An Early Warning System for the Differential Diagnosis of In-Hospital Acute Kidney Injury for Better Patient Outcome: Study of a Quality Improvement Initiative

Abstract

Background: Acute kidney injury (AKI) is a syndrome with heterogeneous causes and mechanisms. An early warning system (EWS) for AKI was created to reduce the incidence and improve outcomes. However, the benefits of AKI-EWS remain debatable. Methods: We launched a project to design and create AKI-EWS for inpatients in our institute. Incidence of AKI and its outcome before and after the implementation of AKI-EWS were collected for analysis. Results: We enlisted a stakeholder map before creating AKI-EWS. We then started an action plan for this initiative. The diagnosis was automatic and based on the definition of Kidney Disease: Improving Global Outcomes (KDIGO). The differential diagnosis of causes of AKI was also automatic. Users are to adjust the threshold of detection. After the implementation of this AKI-EWS, the incidence of AKI fell. The proportion of AKI > 4% was reduced significantly (47.7% and 41.6%, p = 0.010) in patients with serum creatinine measured. The proportion of AKI > 0.9% also dropped significantly (51.67% and 35.94%, p = 0.024) in all inpatients. Trends of AKI outcomes also showed improvement. The loading of consultation of nephrologists decreased by 15.5%. Conclusions: Through well-designed AKIEWS, the incidence of AKI dropped, showing improved outcomes. The factors affecting benefits from AKI-EWS included high-risk identification (individual threshold detection), timely and automatic diagnosis, real-time alerting on electronic health information systems, fast self-diagnosing of the cause of AKI, and coverage of all inpatients.