Determination of lower and upper boundaries for ScvO2: Data(n=203 initial values, two missing values for lactate) from 205 patients at time point h=0 hours were split at a ratio of 70:30into training (143 samples) and test (60 samples) sets as recommended [12].The definition of optimal lower and upper bounds for ScvO2 was solely based on the training set. The independent test set was used to assess the predictive power of the training results. We used 50 replications of 5-fold cross-validation to obtain estimators for the optimal lower and upper bounds for ScvO2. That is, the training set was split up into five random samples of nearly equal size (n=28 or 29, respectively). One of five sample sets was left aside (as the cross-validation test sample) and the remaining four parts were used for training of the classifier to predict the combined endpoint “death, need for veno-venous hemofiltration, or ventilation exceeding 24 hours”. This procedure was repeated for each of the five parts of the training set. To reduce the effect of this random split, 50 replications of 5-fold cross-validation overall leading to 250 random samples were used in addition. The initial use of numerical optimization procedures showed that there exist several local minima for the classification error. Hence, we switched to a more general procedure consisting of a grid search to find the global minimum of misclassification. A grid consisting of values ranging from 19 to 92% at agrid width of 0.1% was applied, as the minimum and maximum observed values were 19.8 and 91.5% for ScvO2, respectively. As a lower bound below 19.8% and a upper bound above 91.5% may be selected by our algorithm it also covers the case of a single cut-off. In each cross validation step the classification error for (731*730/2 =) 266815 pairs of grid points (i.e., all possible lower and upper bounds) was evaluated to define lower and upper bounds and the combination with the smallest misclassification chosen. Thus, this approach allowed to test all possible pairs of values simultaneously to calculate the minimal classification error for a pair of lower and upper limits of normal.The final lower and upper bounds based on the training set were then obtained as the median values of the 250 random replications.Thus, three groups were defined as ‘low’, ‘normal’, and ‘supranormal’ ScvO2, based on their immediate ScvO2 values upon admission to the ICU.

ESM Table 1: Vasoactive and inotropic drugs

low ScvO2-normal ScvO2-supranormal ScvO2-p-value

group group group

Epinephrine

0 hours after admission

No. of patients /total no. 4 / 27 37 / 115 42 / 63 *† < .001

dosage [µg/kg/min]0.33 [0.11 – 0.36]0.05 [0.03 – 0.08]0.06 [0.05 – 0.10] .011

3 hours after admission

No. of patients /total no. 8 / 26 38 / 115 44 / 63 *†< .001

dosage [µg/kg/min]0.11 [0.03 – 0.27]0.05 [0.03 – 0.08]0.07 [0.04 – 0.10] .067

6 hours after admission

No. of patients /total no. 7 / 56 36 / 115 41 / 63 *†< .001

dosage [µg/kg/min]0.07 [0.03 – 0.19] 0.05 [0.03 – 0.08] 0.07 [0.05 – 0.09] .098

9 hours after admission

No. of patients /total no. 7 / 26 38 / 115 39 / 63 *†< .001

dosage [µg/kg/min]0.07 [0.03 – 0.19]0.04 [0.03 – 0.07] 0.07 [0.04 – 0.09] * .044

12 hours after admission

No. of patients /total no. 7 / 26 37 / 115 36 / 63 *† .002

dosage [µg/kg/min]0.07 [0.03 – 0.29]0.04 [0.03 – 0.07]0.07 [0.04 – 0.09] * .027

24 hours after admission

No. of patients /total no. 6 / 26 35 / 115 31 / 63 * .002

dosage [µg/kg/min]0.13 [0.03 – 0.25] 0.04 [0.03 – 0.07] 0.06 [0.03 – 0.09] .128

Norepinephrine

0 hours after admission

No. of patients /total no. 21 / 27 87 / 115 50 / 63 .893

dosage [µg/kg/min]0.06 [0.02 – 0.15]0.06 [0.03 – 0.10]0.10 [0.05 – 0.17] * .004

3 hours after admission

No. of patients /total no. 20 / 26 74 / 115 45 / 63 .403

dosage [µg/kg/min]0.04 [0.02 – 0.11] 0.04 [0.03 – 0.10]0.10 [0.05 – 0.17] * .005

6 hours after admission

No. of patients /total no. 16 / 26 62 / 115 40 / 63 .466

dosage [µg/kg/min]0.06 [0.04 – 0.11]0.04 [0.02 – 0.10] ‡0.06 [0.04 – 0.11] ‡ .102

