AUTHOR=Li Le , Zhang Zhenhao , Xiong Yulong , Hu Zhao , Liu Shangyu , Tu Bin , Yao Yan 

TITLE=Relationship Between Initial Urine Output and Mortality in Patients Hospitalized in Cardiovascular Intensive Care Units: More Is Not Better

JOURNAL=Frontiers in Cardiovascular Medicine

VOLUME=Volume 9 - 2022

YEAR=2022

URL=https://www.frontiersin.org/journals/cardiovascular-medicine/articles/10.3389/fcvm.2022.853217

DOI=10.3389/fcvm.2022.853217

ISSN=2297-055X

ABSTRACT=Backgrounds: Decreased urine output (UO) is associated with adverse outcomes in certain patients, but this effect in the patients admitted for cardiovascular diseases is still unproven. Moreover, the relationship between increased UO and prognosis is also unclear.
Objective: To investigate the relationship between decreased or increased UO and outcomes in patients of cardiovascular intensive care unit (CICU).
Methods: This study was a retrospective cohort analysis based on the medical information mart for intensive care III (MIMIC-III) database. Patients data were extracted from the Beth Israel Deaconess Medical Center (Boston, MA) between 2001 and 2012. With the initial 24-hour UO range from 0.5 to 1.0 mL/kg/h as the reference, participants were divided into several groups. The primary outcome was 30-day mortality. The secondary outcomes were 90-day mortality, ICU mortality, hospital mortality, use of mechanical ventilation (MV) and vasopressor agents in the first 24-hour of ICU. The association between UO and mortality was assessed by a multivariable logistic regression.
Results: A total of 13279 patients admitted to CICU were included. Low UO (< 0.5 mL/kg/h) was strongly associated with 30-day mortality (Unadjusted OR = 3.993, 95% CI: 3.447 to 4.625, p < 0.001), and very high UO (≥ 2.0 mL/kg/h) was also a significantly risk factor for 30-day mortality (Unadjusted OR = 2.069, 95% CI: 1.701 to 2.516, p < 0.001) compared with the reference. The same effects also were shown in the multivariable logistic regression, adjusted by age, gender, vital signs, common comorbidities and use of diuretics, with adjusted OR of 2.023 (95% CI: 1.693 to 2.417, p < 0.001) for low UO and 1.771 (95% CI: 1.389 to 2.256,	p < 0.001) for very high UO. Moreover, both decreased UO and increased UO were risk factors for 90-day mortality, ICU mortality, hospital mortality, use of MV and vasopressor agents.
Conclusion: The decreased and increased UO both were significantly associated with short-term mortality, the relationship between UO and mortality was U-shape rather than linear.

Key words: urine output, mortality, cardiovascular critical care, logistic regression, Lowess smoothing.