Objective: An investigation into the incidence and the clinical implication of discrepancies which may sometimes occur between invasive and non-invasive hemodynamic evaluation in septic patients.
Design: A prospective, consecutive comparison.
Setting: Department of Intensive Care Medicine at a University Hospital.
Patients: 32 patients undergoing therapy for an episode of septic shock.
Interventions: Conventional hemodynamic support (including volume expansion in all cases and inotropic support if necessary) required to obtain a stable hemodynamic status.
Measurement and results: Cardiac output (thermodilution method), cardiac pressures (right heart catheterization) and left ventricular (LV) volumes (two-dimensional echocardiography) were simultaneously recorded. A comparison was thus made between both procedures, particularly concerning preload evaluation and assessment of left ventricular systolic function. Pulmonary artery occlusion pressure measurement was evidence as an unreliable index of LV end-diastolic volume, determining preload. Assessment of LV systolic function by both methods was conflicting in 11 cases out of the 32.
Conclusions: Frequent discrepancies between to invasive and non-invasive procedure were observed. The reasons for these discrepancies, including low vascular resistance, reduced LV compliance, and a possible overestimation of cardiac output by the thermodilution method, are examined in the light of data recorded. It was concluded that invasive hemodynamic evaluation by right heart catheterization in septic patients should be seriously questioned.