9 hours after admission

No. of patients /total no. 13 / 26 59 / 115 37 / 63 .594

dosage [µg/kg/min]0.07 [0.05 – 0.11] ‡0.05 [0.02 – 0.09] ‡0.06 [0.03 – 0.16] ‡ .156

12 hours after admission

No. of patients /total no. 12 / 26 53 / 115 31 / 63 .921

dosage [µg/kg/min]0.07 [0.05 – 0.17]0.04 [0.02 – 0.08] ‡0.09 [0.03 – 0.29] *‡ .009

24 hours after admission

No. of patients /total no. 11 / 26 49 / 115 27 / 63 .999

dosage [µg/kg/min]0.06 [0.03 – 0.14] ‡0.04 [0.02 – 0.08] ‡0.08 [0.04 – 0.24] *‡ .027

Dobutamine

0 hours after admission

No. of patients /total no. 3 / 27 * 0 / 115 2 / 63 .003

dosage [µg/kg/min]3.40 [2.90 – 4.50] -3.90 [3.80 – 4.00] .564

3 hours after admission

No. of patients /total no. 4 / 26 4 / 115 6 / 63 .042

dosage [µg/kg/min]4.00 [3.02 – 4.88]3.91 [3.61 – 6.23]3.65 [3.33 – 4.55] .611

6 hours after admission

No. of patients /total no. 6 / 26 12 / 115 7 / 63 .231

dosage [µg/kg/min]3.80 [3.20 – 4.63]4.20 [3.61 – 5.40] ‡4.00 [3.50 – 5.80] .675

9 hours after admission

No. of patients /total no. 6 / 26 19 / 115 8 / 63 .452

dosage [µg/kg/min]3.80 [3.20 – 4.63]4.20 [3.60 – 5.10] ‡4.00 [3.58 – 5.55] .595

12 hours after admission

No. of patients /total no. 6 / 26 19 / 115 10 / 63 .647

dosage [µg/kg/min]3.35 [2.98 – 4.28]4.20 [3.60 – 5.10] ‡4.00 [3.58 – 5.35] .196

24 hours after admission

No. of patients /total no. 9 / 26 22 / 115 13 / 63 .229

dosage [µg/kg/min]4.20 [3.15 – 4.40]4.20 [3.45 – 5.13] ‡4.06 [3.70 – 5.50] ‡ .593

Milrinone

0 hours after admission

No. of patients /total no. 0 / 27 0 / 115 0 / 63 -

dosage [µg/kg/min] - - - -

3 hours after admission

No. of patients /total no. 1 / 26 0 / 115 1 / 63 .098

dosage [µg/kg/min]0.19 [0.19 – 0.19] - 0.15 [0.15 – 0.15] .317

6 hours after admission

No. of patients /total no. 2 / 26 0 / 115 1 / 63 .018

dosage [µg/kg/min]0.36 [0.24 – 0.48] -0.15 [0.15 – 0.15] .221

9 hours after admission

No. of patients /total no. 2 / 26 0 / 115 2 / 63 .025

dosage [µg/kg/min]0.36 [0.24 – 0.48] -0.16 [0.15 – 0.16] .121

12 hours after admission

No. of patients /total no. 2 / 26 1 / 115 2 / 63 .081

dosage [µg/kg/min]0.36 [0.24 – 0.48]0.33[0.33 – 0.33]0.16 [0.15 – 0.16] .223

24 hours after admission

No. of patients /total no. 2 / 26 1 / 115 2 / 63 .081

dosage [µg/kg/min]0.36 [0.24 – 0.48]0.33[0.33 – 0.33]0.37 [0.31 – 0.42] .999

Data are medians and interquartile range and absolute values

* indicates p< .05 compared to normal ScvO2-group; † indicates p< .05 compared to low ScvO2 -group

‡ indicates p < .05 compared to baseline

ESM Table 2:

