Under‐detection of acute kidney injury in hospitalised patients: a retrospective, multi‐site, longitudinal study

• Published in: Internal medicine journal 2020-03, Vol.50 (3), p.307-314
• Main Authors: Campbell, Craig A., Li, Ling, Kotwal, Sradha, Georgiou, Andrew, Horvath, Andrea R., Westbrook, Johanna, Endre, Zoltan
• Format: Article
• Language: English
• Subjects: acute kidney injury; Acute Kidney Injury - diagnosis; Acute Kidney Injury - epidemiology; Adult; Algorithms; Australia - epidemiology; Creatinine; diagnostic error; Hospitals; Humans; ICD‐10; Incidence; Kidney diseases; Laboratories; laboratory alert value; Longitudinal Studies; New South Wales - epidemiology; Patients; Renal function; Retrospective Studies; Risk Factors
• ISSN: 1444-0903; 1445-5994
• DOI: 10.1111/imj.14264

Background Acute kidney injury (AKI) is a rapid deterioration of renal function, often caused by a variety of co‐existing morbidities complicating its recognition and treatment, leading to short‐ and long‐term adverse clinical outcomes. There are limited data on the incidence of AKI in Australia using the Kidney Disease Improving Global Outcomes creatinine‐based consensus definition. Aim To determine the incidence and estimate the extent of under‐reporting of AKI in four hospitals in the South‐Eastern Sydney/Illawarra regions of New South Wales, Australia. Method A laboratory algorithm based on the Kidney Disease Improving Global Outcomes creatinine‐based definition for AKI was applied retrospectively to laboratory data for adult patients admitted to the study hospitals between 2009 and 2013 to identify those with AKI. The results were compared with the incidence of AKI based on diagnostic codes for AKI reported for the same period. Results AKI was detected in 12.4% of all hospitalisations (46 101/370 969) and 16.4% of patients (31 448/192 133) across the 5‐year study period using the laboratory algorithm. Of these, 72.1% were AKI Stage 1 (33 246/46101). AKI was coded in only 15.9% of hospitalisations with AKI Stage 1 (5294/33 246), 38.5% of hospitalisations with Stage 2 (2381/6185), and 46.8% with Stage 3 (3120/6670). Yearly incidence of laboratory‐identified AKI trended downward between 2009 and 2013, while annual incidence determined by coding trended upward. Conclusion Although coding trends suggested a continuous increase in clinician awareness of AKI across the study period, AKI in hospitalised patients remained significantly under‐reported.