Illness severity modifies the association between sepsis and survival in critically ill patients with acute kidney injury.

backgroundSepsis and acute kidney injury (AKI) are common, often co-occurring ICU syndromes with high mortality. Although illness severity scores such as the Simplified Acute Physiology Score II (SAPS II) are powerful prognostic tools, it is unclear whether baseline illness severity modifies the relationship between sepsis and mortality in AKI. We hypothesized that this association varies across the severity spectrum.

methodsIn a retrospective cohort of 35,926 adults with AKI, we examined the interaction between sepsis (Sepsis-3 criteria) and baseline illness severity (SAPS II) in relation to 30-day survival. We used multivariable logistic regression adjusted for demographics, comorbidities, and organ dysfunction.

resultsA significant interaction between sepsis and illness severity was observed (p < 0.001). At low severity (SAPS II = 20), sepsis was associated with a 4.3 percentage point reduction in 30-day survival, whereas at high severity (SAPS II = 90), it was associated with a 23.6 percentage point increase in survival, representing a net reversal of 27.9 percentage points. This pattern was robust in sensitivity analyses.

conclusionsIn critically ill patients with AKI, the impact of sepsis on survival is strongly modified by baseline illness severity, reversing from harm at low severity to improved survival at high severity. This paradox challenges the assumption of uniform sepsis risk and suggests that, in the sickest patients, sepsis may indicate a state of potentially reversible organ dysfunction. Incorporating illness severity into AKI prognostication may better capture recovery potential, as severity scores alone may underestimate the likelihood of recovery in severe septic AKI.