Table 2 Demographic and clinical characteristics. Values are mean (SD), median (IQR [range]) or number (proportion).

From: Automated pain detection using facial expression in adult patients with a customized spatial temporal attention long short-term memory (STA-LSTM) network

Parameter

Training cohort n = 160

Validation cohort n = 40

Overall n = 200

Age, y

47.4 (11.8)

45.1 (12.8)

46.9 (12.0)

Gender

 Female

108 (67.5%)

32 (80.0%)

140 (70.0%)

 Male

52 (32.5%)

8 (20.0%)

60 (30.0%)

Race

 Chinese

96 (60.0%)

26 (65.0%)

122 (61.0%)

 Malay

28 (17.5%)

6 (15.0%)

34 (17.0%)

 Indian

20 (12.5%)

3 (7.5%)

23 (11.5%)

 Others

16 (10.0%)a

5 (12.5%)b

21 (10.5%)

 Weight, kg

68.9 (17.0)

70.8 (15.6)

69.3 (16.7)

 BMI, kg.m−2

26.1 (5.4)

27.0 (5.1)

26.3 (5.3)

ASA status

 I

42 (26.3%)

11 (27.5%)

53 (26.5%)

 II

106 (66.3%)

26 (65.0%)

132 (66.0%)

 III

12 (7.4%)

3 (7.5%)

15 (7.5%)

Purpose of visit

 Pain consultation

72 (45.0%)

19 (47.5%)

91 (45.5%)

 Gynecological surgery

80 20 (50.0%)

20 (50.0%)

100 (50.0%)

 Procedure

7 (4.4%)

1 (2.5%)

8 (4.0%)

 Pre-procedure

1 (0.6%)

0

1 (0.5%)

Pre-surgical/procedural pain score (at rest)

0 (3 [0–9])

0 (3 [0–7.5])

0 (3 [0–9])

Pre-surgical/procedural EQ-5D-3L health VAS score

72.9 (16.8)

73.0 (18.5)

72.9 (17.1)

Post-surgical/procedural pain score (at rest)

2 (3 [0–10])

2 (3 [0–7])

2 (3 [0–10])

  1. ASA American society of anesthesiologists, BMI body mass index, VAS visual analog scale.
  2. aOthers included Filipino (n = 8), Indonesian (n = 4), Burmese (n = 1), Bangladeshi (n = 1), and Myanmese (n = 1).
  3. bOthers included Filipino (n = 2), Burmese (n = 1), Eurasian (n = 1), and Pakistani (n = 1).