Daniel M. Keller, PhD

November 16, 2012

SAN DIEGO, California — An electronic alerting system for acute kidney injury (AKI) helps general practitioners identify and manage postoperative patients with AKI and some unexpected cases of AKI, according to a new report.

Speaking here at Kidney Week 2012, Edward Stern, MB BS, renal registrar (fellow) at University College London, and clinical lecturer at the Royal Free Hospital Centre for Nephrology, United Kingdom, described the goals of the London Acute Kidney Injury Network (LAKIN).

The network, a collaboration among the 7 London renal units, aims to improve early recognition of AKI, highlight the significance of small absolute increases in serum creatinine, and provide access to standardized clinical management of AKI. It uses a hub-and-spoke model of care to improve access to services and speed transfers from simple bedside medicine through dialysis.

Need to Flag Increases in Creatinine

Dr. Stern cited literature showing that an increase in creatinine of 0.3 mg/dL was associated with a 4-fold increase in in-hospital mortality. A 1.0-mg/dL increase in creatinine was associated with an almost 10-fold increase in mortality and an almost 6-day increase in hospital stay for survivors.

Furthermore, he referred to several studies showing that even mild AKI doubled the risk for chronic kidney disease compared with no AKI and more than doubled the risk for end-stage renal disease. One third of postadmission AKI cases may be avoidable, and in more than 40% of cases there was an unacceptable delay in diagnosis.

Dr. Stern said the investigators wanted to design a system that recognized the significance of absolute small increases in creatinine, which are larger increases in relative terms. "The reporting systems that clinicians see use normal ranges [and flag them] when they're outside the normal range, and that's not a very helpful technique for creatinine results because we don't really think of a normal range in creatinine," he explained. "So we've tried to augment that way of reporting results to clinicians by doing a 'delta check' on creatinine, so looking at the relative change from the previous result."

LAKIN carried out a pilot project in 2 London sites: University College London Hospitals (a 900-bed multisite teaching hospital in London) and Whittington Health (a 500-bed district general hospital in North London).

The investigators designed an automated rule for the hospitals' pathology systems to identify patients whose creatinine levels increased by 50% or more at any time within 90 days of a previous creatinine determination. An increase in creatinine triggered an alert to the requesting clinician on the pathology report flagging the patient as having AKI, along with a link to the LAKIN Web site (www.londonaki.net). At their discretion, the laboratory staff could directly contact the clinician. Dr. Stern said it took 5 minutes to set up the alerting system within the pathology reporting program.

AKI Diagnoses in Hospitalized and Community Patients

At Whittington Health, the first 100 patient alerts occurred during a 50-day period. The patients consisted of 86 inpatients (including those seen in the emergency department), 7 hospital outpatients, and 7 community patients of general practitioners. During this period, the hospital performed 25,017 creatinine tests.

The median time from baseline to the alert was 20 days (range, 1 - 90 days). The mean increase in creatinine at the time of the alert was 1.3 mg/dL (range, 0.2 - 9.6 mg/dL), and the actual creatinine levels at the time of the alerts ranged from 0.5 mg/dL to 11.0 mg/dL, with a change in creatinine ranging from 51% to 1214%. Four of the 100 alerts were considered false positives because of spuriously low baseline values.

According to stratification by Kidney Disease: Improving Global Outcomes (KDIGO) stage, 56 patients had AKI stage I, 29 had AKI stage II, and 15 had AKI stage III. The 90-day mortality rates were 9%, 28%, and 13%, respectively, compared with 1.7% overall mortality for Whittington in-patients.

Unmet Medical Need

The investigators concluded that this around-the-clock system is valuable for the early identification and management of patients with an unmet medical need and is simple to implement.

The investigators are now introducing the system among 5 hospitals in North London, with a goal of having a standardized regional approach to managing AKI.

"Anything that is a stimulus for a small percentage of clinicians to think about the change in the patient's creatinine more quickly is very important," Dr. Stern advised. "This may not revolutionize our management, but it's a small part in this attempt to augment clinicians' recognition of the phenomenon of a change in the patient's creatinine and what the significance of it might be." He noted that by recognizing a problem early, patients may not require more aggressive monitoring and treatment later.

He said his message to nephrologists is that "we have a role in increasing our colleagues' understanding of relative changes in creatinine because I think in most specialties they're used to looking at numbers in the context of a normal range."

He pointed out it would be difficult to do a randomized trial of alerts vs no alerts, but researchers should look over the long term to see whether implementing the system has made a difference in hospital length of stay and mortality in patients with AKI.

LAKIN has formulated a care bundle flow chart for recommended management of all patients who have a 1.5-fold or greater increase in creatinine, available on its Web site, along with other information, guidelines, checklists, and an AKI manual, as well as a comprehensive iPhone app.

Incremental, but Not Insignificant, Increases

Glenn Chertow, MD, MPH, professor of medicine and chief of the Division of Nephrology at Stanford University School of Medicine in Palo Alto, California, commented to Medscape Medical News that he published 1 of the first studies showing that quite small incremental changes in the serum creatinine concentration were associated with rather large increases in the risk for death.

A problem in evaluating AKI in the hospital is that the increase in serum creatinine tends to lag behind the timing of the injury. "So when one uses the serum creatinine concentration or any other solute that has to accumulate before its level rises in the serum, we're already behind the 8-ball with respect to time," he said.

Compounding the problem is that when physicians are making rounds in the morning and laboratory test results are not back, a report of impaired kidney function could be missed for more than a day until the physician makes rounds the next morning. So from the time of an AKI to its recognition may actually be a couple of days.

A real-time electronic alert that does not depend on being available at 1 specific time could therefore make earlier recognition possible. "I would think that acute kidney injury would be 1 type of condition where electronic alerts could help," Dr. Chertow said. "If we use a test like the serum creatinine concentration, the earlier we know that it's rising, the less far behind the 8-ball we are."

There was no commercial funding for the study, and Dr. Stern has disclosed no relevant financial relationships. Dr. Chertow has received consulting fees or honoraria and travel support from Amgen. He is on the board of directors of Satellite Healthcare; is on the scientific advisory board for DaVita Clinical Research; is a consultant to Reata Pharmaceuticals; and has stock options in Allocure, Ardelyx, Home Dialysis Plus, PuraCath, and Thrasos.

Kidney Week 2012. Abstract TH-OR030. Presented November 1, 2012.