Figure 3. Mean data (+/- SEM) on systolic (Ees [upper]) and diastolic (Kc [lower]) function at the different times of measurement (before the start of cardiopulmonary bypass [CPB; Baseline]; immediately after separation from CPB [time 0]; and 5 min [time 5], 10 min [time 10], and 15 min [time 15] later). Data of the two protocols are shown. In protocol 1 (filled circles), preload was optimized just before the start of the separation procedure from CPB. In protocol 2 (filled squares), preload was optimized 10 min before the anticipated separation time and kept constant during the period before the separation procedure was started. Duration of the separation procedure was similar in both groups. In protocol 1, ventricular function after separation from CPB was impaired, with a decrease in Ees and an increase in Kc, and a progressive improvement of ventricular function within the first 10 min after separation from CPB. These observations were not present in protocol 2, in which ventricular function remained unchanged and comparable to baseline values. *Difference between both groups is statistically significant at P < 0.01.