Intra- and postoperative heart rate, mean arterial pressure and volume therapy

low ScvO2- / normal ScvO2- / supranormal ScvO2- / p-value
group / group / group
n=27 / n=115 / n=63
Intraoperative values
heart rate [bpm]
after induction of anaesthesia / 63 [53 – 70] / 64 [54 – 74] / 62[55 - 73] / .849
after heparin bolus / 70.5 [59.8 - 83.5] / 70 [61 - 78.3] / 71 [60 – 83] / .850
after protamin bolus / 90 [88 - 99] / 92 [87.5 - 98] / 91 [87 – 97] / .914
mean arterial pressure [mmHg]
after induction of anaesthesia / 81 [65 - 88.8 ] / 80 [73 – 89] / 76.0 [66 - 84]* / .033
after heparin bolus / 70.5 [63.5 - 79.3] / 70 [64 – 78.3] / 66.0 [60 – 75] / .144
after protamin bolus / 72.5 [64.5 - 86.8] / 70 [64 - 75] / 67.0 [59 - 72] / .037
volume replacement
crystalloid solutions [ml/h] / 465.1 [384.6 - 606.1] / 512.8 [375 - 714.3] / 517.2 [370.2 - 714.3] / .689
colloid solutions [ml/h] / 0 [0 - 156.3] / 0 [0 - 116.3] / 0 [0 - 151.5] / .145
packed red cells [n] / 0 [0 - 1] / 0 [0 - 2] / 0 [0 - 2] / .406
fresh frozen plasma [n] / 0 [0 – 0] / 0 [0 – 0] / 0 [0 - 0] / .994
platelet concentrate [n] / 0 [0 - 1]* / 0 [0 – 0] / 0 [0 - 0] / .006
intensive care unit values
heart rate [bpm]
0 hours after admission / 94 [90 - 100] / 91 [84 - 99] / 91 [88 – 100] / .389
3 hours after admission / 93 [90 - 100.5] / 95 [90 - 100] / 92 [89 -100] / .602
6 hours after admission / 92.5 [90 - 102] / 94 [88 - 100] / 94 [90 - 101.5] / .329
9 hours after admission / 91.5 [90 - 100.5] / 92 [85 - 100] / 95 [90 - 100.3] / .270
12 hours after admission / 91.5 [90 - 100] / 92 [85 - 100] / 94 [89 – 100] / .594
24 hours after admission / 91 [90 - 100.5] / 92 [84 - 100] / 94 [89 - 99] / .679
mean arterial pressure [mmHg]
0 hours after admission / 75 [66 – 85] / 75 [66 - 82] / 75 [66 – 82] / .964
3 hours after admission / 72 [68.5 – 83.2] / 78.0 [72 - 85] / 74 [66 - 78]* / .005
6 hours after admission / 75 [69 - 84] / 76.0 [69 – 84] / 71 [67 - 77.3] / .051
9 hours after admission / 74.5 [65.8 – 78] / 77 [69 - 84] / 73 [67 – 80] / .063
12 hours after admission / 76 [70 - 79] / 77 [70 – 83] / 72 [68 - 78]* / .038
24 hours after admission / 76 [70 - 79.8] / 77 [70 – 83] / 73 [68 - 78] / .089
volume replacement
crystalloid solutions [ml/h] / 87 [58.8 - 166.7] / 125.0 [74.1 - 176.5] / 103.5 [71.4 - 166.7] / .561
colloid solutions [ml/h] / 125 [66.7 - 166.7] / 104.2 [71.4 - 172.4] / 100 [62.5 - 166.7] / .871
packed red cells [n] / 1 [0 – 2]* / 0 [0 – 0] / 0 [0 - 2] / <.001
fresh frozen plasma [n] / 0 [0 – 4] / 0 [0 - 0] / 0 [0 – 4] / .138
platelet concentrate [n] / 0 [0 - 1] / 0 [0 - 0] / 0 [0 – 0] / .078
aiDCO2
0 hours after admission / 0.5 [0.2 - 1.2] / 0.5 [0.2 - 0.9] / 0.4 [0.2 - 0.7] / .565
3 hours after admission / 0.5 [0.3 - 1.1] / 0.6 [0.4 - 1.1] / 0.6 [0.3 - 1.0] / .440
6 hours after admission / 1.2 [0.5 - 1.5] / 0.9 [0.5 - 1.3] ‡ / 0.6 [0.3 - 1.3] / .172
9 hours after admission / 0.9 [0.5 - 1.5] / 0.9 [0.5 - 1.2] ‡ / 0.8 [0.3 - 1.5] ‡ / .873
12 hours after admission / 1.2 [0.5 - 1.7] / 0.9 [0.5 - 1.2] / 0.9 [0.3 - 1.4] / .572
24 hours after admission / 1.2 [0.5 - 1.7] / 0.9 [0.5 - 1.2] / 0.9 [0.4 - 1.5] ‡ / .341

Data are medians and interquartile range and absolute values. * indicates p< .05 compared to normal ScvO2-group;† indicates p< .05 compared to low ScvO2-group, ‡ indicates p < .05 compared to baseline