PROJECT SUMMARY/ABSTRACT Acute Kidney Injury (AKI) is a common complication in hospitalized patients and is associated with increased morbidity and mortality. My objective is to transform the medical care of hospitalized patients at risk for the development of AKI through the utilization of an automated, real-time, electronic health record risk assessment score allowing for early standardized nephrology intervention. The current gold standards for AKI detection (serum creatinine (SCr) and urine output) are imperfect flawed biomarkers; they are neither sensitive nor specific for AKI and often do not change until 24 to 48 hours after the initial injury. This time-lag is often compounded by the fact that nephrologists are often not called until severe AKI is present (e.g. a tripling of SCr from baseline). At the University of Chicago (UofC) we have data demonstrating that on average ward- based consult for AKI occur on hospital day 4 after a 1.6 mg/dL increase in SCr from baseline with over 40% of patients already having Stage 2 or 3 AKI. Using a multi-center cohort we derived and validated an Electronic Health Record (EHR)-based AKI risk assessment model to predict the development of SCr based AKI in hospitalized patients using patient vital signs, labs and demographics (called Electronic Signal to Prevent AKI E-STOP-AKI). E-STOP-AKI, which is entirely derived from data freely available in the EHR, accurately predicts the future development of stage 3 AKI a median (IQR) of 35 (14-97) hours before any evidence of SCr-based AKI with an AUC of 0.83. We will combine the power of E-STOP-AKI with an intervention that has been repeatedly shown to improve patient outcomes across several clinical settings, nephrology consultation. Early nephrology consultation for AKI has been associated with improved patient outcomes, lower peak serum creatinine, less severe AKI, shorter length of hospital stay, increased renal recovery post-AKI and reduced morbidity /mortality. These same studies mirrors our data in that the average UofC nephrology consultation is not called until 2 days after there is clinical evidence of SCr ?based AKI and that 35-50% of AKI patients are not referred to nephrology until 5 days after their clinical diagnosis of AKI. As such we seek to combine these 2 tools, E-STOP-AKI and early standardized nephrology consultation in a randomized trial aimed to mitigate severe AKI in high risk patients. Patients at high risk for AKI, as measured by an elevated E-STOP-AKI score will be randomized to receive an early structured individualized nephrology consult which will explicitly comment on fundamental issues which have been shown to impact AKI severity and outcomes (Differential Diagnosis of AKI, nephrotoxins/ drug dosing and volume status /renal perfusion) versus usual care. We hypothesize that combining the precision medicine approach of E-STOP AKI with early standardized real-time individualized nephrology-centered care in ward patients will improve patient outcomes (as measured by lower peak SCr, less severe AKI, fewer ICU transfers, shorter LOS and less mortality). If this hypothesis is correct it will serve as a major paradigm shift in AKI consultative care, using an electronic risk assessment tool to bring the expert physician to the high-risk patient's bedside several days earlier than the current standard of care.

R21DK113420

2018-08-01

2021-07-31