Skip to main content
. 2023 Nov 20;13(6):1460–1487. doi: 10.3390/clinpract13060130

Table 3.

Comparison of correct response proportions of GPT-4 for MIR examination in Spanish by specialty.

Specialty % Correct N Specialty % Correct N
Pathology 100% 1 Forensics & Legal Medicine 50.0% 2
Cardiology 77.8% 9 Family Medicine 88.9% 9
Orthopedic Surgery and Traumatology 100% 10 Nephrology 75.0% 4
dermatology 100% 1 Pneumology 66.7% 6
digestive 100% 6 Neurology 90.0% 10
endocrinology 100% 8 Ophthalmology 100% 3
epidemiology 71.4% 7 gynecologic oncology 100% 2
Plastic Surgery 100% 2 Oncology 100% 2
Ethics 100% 4 ENT (Otorhinolaryngology) 100% 3
pharmacology 40% *** 5 palliative care 100% 2
physiopathology 100% 7 Pediatrics 90.0% 10
genetics 100% 2 Public Health 66.7% 3
geriatrics 100% 6 Psychiatry 100% 8
gynecology 88.9% 9 Rheumatology 77.8% 9
hematology 100% 5 Critical Care 33.3% ** 3
infectious diseases 57.1% * 7 emergency medicine 83.3% 6
immunology 100% 7
Maxillofacial
surgery
100% 2 TOTAL 182

Adjusted residual analysis. * p < 0.05; ** p < 0.01; *** p < 0.001.