Table 1 Patient characteristics in combined training and testing cohort

From: Predicting arrhythmia recurrence post-ablation in atrial fibrillation using explainable machine learning

 

Recurrence

No Recurrence

p-value

N (%)

40 (59.7%)

27 (40.3%)

n/a

Age (years)

65 (53–77)

59 (46–72)

0.4896

BMI

31.2 (20.4–42.0)

29.9(22.5–37.5)

0.9747

Sex (male)

60.0%

55.6%

0.7251

Hypertension

57.5%

59.3%

0.8929

Hyperlipidemia

37.5%

40.7%

0.7968

Coronary artery disease

20%

14.8%

0.5966

OSA

32.5%

33.3%

0.7091

Stroke/TIA

10.0%

3.7%

0.3469

Diabetes

10.0%

7.4%

0.7272

COPD

10.0%

0%

0.09566

Cigarette use

17.5%

18.5%

0.9233

Thyroid disease

15.0%

7.4%

0.6570

CHF

32.5%

22.2%

0.3679

Pre-Ablation left atrial LGE (%)

15.6% (11.48–21.9%)

17.6% (8.95–20.25%)

0.9033

AFib ablation type:

Cryo PVI only

35.0%

44.4%

0.4444

Cryo PVI + RF SM

5.0%

3.7%

0.8166

RF PVI only

17.5%

11.1%

0.4814

RF PVI + SM

42.5%

40.7%

0.8929

  1. Variables are reported as median (IQR) and compared with Mann-Whitney U tests as appropriate. For binary variables, proportions were reported and compared with χ2 tests.
  2. BMI body mass index, CHF congestive heart failure, COPD chronic obstructive pulmonary disease, LGE late gadolinium enhancement, OSA obstructive sleep apnea, PVI pulmonary vein isolation, RF radiofrequency, SM substrate modification, TIA transient ischemic attack.
Back to article page