Automated AKI Alert Prompts Change of Care, Shows No Impact on Patient Outcomes

Published on: 

Use of the electronic AKI alert system led to slight improvements in intervention rates and AKI diagnosis, although kidney function and patient-centered outcomes did not differ from patients receiving usual care.

Despite having a notable impact on the diagnostic rate of acute kidney injury (AKI) and the proportion of patients receiving multiple interventions, findings from a recent study suggest the use of an electronic AKI alert system does not improve clinical outcomes.1

The programmatic AKI alert system monitored changes in patients’ serum creatine (SCr) and notified clinicians by phone when potential AKI episodes were detected, although kidney function and other patient-centered outcomes in this group did not differ from patients receiving usual care.1

“The occurrence of AKI increases the risk of in-hospital mortality and the development of cardiovascular disease and chronic kidney disease, with substantial resource and economic implications. Early detection and prompt intervention are critical for improved patient outcomes,” wrote Hui-Juan Mao, MD, PhD, professor in the department of nephrology at The First Affiliated Hospital of Nanjing Medical University in China, and colleagues.1

A sudden episode of kidney failure or kidney damage, AKI causes a buildup of waste in the blood and disrupts the balance of bodily fluid. Prompt detection of AKI is essential because it can lead to chronic kidney disease, kidney failure, heart disease, or death. The recent development of an electronic AKI alert has shown promise for early AKI detection and potentially improved patient outcomes, although current research has yielded mixed results.1,2

To assess the viability of an AKI alert combined with a care bundle for improving the care and clinical outcomes of hospitalized patients with AKI, investigators conducted a single-center, double-blind, parallel-group, randomized clinical trial in a tertiary teaching hospital in Nanjing, China, from August 1, 2019, to December 31, 2021.1

Participants included inpatient adults ≥ 18 years of age with hospital-acquired AKI, defined using the Kidney Disease: Improving Global Outcomes (KDIGO) creatinine criteria. Exclusion criteria were baseline estimated glomerular filtration rate (eGFR) < 15 mL/min/1.73m2, admission diagnosis of end-stage kidney disease, history of kidney transplant, AKI occurring outside the hospital, hospitalization for < 24 hours, and baseline SCr levels < 0.5 mg/dL.1

Based on these criteria, investigators identified 2208 patients and randomized them in a 1:1 ratio to either the alert group or the usual care group, stratified by medical vs surgical ward and by intensive care unit (ICU) vs non-ICU setting.1

The intervention in the alert group was an electronic AKI alert using changes in SCr to facilitate early detection of AKI through a computerized algorithm and notify clinicians via phone of an AKI episode. A care bundle accompanied the AKI alert and consisted of general, non-individualized, and nonmandatory AKI management measures.1

The primary outcome was maximum change in eGFR within 7 days of randomization. Secondary patient-centered outcomes included death, dialysis, AKI progression, and AKI recovery, while care-centered outcomes included diagnostic and therapeutic interventions for AKI.1

A total of 2208 patients who were predominantly male (70.7%) with a median age of 65 (Interquartile range [IQR], 54 to 72) years were randomized to the alert group (n = 1123) or the usual care group (n = 1085).1

Within 7 days of randomization, median maximum absolute changes in eGFR were 3.7 (IQR, −6.4 to 19.3) mL/min/1.73 m2 in the alert group and 2.9 (IQR, −9.2 to 16.9) mL/min/1.73 m2 in the usual care group (P = .24). Investigators pointed out there was also no difference in the median maximum relative changes in eGFR within 7 days after randomization between the alert (9.4%; IQR, −14.8% to 47.0%) and usual care (7.7%; IQR, −18.6% to 44.0%; P = .26) groups.1

The groups also did not differ for median maximum absolute changes in SCr level (0.0; IQR, −0.5 to 0.2 mg/dL for both groups; P = .59) and maximum relative changes in SCr level (−1.9%; IQR, −33.5% to 15.0% vs −0.7%; IQR, −31.8% to 17.1%; P = .40) within 7 days of randomization. Patient-centered secondary outcomes for in-hospital death, in-hospital dialysis, dialysis within 7 days, death within 7 days, death within 28 days, death within 90 days, in-hospital AKI progression, highest AKI stage achieved, AKI recovery at discharge, and incidence of survival with dialysis dependency at 90 days were similar between the alert and usual care groups.1

Of note, significantly more patients in the alert group received intravenous fluids within 2 days of randomization and had urinalysis, fluid intake and output measurements, subsequent SCr level measurements, kidney ultrasonography, and corrected anemia by reaching hemoglobin > 9.0 g/dL compared to the usual care group (all P <.001).1

Investigators called attention to several potential limitations to the study, highlighting its single-center design, potentially weak care bundle accompanying the AKI alert, and use of the KDIGO creatine standard to design AKI alerts due to lack of urine output data.1

“We found that AKI alerts could change the care of AKI, but these changes did not improve short-term kidney function and other clinical outcomes. These findings suggest that a combination of high-quality interventions and AKI alerts is warranted in future clinical practice,” investigators concluded.1


  1. Li T, Wu B, Li L, et al. Automated Electronic Alert for the Care and Outcomes of Adults With Acute Kidney Injury: A Randomized Clinical Trial. JAMA Netw Open. 2024;7(1):e2351710. doi:10.1001/jamanetworkopen.2023.51710
  2. National Kidney Foundation. Acute Kidney Injury (AKI). Accessed January 26, 2024.