Scientific Research Project Number: MA 2021 05
Place: Amsterdam UMC, location AMC, department of Medical Informatics
Comparison of models to diagnose AKI on ICU data
Acute kidney injury (AKI) is a common and potentially life-threatening condition that affects approximately one in five inpatient admissions in the US. Its occurrence is particularly frequent in the intensive care unit (ICU) patients. Thus, prediction of this condition in the ICU setting is of high relevance. Clinically detection of AKI uses serum creatinine increase as a marker of acute decline in renal function. The lag of such an increase behind the renal injury therefore results in delayed treatment and preventative alerts could therefore empower clinicians to act before a major clinical decline has occurred.
Three models should be considered: the VUmc model (deployed in the UMC EHR based on EPIC), the NHS AKI model, used nationwide in the UK, and the KDIGO scanner, a model solely based on the KDIGO guidelines developed at the AMC dept. of medical informatics.
A first step will consist in comparing how the three models define and implement rules to determine AKI, while the second step will be assessing how different the model perform in identifying AKI patients on ICU data in terms of false positives (i.e. patient identified as having AKI while they do not) and false negatives (patient having AKI not identified), considering the KDIGO guidelines the gold standard. To perform the latter step, models may be reproduced (i.e. developed) in a controlled environment.
1. Do the different models preselected, define and implement AKI in a similar manner? Which are the similarities and which are the differences?
2. How well can the NHS and VUmc correctly identify AKI patients compared to the KDIGO standard?
A deep analysis in the similarities and differences in defining and implementing rules to detect AKI among the three models.
Three models which accurately reproduce the NHS, VUmc, and KDIGO algorithms, respectively.
A comparison of models in identifying AKI patients on ICU data in terms of false positives and false negatives.
A master thesis written in a form of a scientific article.
Mentor: Iacopo Vagliano, Amsterdam UMC, location AMC, department of Medical Informatics, email@example.com
1. Hill R, Selby N. Acute Kidney InjuryWarning AlgorithmBest Practice Guidance. https://www.thinkkidneys.nhs.uk/wp-content/uploads/2014/12/AKI-Warning-Algorithm-Best-Practice-Guidance-final-publication-0112141.pdf
2. Kidney Disease Improving Global Outcomes KDIGO Acute Kidney Injury Work Group KDIGO clinical practice guideline for acute kidney injury. (2012). Kidney Int. suppl 2:1–138. https://kdigo.org/guidelines/acute-kidney-injury
Iacopo Vagliano, Amsterdam UMC, location AMC, department of Medical Informatics, firstname.lastname@example.org