Table 2 Patient characteristics.

From: Machine learning to predict late respiratory support in preterm infants: a retrospective cohort study

Characteristics

All

Training data set

Testing data set

p value

Patient number, n (%)

3200 (100%)

2240 (70%)

960 (30%)

 

GA, week, mean ± SD

27.47 ± 2.35

27.48 ± 2.35

27.44 ± 2.35

0.613

BW, per 100 g, mean ± SD

9.28 ± 2.80

9.27 ± 2.80

9.29 ± 2.79

0.832

SGA*, n (%)

1014 (32%)

725 (30%)

289 (32%)

0.029**

Sex, female, n (%)

1496 (47%)

1,050 (47%)

446 (46%)

0.829

Outborn, n (%)

203 (6%)

132 (6%)

71 (7%)

0.110

Maternal history

 Antenatal corticosteroid, n (%)

2781 (87%)

1950 (87%)

831 (87%)

0.706

 Antenatal MgSO4, n (%)

1926 (60%)

1349 (60%)

577 (60%)

0.950

 Chorioamnionitis, n (%)

580 (18%)

397 (18%)

183 (19%)

0.367

 Maternal hypertension, n (%)

747 (23%)

521 (23%)

226 (24%)

0.862

 Caesarean section, n (%)

2,239 (70%)

1580 (71%)

659 (69%)

0.285

 Multiple births, n (%)

936 (29%)

658 (29%)

278 (29%)

0.812

  1. BW Birth weight; GA Gestational age; SD Standard deviation; SGA Small for gestational age.
  2. *Reference: Hsieh21.
  3. **p < 0.05.