RELATION OF ECHOCARDIOGRAPHIC PRELOAD INDEXES TO STROKE VOLUME IN CRITICALLY-ILL PATIENTS WITH NORMAL AND LOW CARDIAC INDEX

Objective: To examine the usefulness of preload indices obtained by tr ansoesophageal echocardiography (TOE) for estimating stroke volume at various levels of cardiac index. Design: Prospective clinical study. S etting: Intensive care unit with surgical patients. Patients: 16 venti lated patients monitored via Swan-Ganz catheterization and TOE. Interv entions: Echocardiographic images of left ventricular cross-sectional short-axis areas were analysed for the preload indices end-diastolic a rea (EDA), stroke area and end-diastolic wall stress. The relation bet ween these indices and stroke volume, calculated from thermodilution c ardiac output, was analysed in all patients and in nine patient groups discriminated by various ranges in heart rate (less than or equal to 70 to > 110 beats/min), pulmonary artery occlusion pressure (less than or equal to 8 to > 12 mmHg) and cardiac index (less than or equal to 3.0 to > 4.2 l/min per m(2)). Measurements and results: Overall stroke volume (n = 155) correlated significantly (p < 0.0001) with EDA (r = 0.89) and stroke area (r = 0.80). The correlation with end-diastolic w all stress was non-significant (r = 0.51). Linearity in the relation b etween stroke volume and EDA or stroke area was independent of variati ons in heart rate and pulmonary artery occlusion pressure. Stroke volu me correlated well with EDA and stroke area, when cardiac index was no rmal or high, but the relation slightly deteriorated (r = 0.63 to less than or equal to 0.72) when the cardiac index was low. Changes in EDA and stroke area by more than 1, 2 or 3 cm(2) were weak predictors for changes in stroke volume greater than 20 %. Conclusions: Stability of the relation between echocardiographic preload indices and stroke vol ume emphasize the potential of TOE for continuous preload monitoring i n the critically ill.