AUTHOR=Moor Michael , Rieck Bastian , Horn Max , Jutzeler Catherine R. , Borgwardt Karsten 

TITLE=Early Prediction of Sepsis in the ICU Using Machine Learning: A Systematic Review

JOURNAL=Frontiers in Medicine

VOLUME=Volume 8 - 2021

YEAR=2021

URL=https://www.frontiersin.org/journals/medicine/articles/10.3389/fmed.2021.607952

DOI=10.3389/fmed.2021.607952

ISSN=2296-858X

ABSTRACT=Background: Sepsis is among the leading causes of death in intensive
care units (ICU) worldwide; its recognition, particularly in the
early stages of the disease, remains a medical challenge.
The existence of digital health data has created a setting in
which machine learning can be used for digital biomarker discovery, with
the ultimate goal to advance early sepsis recognition.

Objective: To systematically review and evaluate studies employing
machine learning for the prediction of sepsis in the ICU.

Data sources: Using Embase, Google Scholar, PubMed/Medline, Scopus, and
Web of Science, we systematically searched literature for
machine learning-driven sepsis onset prediction for ICU patients.

Study eligibility criteria: All peer-reviewed articles using machine
learning for the prediction of sepsis onset in adult ICU patients were
included. Studies focusing on non-ICU patients were excluded.

Study appraisal and synthesis methods: A systematic review was performed
according to the PRISMA guidelines, followed by a quality assessment of
all eligible studies.

Results: Out of 974 identified articles, 22 and 21 met inclusion
criteria for the systematic review and quality assessment, respectively.
Numerous machine learning algorithms were applied to refine the
early prediction of sepsis. The quality of the studies ranged from
"poor" (satisfying ≤40% of the quality criteria) to "very good" (satisfying
≥ 90%). The majority of the studies (n=19) employed an
offline training scenario combined with horizon evaluation, while two
studies implemented an online scenario. Massive inter-study
heterogeneity in terms of models, sepsis definitions,
prediction time windows, and outcomes precluded a meta-analysis.
Only 2 studies provided source code and data sources.

Limitations: Articles were only eligible for inclusion when employing
machine learning algorithms for the prediction of sepsis onset in the
ICU. This restriction led to the exclusion of studies focusing on the
prediction of septic shock, sepsis-related mortality, and patient
populations outside the ICU.

Conclusions and key findings: A growing number of studies employs
machine learning to optimise the early prediction of sepsis through
digital biomarker discovery. This review highlights several
shortcomings of current approaches, including low comparability and
reproducibility. We gather recommendations how these
challenges can be addressed before deploying models in prospective
analyses.