NINJA Program Successfully Reduces AKI from Nephrotoxic Medications

Exposure to nephrotoxic medications is a major cause of acute kidney injury (AKI) in hospitalized children, increasing the costs and length of hospital stays. In one study of pediatric patients not in the intensive care unit, 86% were exposed to a potentially nephrotoxic medication at some point during hospitalization (1). When children receive three or more nephrotoxic medications in the same day, the rates of AKI double (1). In some patients, the damage is permanent, leading to chronic kidney disease.

Stuart L. Goldstein, MD, director of the Center for Acute Care Nephrology at the Cincinnati Children’s Hospital Medical Center, and his colleagues were motivated to do something about the high nephrotoxic burden experienced by their patients.

When they looked at the epidemiology of nephrotoxic medication exposure, they found that patients were having their kidney functions reliably measured by creatinine screening only about 50% of the time. At the time, nephrotoxic AKI was viewed as a largely nonmodifiable adverse event for hospitalized patients.

“We felt if we had more reliable kidney function assessment, we could identify acute kidney injury earlier and potentially decrease unnecessary nephrotoxic medication exposure or react to changes in kidney function sooner,” Goldstein said.

Goldstein and colleagues developed the Nephrotoxic Injury Negated by Just-in Time Action (NINJA) program in the hopes of decreasing nephrotoxic medication exposure at their center (2). The program leveraged electronic health records to provide notifications to pharmacists who were coordinating with clinical teams. These alerts recommended obtaining daily creatinine levels in all patients not in the intensive care unit who were deemed at high risk for AKI secondary to nephrotoxic medications. Patients were defined as high risk if they had received intravenous aminoglycosides for 3 or more days or were receiving three or more nephrotoxic medications on the same day.

“We know children need to get nephrotoxic medications,” said Goldstein. “They are lifesaving. We weren’t telling anyone they couldn’t give a specific medication. But our overall guiding vision is that children should get the nephrotoxic medications they need for only the duration that they need them.”

At Cincinnati Children’s, the NINJA program resulted in sustained reductions in nephrotoxic exposures and rates of related AKI. In results reported in 2016, Goldstein and colleagues reported single-center results over a 43-month study period. Overall, the exposure rate to nephrotoxic medications decreased by 38%, and the rate of AKI decreased by 64% (3). Moreover, the team demonstrated that the program had not had a deleterious secondary effect in persistent infections—a potential concern if clinicians had switched to less nephrotoxic but less efficacious medications.

A recent study in Kidney International demonstrates the broader potential of the NINJA program through a prospective trial across nine diverse pediatric centers in the United States (4). Details of the program implementation varied, but all included an interprofessional 2-day learning session. At each center, an alert triggered by the electronic health record prompted medication review (and substitution if appropriate) and also assessment of daily creatinine values for high-risk patients.

The primary endpoint was the number of episodes of AKI per 1000 patient hospital days. The centers varied in time to maturity: the speed at which they fully implemented the goals of the NINJA program. AKI was defined by the Kidney Disease Improving Global Outcomes serum creatinine criteria (an increase of 50% or 0.3 mg/dL from baseline). Medications to be included in the algorithm were reviewed and regularly updated by a consensus committee (5).

The authors estimated that over the 2-year study, 242 cases of AKI were prevented. Statistical process control analysis showed that the program yielded a 23.8% decrease in the rates of AKI from nephrotoxic exposure: an improvement in prevalence from 1.7 to 1.3 episodes per 1000 patient hospital days. The analysis also showed a 36.7% reduction in the rates of AKI per exposure to nephrotoxic medication (4).

These decreases were largely driven by the centers with initially higher rates of AKI. Of the six centers that began the initiative with relatively high rates of AKI, five ultimately achieved a significantly lower rate. By contrast, the three centers that started with relatively low rates of AKI did not demonstrate improvements in rates. Eight of the nine centers ended the study clustered around the same rate of AKI (4).

Those rates were exactly what they saw in Cincinnati, Goldstein said. “I think the novel aspect of this quality improvement project is that I think we’ve demonstrated not only what centers should strive for, but what may be achievable through our intervention in the way that we’ve designed it.” He suggested that future innovations may be necessary to show additional improvements.

Although the original grant sponsoring the NINJA initiative has been completed, the sites continue to gather data. Goldstein and colleagues hope to obtain additional grant funding and use the existing NINJA structure to analyze different potential variables that might lead to even lower rates of AKI, such as specific nephrotoxic medication combinations, current level of kidney function, and genetic predispositions.

The Children’s Hospitals’ Solutions for Patient Safety is a network of over 135 pediatric hospitals in the United States and Canada aimed at reducing hospital-related medical conditions. Through a grant by the Centers for Medicare & Medicaid Services, it has rolled out NINJA in about 30 hospitals, with the ultimate goal of using NINJA throughout the entire network.

The original algorithm created at Cincinnati Children’s Hospital has since been licensed to the VigiLanz digital healthcare intelligence firm, but any electronic health records platform can interface with the NINJA program. Some institutions are assessing the costs of developing their own interface versus adopting the VigiLanz platform.

In adult patients, quality improvement projects focused on preventing AKI have shown mixed results (6–8). Most of them studied the effects of alerting clinicians after patients already had acute kidney damage. By contrast, the NINJA approach is upstream, identifying patients at risk and attempting to lower that risk. Unlike some other interventions, NINJA does not mandate specific clinical action other than surveillance serum creatinine values for high-risk patients.

Currently, other investigators are evaluating the NINJA platform in adult populations. “This has all been demonstrated in the pediatric population,” said Goldstein. “There is no reason this should not translate well to hospitalized adults.”