Mean score, length and points per 100 words of each group
| Group | Mean score | Mean length (words) | Points per 100 words |
| A: Random doctor | 6.0 (5.5 to 6.6) | 320 (280 to 350) | 2.2 (1.9 to 2.5) |
| B: Top-tier doctor | 7.2 (6.8 to 7.6) | 350 (320 to 380) | 2.3 (2.0 to 2.6) |
| C: GPT-4 | 4.5 (3.9 to 5.0) | 290 (280 to 300) | 1.6 (1.4 to 1.7) |
95% CIs in parenthesis. The points per 100 words is the average of the points per 100 words for each case.