Table 2 Lab characteristics of patient cohorts used in the study (as mean ± standard deviation).

From: Towards artificial intelligence-based learning health system for population-level mortality prediction using electrocardiograms

 

Full Data (n = 601,307)

Development set (n = 361,585)

Holdout set (n = 239,722)

Random ECG per patient in holdout set (n = 30,076)

Age (years)

64.82 ± 17.57

64.90 ± 17.53

64.70 ± 17.62

64.74 ± 17.87

Sex (Male in %)

55.84

55.81

55.88

56.63

Lab measurements

GFR (mL/min)

66.01 ± 28.03

65.85 ± 28.04

66.25 ± 28.03

69.66 ± 27.94

Creatinine (umol/L)

116.46 ± 117.03

117.20 ± 119.62

115.35 ± 113.00

107.52 ± 100.63

Hemoglobin (g/L)

126.53 ± 23.23

126.46 ± 23.27

126.65 ± 23.18

127.81 ± 23.32

Potassium (mmol/L)

4.02 ± 0.64

4.02 ± 0.64

4.02 ± 0.63

3.99 ± 0.62

Sodium (mmol/L)

137.39 ± 4.35

137.38 ± 4.36

137.41 ± 4.33

137.39 ± 4.39

Troponin I (ug/L)

2.24 ± 10.07

2.23 ± 10.09

2.24 ± 10.04

1.99 ± 9.49

  1. ECG electrocardiogram, GFR glomerular filtration rate, N number.