Table 4. Comparison of information sources.

Variable	Outcomes	Standard information (median and 25th; 75th percentiles)	ChatGPT responses (median and 25th; 75th percentiles)	p-value*
How useful was the information?	0 = very little; 5= = really useful	4 (3;4)	5 (4;5)	0.0002
How understandable was the information?	0 = very difficult; 5 = very easy	4 (4;5)	5 (4;5)	0.0005
Did the information make you nervous about getting the side effects?	0 = very little; 5 = very much	4 (3;5)	3 (2;4)	0.000
Which of the two information texts do you prefer?	N (proportion)	7 (15.2%)**	36 (78.3%)**	-