Skip to main content

Cancer patients with sepsis: Prognostic insights from a population-based cohort study in Norway.

Cancer epidemiology, biomarkers & prevention : a publication of the American Association for Cancer Research, cosponsored by the American Society of Preventive Oncology2025-12-01PubMed
Total: 75.5Innovation: 7Impact: 8Rigor: 8Citation: 7

Summary

In a nationwide cohort of 222,832 sepsis admissions, 16.9% had cancer and experienced substantially higher in-hospital mortality, especially with metastatic disease. Relative risks were highest among younger adults, females, and Gram-negative sepsis, emphasizing cancer-type- and metastasis-specific prognostication.

Key Findings

  • Among 222,832 sepsis admissions, 16.9% had cancer; in-hospital mortality was 16–17% (non-metastatic) vs ~26–27% (metastatic).
  • Adjusted RR of death vs non-cancer: non-metastatic 1.39 (men)/1.63 (women); metastatic 2.27 (men)/2.75 (women).
  • Risks were highest in younger metastatic patients and in Gram-negative sepsis.

Clinical Implications

Prioritize early aggressive management in metastatic and high-risk cancer subgroups, ensure Gram-negative coverage when appropriate, and integrate cancer-specific variables into sepsis triage and prognostic tools.

Why It Matters

The study provides precise, population-level estimates of sepsis mortality risk stratified by cancer type, metastasis, age, sex, and pathogen class, directly informing oncology–critical care risk models and resource allocation.

Limitations

  • Administrative coding may misclassify sepsis and cancer status; limited clinical granularity (e.g., SOFA, treatment details)
  • Observational design cannot eliminate residual confounding

Future Directions

Integrate clinical severity scores and treatment variables to refine cancer-specific sepsis prognostic models; evaluate targeted pathways for Gram-negative sepsis in oncology populations.

Study Information

Study Type
Cohort
Research Domain
Prognosis
Evidence Level
III - Population-based retrospective cohort using national registry data
Study Design
OTHER