to reduce aminoglycoside-linked kidney damage and sends e-mails to his group about patients at risk.
“If you make it easy, generally most of the folks are eager to do the right thing,” Vaidya said.
The next step for such automated monitoring programs is to use biomarkers like urinary neutrophil gelatinase-associated lipocalin (NGAL) to identify patients earlier through noninvasive urine tests, something both Goldstein and the Universtity of Alabama Children’s team are working on. The Phoenix Children’s team also has added NGAL to its dashboard. Goldstein is also working with multiple collaborators to find ways to use genetic data to personalize patients’ AKI risk assessments.
All of these programs are part of a larger shift in the field of nephrology to become more proactive, said Keshani. He said this new model of nephrology care is more similar to the way infectious disease specialists practice by coming in each morning and reviewing all the antibiotics to ensure judicious use.
“This is the future that is coming into [nephrology] practice very, very quickly,” Keshani said.
