Urine Output as a Novel Predictor for In-Hospital Mortality in Acute Pulmonary Embolism Patients: Training With the MIMIC Database and Validation With Independent Cohort

Abstract

Background: Identifying high-risk patients with acute pulmonary embolism is vital for improving disease prognosis. However, current guidelines and research on risk factors are insufficient to meet clinical needs. This study was aimed at exploring novel risk factors to predict in-hospital mortality.

Methods: We utilized a patient cohort from the Medical Information Mart for Intensive Care Version IV (MIMIC-IV) database as training cohort. Major analyses included screening risk factors for in-hospital mortality, correlation analysis via smooth curve fitting, multivariate Cox regression, and subgroup analysis. The findings were further validated with our own institute patient cohort.

Results: Among 1463 adult patients with acute pulmonary embolism in the MIMIC-IV database, the overall in-hospital mortality rate was 17.8%. A nonlinear correlation was observed between urine output and in-hospital mortality. A urine discharge less than 0.85 mL/kg/h was used as the threshold and was negatively associated with the risk for in-hospital death. Compared to patients with urine value < 0.5 mL/kg/h, the risk for in-hospital mortality reduced by 36% and 48% in patients with urine values of 0.5–0.85 mL/kg/h and > 0.85 mL/kg/h, with the hazard ratios of 0.64 (0.47, 0.87) and 0.52 (0.38, 0.72), respectively. This association remained significant in the subgroup analysis after adjusting for age, gender, hypotension, and low oxygen saturation. Our validation patient cohort (n = 151) further confirmed the strong association of the urine value with in-hospital mortality and consistent cutoff value.

Conclusion: Our study revealed a negative association of urine output with in-hospital mortality in acute pulmonary embolism patients, with the optimal urine output being significantly higher than the value of other critical illnesses.