RESUMEN
Objectives: To investigate the epidemiology, management and predictors of mortality in severe sepsis. Design and Methodology: Prospective observational study in 4 Trinidadian Intensive Care Units (ICU) over a one year period August 2017-August 2018. Physiologic and treatment data was collected on admission to ICU and patients were followed up until ICU discharge and then at 28 days to determine mortality. Results: 163 patients fit the criteria for severe sepsis and were enrolled. Twenty-eight day and ICU mortality rate were 42% (68) and 34 % (56) respectively. Case distribution by the various hospitals were San Fernando General Hospital 62% (101), Port of Spain General Hospital 16% (26), Eric Williams Medical Sciences Complex 12.3% (20) and Sangre Grande Hospital 9.8% (16). The most common source of sepsis was pulmonary (54%) followed by abdominal (17%) and urological sepsis (14%). Acute Kidney Injury (AKI) was present in 71% (115) patients and 43% (46) of patients with stage 2 and above AKI received Renal Replacement Therapy (RRT) in the ICU. In the regression model, the only factors that were found to predict both 28 day and ICU mortality were stage 2 AKI [OR 6.2 (95% CI 1.7- 23.1, p = .007)], stage 3 AKI [OR 7.2 (95% CI, 2.1-24.0, p=.001)] , mean arterial pressure of < 60mmHg in the first 24 hours [OR 10.8 (95% CI 1.7-68.1, p= .001)], presence of either moderate-severe Acute Respiratory Distress Syndrome [OR 4.1 (95% CI 1.8 9.2, p = .002)] and Acute Physiology, Age, Chronic Health Evaluation (APACHE) II score [OR 1.08 (95% CI, 1.0-1.2, p=.039)]. Conclusion: Severe sepsis is associated with a high hospital mortality rate and this sepsis burden varies according to region. Limited access to RRT remains a problem in certain centers.