Prognostic factors and utility of scoring systems in patients with hematological malignancies admitted to the intensive care unit and required a mechanical ventilator.

RATIONALE: Mortality of patients with hematological malignancy requiring mechanical ventilation is high. Neither early prognostic indicators nor scoring systems that discriminate survivors from non-survivors to aid in end-of-life decision making have been identified.

OBJECTIVE: To assess the outcomes, prognostic factors, and scoring systems of acute respiratory failure requiring endotracheal intubation in the intensive care unit (ICU) in patients with hematological malignancies.

METHODS: Retrospective cohort study in the medical ICU of a tertiary hospital. Thirty-three critically ill patients with hematological malignancies requiring mechanical ventilation were analyzed for demographic data, ICU survival, type of malignancy state of disease, reasons for hospitalization and ICU admission, peripheral blood parameters and scoring systems (APACHE II, SOFA, SAPS II, and MODS) during ICU stay All recorded variables were evaluated for prognostic relevance by univariate and multivariate analyses.

MEASUREMENTS AND RESULTS: Overall ICU mortality was 74%. Univariate analysis revealed statistically significant differences in red blood cell count, hemoglobin, mean arterial pressure, coagulation studies, as well as the presence of oliguria, multi-organ failure, vasopressor requirement, pneumonia, blood product requirement. APACHE II, SOFA, and SAPS II scales also revealed similar statistical significance in outcome. However, multivariate analysis did not reveal any independent prognostic factors statistically. Among these, hemoglobin level appears to be the strongest trend (p = 0.0577) for survival.

CONCLUSIONS: Mortality of patients with hematological malignancies requiring mechanical ventilation remains high. No single independent risk factor for ICU mortality was identified with multivariate logistic regression analysis. Prognostic scoring systems do not yield adequately reliable information to be used exclusively for end-of-life decision making in individual patients.