Nicole Kovacic Scherrer, BCCCP, PharmD
West Virginia University Hospital
Morgantown, West Virginia
Disclosure information not submitted.
Erin Barreto, MSc, PharmD, FCCM
Associate Professor of Pharmacy and Medicine
Mayo Clinic
Rochester, Minnesota
Disclosure information not submitted.
Chris cook, MD
Associate Prof
WVU Medicine, J.W. Ruby Memorial Hospital, United States
Disclosure information not submitted.
Ankit Sakhuja, MBBS, FACP, FASN, FCCP
Assistant Professor Of Medicine
WVU Medicine, J.W. Ruby Memorial Hospital, United States
Disclosure information not submitted.
Title: Impact of an Electronic Acute Kidney Injury Alert on Nephrotoxin Use in a Cardiac Surgery ICU
Introduction: Electronic alerts are known to improve rates of acute kidney injury (AKI) detection, but less well understood is their impact on nephrotoxic medication use. The aim of our study was to assess the impact of an electronic AKI alert on nephrotoxic medication use within 30 hours after development of AKI among patients undergoing cardiac surgery.
Methods: This was a retrospective single center study of critically ill patients hospitalized in a cardiac surgery intensive care unit (ICU) at a tertiary care center from January 1, 2020 to December 17, 2020. The AKI alert went live in our ICU on July 1, 2020. The outcome of interest was nephrotoxic medication use in the 6 months before and after initiation of a passive electronic AKI alert. The AKI alert triggered when patients met kidney disease improving global outcomes (KDIGO) Stage 1 AKI (increase in serum creatinine by 0.3 mg/dL or more) within the previous 52 hours. Patients with alerts were compared to those without alerts who met the same AKI criteria. Multivariable logistic regression adjusted for age, sex and pre-AKI nephrotoxin exposure was used to assess the relationship between the AKI alert and nephrotoxin use.
Results: Of 184 patients with AKI, 51 were in the alert group and 133 in the no alert group. The alert group was older (mean age 68±11 years vs 59±17 years, p< 0.001) but with similar sex distribution (41% vs 37% females, p=0.7) and similar use of nephrotoxic medications within 30h before AKI (14% vs 16%, p=0.7). Nephrotoxin use 30h post AKI was numerically lower in the alert group but was not statistically significantly different (18% vs 28%, p=0.2). In multivariable models the presence of the AKI alert did not predict decreased use of nephrotoxins within 30h after AKI (OR 0.80; 95% CI 0.33-1.96).
Conclusions: The use of passive AKI alerts in this study was not associated with a statistically significant decrease in use of nephrotoxic medications. Further studies are needed to understand optimal approaches to utilize electronic AKI alerts in ICU patients.