ChatGPT-4o performance in predicting the final diagnosis of febrile returning travellers presented to the emergency department at Sheba Medical Centre during 2009–2024 (N = 114)
| . | ChatGPT-4o correct hits . | |
|---|---|---|
| . | N . | % (95% CI) . |
| Primary outcome: | ||
| Success rate in predicting the final diagnosis when requested to specify the top three differentials | ||
| All cases (N = 114) | 89 | 78.1 (69.4, 85.3) |
| Only malaria cases (N = 48) | 47 | 97.9 (88.9, 100) |
| Only dengue cases (N = 19) | 18 | 94.7 (74.0, 99.9) |
| Secondary outcomes: | ||
| Success rate in predicting the final diagnosis when requested to specify the single most likely diagnosis | ||
| All cases (N = 114) | 78 | 68.4 (59.1, 76.8) |
| Only malaria cases (N = 48) | 43 | 89.6 (77.3, 96.5) |
| Only dengue cases (N = 19) | 16 | 84.2 (60.4, 96.6) |
| Success rate in recommending a diagnostic test that could confirm the final diagnosis | ||
| All cases (N = 114) | 94 | 82.5 (74.2, 88.9) |
| Only malaria cases (N = 48) | 48 | 100 (93–100) |
| Only dengue cases (N = 19) | 18 | 94.7 (74.0, 99.9) |
| . | ChatGPT-4o correct hits . | |
|---|---|---|
| . | N . | % (95% CI) . |
| Primary outcome: | ||
| Success rate in predicting the final diagnosis when requested to specify the top three differentials | ||
| All cases (N = 114) | 89 | 78.1 (69.4, 85.3) |
| Only malaria cases (N = 48) | 47 | 97.9 (88.9, 100) |
| Only dengue cases (N = 19) | 18 | 94.7 (74.0, 99.9) |
| Secondary outcomes: | ||
| Success rate in predicting the final diagnosis when requested to specify the single most likely diagnosis | ||
| All cases (N = 114) | 78 | 68.4 (59.1, 76.8) |
| Only malaria cases (N = 48) | 43 | 89.6 (77.3, 96.5) |
| Only dengue cases (N = 19) | 16 | 84.2 (60.4, 96.6) |
| Success rate in recommending a diagnostic test that could confirm the final diagnosis | ||
| All cases (N = 114) | 94 | 82.5 (74.2, 88.9) |
| Only malaria cases (N = 48) | 48 | 100 (93–100) |
| Only dengue cases (N = 19) | 18 | 94.7 (74.0, 99.9) |
ChatGPT-4o performance in predicting the final diagnosis of febrile returning travellers presented to the emergency department at Sheba Medical Centre during 2009–2024 (N = 114)
| . | ChatGPT-4o correct hits . | |
|---|---|---|
| . | N . | % (95% CI) . |
| Primary outcome: | ||
| Success rate in predicting the final diagnosis when requested to specify the top three differentials | ||
| All cases (N = 114) | 89 | 78.1 (69.4, 85.3) |
| Only malaria cases (N = 48) | 47 | 97.9 (88.9, 100) |
| Only dengue cases (N = 19) | 18 | 94.7 (74.0, 99.9) |
| Secondary outcomes: | ||
| Success rate in predicting the final diagnosis when requested to specify the single most likely diagnosis | ||
| All cases (N = 114) | 78 | 68.4 (59.1, 76.8) |
| Only malaria cases (N = 48) | 43 | 89.6 (77.3, 96.5) |
| Only dengue cases (N = 19) | 16 | 84.2 (60.4, 96.6) |
| Success rate in recommending a diagnostic test that could confirm the final diagnosis | ||
| All cases (N = 114) | 94 | 82.5 (74.2, 88.9) |
| Only malaria cases (N = 48) | 48 | 100 (93–100) |
| Only dengue cases (N = 19) | 18 | 94.7 (74.0, 99.9) |
| . | ChatGPT-4o correct hits . | |
|---|---|---|
| . | N . | % (95% CI) . |
| Primary outcome: | ||
| Success rate in predicting the final diagnosis when requested to specify the top three differentials | ||
| All cases (N = 114) | 89 | 78.1 (69.4, 85.3) |
| Only malaria cases (N = 48) | 47 | 97.9 (88.9, 100) |
| Only dengue cases (N = 19) | 18 | 94.7 (74.0, 99.9) |
| Secondary outcomes: | ||
| Success rate in predicting the final diagnosis when requested to specify the single most likely diagnosis | ||
| All cases (N = 114) | 78 | 68.4 (59.1, 76.8) |
| Only malaria cases (N = 48) | 43 | 89.6 (77.3, 96.5) |
| Only dengue cases (N = 19) | 16 | 84.2 (60.4, 96.6) |
| Success rate in recommending a diagnostic test that could confirm the final diagnosis | ||
| All cases (N = 114) | 94 | 82.5 (74.2, 88.9) |
| Only malaria cases (N = 48) | 48 | 100 (93–100) |
| Only dengue cases (N = 19) | 18 | 94.7 (74.0, 99.9) |
This PDF is available to Subscribers Only
View Article Abstract & Purchase OptionsFor full access to this pdf, sign in to an existing account, or purchase an annual subscription.
