Diagnostic Concordance Between Nephrology Consultants and Remote AKI Team

Background: AKI is common and strongly associated with mortality and CKD. Although evaluating AKI in real-time has the advantage of proximity to the insult, later assessment has the benefit of time and more diagnostic information. We aimed to assess AKI etiology concordance between nephrology consultants and remote physicians that examined the chart at the time of AKI detection. Methods: We analyzed patients enrolled in the Kidney Action Team(KAT)-AKI trial(2021–2024) upon meeting KDIGO-creatinine criteria for AKI for whom a remote KAT physician documented a suspected etiology upon AKI detection. We compared characteristics and clinical outcomes of patients who had a nephrology consult within 48h of AKI detection to those who did not and assessed concordance between KAT and consultant-determined AKI etiology. Results: Of 2848 patients meeting eligibility criteria, 18% had nephrology consult within 48h of AKI detection. Those with nephrology consult had higher baseline creatinine(sCr), sCr at AKI detection and median SOFA scores, and greater CKD, DM, cirrhosis and CHF burden(Table1). Those with nephrology consult had worse 14-day AKI progression and mortality(p<0.05). Excluding those with only 1 note, the most commonly suspected etiologies were similar and included pre-renal, multifactorial, ATN, CRS and CIN, with the top 3 diagnoses within multifactorial being pre-renal, CIN, and ischemic ATI. The most frequently reclassified etiology was hypovolemic pre-renal, followed by multifactorial, ATN, CRS and obstruction(Figure1). The kappa for diagnostic agreement with consultants was low (k=0.15). Diagnostic disagreement was not associated with worse mortality or AKI progression. Conclusion: The ability of expert physicians to determine the etiology of AKI at the time of AKI diagnosis is highly limited. Most AKI etiology remains unclear during follow up. There is a need for more specific phenotyping at AKI onset to improve long term outcomes. Funding: Other U.S. Government SupportTable 1 - Characteristic*--------------------------AgeFemale sexBaseline sCr (mg/dL) sCr at AKI detection (mg/dL)SOFA scoreCKDHTNDMCirrhosis CHF Consult (n=442)-------------------------73 (62, 82)206 (47%)1.7 (1.2, 2.3)2.2 (1.6, 2.8)3.0 (2.0, 4.0)235 (53%)366 (83%)241 (55%)46 (10%236 (53%) No Consult (n=2406)--------------------------------------74 (63, 83)1136 (47%)1.2 (0.9, 1.5)1.5 (1.2, 1.9)2 (1, 3)964 (40%)1981 (82%)1112 (46%)141 (6%)1039 (43%) p-value------------------0.0210.800<0.001<0.001<0.001<0.0010.9000.002<0.001<0.001 *median (IQR) or n (%)