Correlation between changes in systolic time intervals and left ventricular end-diastolic diameter after preload reduction. Non-invasive monitoring of pharmacological intervention.