SESSION TITLE: Critical Respiratory Care SESSION TYPE: Original Investigations PRESENTED ON: 10/08/2018 07:30 AM - 08:30 AM PURPOSE: To identify risk factors of maternal mortality in pregnant patients under mechanical ventilation in six intensive care units in Colombia (ICU). METHODS: A retrospective observational study conducted in 6 hospitals in Colombia, South America, including all obstetric patients admitted to the ICU who required mechanical ventilation for more than 24 hours between 2006 and 2016. We reviewed medical records, demographic data, but also clinical and laboratory variables. Patients were divided into survivors and non-survivors for statistical analysis. The categorical variables were presented in absolute numbers and percentages. Continuous variables, using mean and standard deviation or median and interquartile range, as appropriate. To determine the relationship between the categorical variables, we used the chi-square test or Fisher's exact test. P values less than 0.05 were considered statistically significant. The association between ventilation and mortality with clinical variables, interventions and organ dysfunctions was estimated by logistic regression analysis. A discrimination analysis was performed by area under the curve (AUC) for continuous clinical and laboratory variables, evaluating their ability to predict mortality. The statistical analysis was performed using Statistical (STATA 14). RESULTS: There were 2116 ICU admissions during the study period, including 299 (14%) pregnant women with acute respiratory failure and mechanical ventilation requirement. The median age was 26 years (SD 7,2), with an average gestational age of 30.5 weeks (IQR 6 – 41). Main diagnoses in pregnant patients under mechanical ventilation were grouped as follows: obstetric hemorrhage (n = 97, 32.44%), hypertensive disorders of pregnancy (n = 85, 28.42%), maternal sepsis (n = 80, 26.75%) and other causes (n=37, 12,37%). Variables related with higher mortality after the multivariate analysis were: vasopressor requirement (OR: 2.81, 95% CI 1.07 - 7.39), transfusions (OR: 2.26, 95% CI 1.14 - 4.48), neurological dysfunction (OR: 2.81, 95% CI 1.12 - 7.04), coagulopathy (OR: 2.71, 95% CI 1.30 - 5.62) and ARDS (2.58, 95% CI 1.22 - 5.49). Nulliparity was associated with a lower risk of the death (OR: 0.31, IC95% 0.16 - 0.60). The clinical variables that showed an acceptable discrimination capacity as predictors of mortality were the Mean Blood Preassure (AUC 0.69 IC 95% 0.63 - 0.74) and pH (AUC 0.65 IC 95% 0.59-0.71). CONCLUSIONS: The mortality related to mechanical ventilation in obstetric patients has the same associated factors as in the non-obstetric population. ARDS, the presence of coagulopathy and neurological dysfunction are clinical factors associated with mortality; transfusions and use of vasopressors are also related to worse outcomes. CLINICAL IMPLICATIONS: A retrospective study that provides information on the prognostic factors associated with mortality in pregnant patients undergoing mechanical ventilation. DISCLOSURES: No relevant relationships by Camilo Bello Muñoz, source=Web Response No relevant relationships by Diana Borre, source=Web Response recipient of research equipment support relationship with respironics Please note: $20001 - $100000 Added 11/30/2017 by Ghada Bourjeily, source=Web Response, value=Grant/Research Support No relevant relationships by Francisco Camargo-Assis, source=Web Response No relevant relationships by Arturo Cardona, source=Web Response No relevant relationships by ALEJANDRO CASTRO-SANGUINO, source=Web Response No relevant relationships by Carmelo Dueñas-Castell, source=Web Response no disclosure on file for Oscar Lavalle; No relevant relationships by Carlos López-Arrieta, source=Web Response No relevant relationships by HUGO ANDRES PEREZ RAMON, source=Web Response No relevant relationships by Tomas Rodriguez Yanez, source=Web Response No relevant relationships by Jose Rojas-Suarez, source=Web Response