Summary

Background: Acute kidney injury remains a common significant clinical problem. Yet there are scant data in Israel on the incidence of hospital-acquired AKI[1] and on diagnosis validity.

Objectives: To describe the epidemiology of AKI among hospitalized patients in the Western Galilee Hospital, Nahariya, compare discharge summaries to laboratory diagnosis, and investigate the impact of AKI on mortality and length of stay.

Methods: Computerized medical and laboratory data of 34,802 hospitalized subjects were collected. AKI was diagnosed according to three different definitions. We calculated the sensitivity and specificity of AKI based on ICD-9 diagnosis compared to patient's laboratory data as the gold standard.

Results: The overall AKI annual incidence rate was 1–5.1%, depending on the AKI definition used. The incidence of AKI based on ICD-9 diagnosis was significantly lower compared to the laboratory-based diagnosis. Average in-hospital length of stay was 2.4 times longer among patients with AKI compared to subjects without this condition. Furthermore, the in-hospital death rate among AKI patients was 14 times higher than among non-AKI hospitalized subjects, with a positive association between AKI severity and risk of death.

Conclusions: Using AKI laboratory diagnosis as the gold standard revealed ICD-9 diagnosis to be 9.1% sensitive and 99.4% specific. Hospital-acquired AKI is a major contributor to prolonged length of stay and high mortality rates; therefore, interventions to reduce in-hospital disease incidence are required.