To identify urine-output thresholds associated with mortality equivalent to creatinine-defined KDIGO AKI stages.
Approach:
Cohort: Reanalysis of 5,698 ICU admissions involving 4,999 adults without pre-existing chronic kidney disease at one Swedish general ICU.
Analysis: Minimum sustained urine output was assessed over overlapping 3-, 6-, 12-, and 24-hour intervals. Age- and sex-adjusted models estimated its association with 14-day mortality.
Key Findings:
Under current KDIGO urine-output criteria, mortality was underestimated by 37% at stage 1 and 48% at stage 2, and overestimated by 34% at stage 3, relative to creatinine-defined staging.
Mortality-equivalent urine-output thresholds were identified for each creatinine-defined stage over 3-, 6-, and 12-hour intervals. For 24-hour intervals, thresholds were identified for stages 1 and 2, but not stage 3.
All identified thresholds were below the current stage 1 cutoff of 0.5 mL/kg/h.
Urine output below 0.24 mL/kg/h sustained for 3 hours was associated with mortality equivalent to creatinine-defined stage 1 AKI.
Interpretation:
The authors report discordance between current urine-output staging and creatinine-defined mortality risk. The proposed thresholds are prognostic, not causal.
Limitations:
The analysis involved a single tertiary-care ICU and requires validation in other settings and ICU populations.
Adjustment included only age and sex; residual confounding cannot be excluded.
Diuretic use, fluid status, and continuous renal replacement therapy were not considered.
No corresponding analysis of tissue-injury markers was conducted.
Conclusion:
The study identifies urine-output thresholds aligned with creatinine-defined mortality risk and reports that current criteria underestimate mortality at stages 1 and 2 and overestimate it at stage 3. The authors present the thresholds as a reference for future validation.
New multisociety guidance identifies menstrual irregularities as potential early markers of cardiovascular risk, extending prevention efforts into adolescence.
