The role of an electronic alert system to detect acute kidney injury in hospitalized patients: DETECT-H Project

Abstract Background and aims Acute kidney injury (AKI) is associated with higher mortality and length of stay (LOS) for hospitalized patients. To improve outcomes, an electronic detection system could be a useful tool for early diagnosis. Methods A fully automated real-time system for detecting decreased glomerular filtration rate in adult patients was developed in our hospital, DETECT-H project. AKI was established according to KDIGO guidelines. Results In six months, 1241 alerts from 11,022 admissions were issued. Overall incidence of AKI was 7.7%. Highest AKI stage reached was: stage 1 (49.8%), 2 (24.5%) and 3 (25.8%), in-hospital mortality was 10.9%, 22.7%, 33.9% respectively and 57.1% in AKI requiring dialysis; mortality in stable CKD was 4.3%. Median LOS was 8 days versus 5 days for all patients. AKI was associated with a mortality of 3.18 (95% CI 1.80–5.59) and a LOS 1.52 (1.11–2.08) times as high as that for admissions without AKI. Multivariate analysis indicated that a LOS higher than 8 days was associated with AKI. Previous CKD was noted in 31.9% and AKI in 45.3% at discharge. As compared to the use of the detect system, only one third of CKD patients and half of AKI episodes were identified. Conclusions CKD and in-hospital AKI are under-recognized entities. Mortality and LOS are increased in-hospital patients with renal dysfunction. AKI severity was associated with higher mortality and LOS. An automated electronic detection system for identifying renal dysfunction would be a useful tool to improve renal outcomes.