Transpulmonary thermodilution is increasingly used for hemodynamic monitoring of critically ill patients. Injection of a cold saline bolus in the central venous circulation is a prerequisite for transpulmonary thermodilution measurements. Superior vena cava access is typically used for injection. This access, however, is not feasible or available in all intensive care patients (e.g., in burn victims or due to contraindications for Trendelenburg position). The present study investigates whether femoral vein access can be used to obtain clinically acceptable values.Design:
Open prospective trial performed between September 2005 and April 2006.Settings:
Medical intensive care unit at a university hospital.Patients:
Eleven critically ill patients monitored by transpulmonary thermodilution.Interventions:
None.Measurements and Main Results:
A total of 44 measurements in 11 intensive care patients were performed with the Pulsion PICCO Plus device to compare cardiac output, extravascular lung water index, and global end-diastolic volume index after central venous injection of the cold saline bolus via femoral and jugular venous access. Bland–Altman analysis revealed that catheter insertion site does not relevantly influence cardiac output and extravascular lung water index. The bias between femoral and jugular injection was +0.16 L/min for cardiac output and +0.23 mL/kg for extravascular lung water index. Global end-diastolic volume index values, however, show a constant overestimation of +140.73 mL/m2 after femoral injection, as obtained by Bland–Altman analysis. This overestimation can be explained by a longer mean transit time due to a longer distance of catheter tip and right atrium for a femoral catheter.Conclusions:
Transpulmonary thermodilution measurements with a cold saline bolus via a femoral catheter provide clinically reliable cardiac output and extravascular lung water index values. Concerning global end-diastolic volume index, there is a good correlation as well, but in the interpretation of the results, an overestimation has to be taken into account.