Study Identifies Two Sepsis Kidney Injury Subtypes

Researchers found a high-risk group linked to higher 28-day mortality in both a Chinese cohort and MIMIC-IV.

Updated on Nov. 19, 2026 in Diseases — General

Study Identifies Two Sepsis Kidney Injury Subtypes

Researchers identified a two-class clinical signature of sepsis-associated acute kidney injury in a cohort of 282 adults from Zigong, China, and replicated a similar partition in MIMIC-IV, which included 14,949 adults. The high-risk class showed higher 28-day mortality in both cohorts.

Why it matters

Sepsis-associated acute kidney injury is a serious complication, and separating patients into distinct subphenotypes could help clinicians recognize those at greatest risk early. The modest gain in discrimination when the subphenotype was added to age, sex and SOFA suggests it may complement, rather than replace, existing severity measures.

The study compares a 282-adult discovery cohort from Zigong with 14,949 adults in MIMIC-IV, reporting an adjusted 28-day mortality hazard ratio of 2.65 (1.70-4.15) and 1.70 (1.57-1.84) respectively. Whether the subphenotypes perform similarly in other populations and clinical settings is not addressed by the data reported.

The details

Subphenotypes were derived by consensus clustering of eighteen first-day variables, with two classes prespecified in the discovery cohort and the number of classes left free in MIMIC-IV. The high-risk class showed higher lactate and bilirubin and lower platelet count, while peak temperature was the only discordant variable between the two cohorts.

Timeline

  1. Article published as an accepted peer-reviewed research preprint on 2026-11-19.

Health Landscape

The study uses the MIMIC-IV critical care database as an external validation cohort, testing whether a subphenotype found in a single Chinese hospital population replicates in a large US-derived ICU dataset. Its replication across both cohorts places it within the broader effort to move sepsis care from one-size-fits-all criteria toward data-driven patient classification.

For patients and clinicians, the finding points toward a possible way to flag sepsis patients with acute kidney injury who face higher short-term risk. Any practical use would depend on further validation before the classification could inform day-to-day treatment decisions.

The takeaway

Recognizing that sepsis-associated acute kidney injury may fall into distinct risk groups could help clinicians think about how patients are stratified, though the research remains preliminary and requires further testing before influencing care.

Further reading

Explore more on Diseases — General for broader context on how disease classification shapes treatment research.

Source note: This article includes information reported by Nature.