AI Science Tracker
EN

Learned sepsis groups show different observed hospital mortality in an ICU cohort

Wuhan University; Zhongnan Hospital of Wuhan University · 25 Aug 2022 · biology

Discovery

Learned sepsis groups show different observed hospital mortality in an ICU cohort

Source check: Source checked. Not independent scientific confirmation.

Finding: K-means clustering of eleven routine first-day measurements grouped 8,817 sepsis records into two clinical patterns. Observed hospital mortality differed between groups and remained associated after statistical adjustment. This is a retrospective cohort relationship, without established biological subtypes, causal treatment effects or validated bedside deployment.

Technology: Unsupervised K-means clustering of routine ICU clinical measurements

Publication: Peer-reviewed paper · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.

AI contribution
K-means learned multivariable patient groups without using mortality as its clustering target; subsequent outcome analysis linked those learned groups to observed hospital mortality. PCA visualization alone and recovery of known individual severity markers are not the represented result.
Test and result
The source selected 8,817 MIMIC-IV sepsis patients and used eleven expert-selected variables measured during the first ICU day. It compared two-to-eight clusters and selected two using clustering criteria. Hospital mortality was 29.4% in the smaller group versus 8.5% in the larger group; adjusted odds ratio was 2.214, 95% CI 1.780–2.754. Outcome differences are actual cohort observations beyond cluster-separation scores.
Limitation / still unknown
The groups were derived and compared in the same retrospective single-center database, without external transportability testing. They may reflect severity and clinical heterogeneity rather than distinct biological entities. Missing lactate and oxygenation values required imputation; absent baseline SOFA and unmeasured infection characteristics may bias case definition and associations. No causal therapy-response difference, prospective prediction or operational benefit was shown.
Evidence sources and annotation context

Evidence context (descriptions follow the cited sources; test type is a catalog classification):

Evidence sources: https://link.springer.com/article/10.1007/s40121-022-00684-y

Source checked · editorial details

The source page was checked against this entry. Independent confirmation is not assessed by this label.

Source type: paper. Paper: not linked in this catalog; this does not mean no paper exists.

Review date: 21 Nov 2026. Review scope: Read actual primary sepsis eligibility, variable selection/imputation, K-means cluster-number selection and adjusted-outcome Methods, group mortality Results and limitations. Supplementary implementation and individual records not independently reconstructed. Exact original article front DOI/title and fresh PubMed primary relationship metadata checked..

No follow-up recorded; this is not a claim that no follow-up exists. Follow-up review scope and date are not recorded. No source-supported relationships recorded.

Editorial assessment: Impact 1/5 · Validation 2/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.

Open in the full catalog for filters, charts and related discoveries. Browse biology discoveries.