Table 3 ChatGPT performance accuracy in the MRCOG part two

From: Exploring the capabilities of ChatGPT in women’s health: obstetrics and gynaecology

Knowledge area

Correct

Incorrect

Total

Antenatal care

48 (39.0%)

75 (61.0%)

123

Clinical skills

9 (52.9%)

8 (47.1%)

17

Core surgical skills

28 (40.0%)

42 (60.0%)

70

Early pregnancy care

24 (52.2%)

22 (47.8%)

46

Gynaecological oncology

29 (50.9%)

28 (49.1%)

57

Gynaecological problems

107 (51.2%)

102 (48.8%)

209

Management of delivery

17 (56.7%)

13 (43.3%)

30

Management of labour

16 (35.6%)

29 (64.4%)

45

Maternal medicine

95 (55.6%)

76 (44.4%)

171

Postoperative care

13 (56.5%)

10 (43.5%)

23

Postpartum problems

18 (56.2%)

14 (43.8%)

32

Sexual & reproductive health

17 (51.5%)

16 (48.5%)

33

Subfertility

18 (50.0%)

18 (50.0%)

36

Teaching & research

25 (58.1%)

18 (41.9%)

43

Urogynaecology & pelvic floor problems

34 (63.0%)

20 (37.0%)

54

Total

498

491

989

  1. Part Two comprises single best answer (SBA) and extended matching questions (EMQ) from 15 knowledge areas (subjects). Accuracy across the knowledge areas did no vary significantly (p = 0.10, χ² = 21.05). Values in brackets denote the percentage proportion (%).