ICU Open-Access Databases for Artificial Intelligence in Sepsis: Balancing Innovation With Data Quality Challenges

Q: What does this research mean for the field?

Major open-access ICU databases exhibit substantial data quality limitations, including missingness, redundancy, and semantic variability, which necessitate rigorous preprocessing for valid AI-driven sepsis research. Novelty: ClaimNovelty.INCREMENTAL. Consensus alignment: ConsensusAlignment.NEUTRAL.

Joanne Jegou; Ioana Manolescu; Stéphane Ruckly; Jean-François Timsit; Michaël Thy

doi:10.1097/ccm.0000000000007319

August 24, 2026Critical Care Medicine

ICU Open-Access Databases for Artificial Intelligence in Sepsis: Balancing Innovation With Data Quality Challenges

View at Publisher

Ask AI

Bookmark

Share

Key Points

To evaluate the suitability of three major open-access ICU databases (MIMIC-IV, eICU-CRD, and AmsterdamUMCdb) for AI-driven sepsis research, focusing on data completeness, consistency, standardization, and usability.
Analyzed records from MIMIC-IV (2008–2019), eICU-CRD (2014–2015), and AmsterdamUMCdb (2003–2016) using structured queries and metadata inspection.
Quantified missing data, semantic inconsistencies, typographical variations, and structural redundancies across diagnoses, antibiotic therapy, microbiology, and mortality markers.
MIMIC-IV showed extensive missingness in medication administration route (49%) and dose (58%), 99.9% missing fields in microbiology quantity, and 174 near-duplicate diagnostic variants across 17,557 unique strings.
eICU-CRD lacked diagnoses in 1.9% of admissions with mortality discordance in 7,319 stays, whereas AmsterdamUMCdb lacked diagnoses in 61.4% of admissions alongside mixed English-Dutch terminology.
All three databases lacked reliable sepsis diagnosis timestamps, severely limiting time-sensitive evaluations of early clinical interventions.

No abstract yet

We only have this paper's title and publication details.

Read it at the source

Ask about this paper

Answers are grounded in this paper and cite their sources.

Implication

Comparative analysis reveals widespread data missingness and semantic inconsistency across open-access ICU databases, highlighting the need for rigorous preprocessing in sepsis AI research.

Discussion

No takes yet. Share an insight, caveat, or question.

Cite This Study

Jegou et al. (2026) studied this question.

synapsesocial.com/papers/6a8e9b43451774b83f3b3f6e https://doi.org/10.1097/ccm.0000000000007319

View at Publisher

Ask AI

Bookmark

Share

ICU Open-Access Databases for Artificial Intelligence in Sepsis: Balancing Innovation With Data Quality Challenges

Key Points

No abstract yet

Ask about this paper

Implication

Discussion

Cite This Study

ICU Open-Access Databases for Artificial Intelligence in Sepsis: Balancing Innovation With Data Quality Challenges

Key Points

No abstract yet

Ask about this paper

Implication

Discussion

Cite This Study

Keep reading

Keep reading