Alveolar Dead Space Ventilation Ratio as an Early Predictor of Acute Respiratory Distress Syndrome in Severe Sepsis and Septic Shock Patients

ABSTRACT

PURPOSE: Examine the clinical utility of the alveolar dead space ventilation ratio (VdA/VT) as a predictor of acute respiratory distress syndrome (ARDS) in severe sepsis and septic shock patients. METHODS: A prospective observation study was done for 113 patients with severe sepsis and septic shock seen at the emergency department of a university hospital from January 2005 to June 2007. Therapies in the emergency department included central venous access, antibiotics, fluid resuscitation, mechanical ventilation, vasopressors and inotropes as required. The major outcome assessed was the development of ARDS within 3 days after admission. Hemodynamic variables, arterial blood gas values, serum lactate concentration, and estimated VdA/VT were evaluated at presentation (0 hour) and at 4 hours. Briefly the estimated VdA/VT was calculated by dividing the deference of the arterial CO2 and end-tidal CO2 by the PaCO2 value. Data were presented as median+/-SD. RESULTS: ARDS developed in twenty-two patients (<24 hours 17 persons, 24~48 hour 4 persons, 48~72 hour 1 person). Patients who developed ARDS had significantly higher age, higher frequency of pneumonia, greater use of mechanical ventilation and dubutamine during ED therapy, and higher sepsis related organ failure assessment (SOFA) scores. The in-hospital mortality of patients with ARDS was significantly higher than that of patients without ARDS (54.5% vs. 15.4%, p<0.001). Pneumonia, use of dobutamine during ED therapy, and VdA/VT at 4 hours were independent predictive factors for the development of ARDS. The area under the receiving operating characteristic curve for predicting ARDS was 0.891 (95% CI; 0.808-0.980) with a value of VdA/VT at 4 hours. The cut off value of VdA/VT at 4 hours was 0.25 (sensitivity 81.8%, specificity 93.3%). At 4 hours, patients with VdA/VT equal to or greater than 0.25 under resuscitation showed a high rate of fluid and high inhospital mortality when compared with patients with VdA/VT <0.25 (CVP<10 cmH2O; 37.5% vs. 16.9%, p=0.047, mortality; 75.0% vs. 4.5%, p<0.001). In patients with VdA/VT equal to or greater than 0.25 at 0 hour, patients without ARDS showed significantly improvement of VdA/VT at 4 hours. CONCLUSION: VdA/VT was found to be an independent predictive variables for ARDS in the early in-hospital period. Improvement of VdA/VT through early goal directed therapy in emergency department may decrease the development of ARDS in severe sepsis and septic shock patients.