Table 2.

Weighted regression analysis of expecting benefit of using ChatGPT to improve health: univariable and multivariable analysis with and without accounting for trust (n = 1787).

		Univariable		Multivariable without system trust		Multivariable with system trust
		OR	P	OR	P	OR	P
Comfort with use	Comfort with use of ChatGPT in society	1.91	1.9 × 10-16	1.86	5.9 × 10-15	1.68	2.5 × 10-10
Tech-savviness	I consider myself a tech-savvy person	1.15	.072	1.07	.544	1.00	.981
Privacy concerns	I worry that private information about my health could be used against me	0.81	9.9 × 10-4	0.83	.014	0.88	.095
Access	I feel like I have options about where I receive my medical care	1.41	4.9 × 10-7	1.13	.169	1.03	.768
Respect	I feel respected when I seek healthcare	1.52	1.1 × 10-8	1.29	.013	1.05	.646
Trust	System trust index	1.34	6.6 × 10-20			1.26	1.6 × 10-11
Heard of ChatGPT	Yes (Reference)
	No	1.31	.064	1.30	.151	1.36	.100
Sex	Male (Reference)
	Female	0.74	.023	0.72	.025	0.69	.010
Age	18-29 (Reference)
	30-44	1.20	.397	1.37	.172	1.41	.14
	45-59	1.26	.252	1.56	.075	1.76	.025
	60+	1.29	.171	1.85	.007	2.04	.002
Race/ethnicity	White, non-Hispanic (Reference)
	Black, non-Hispanic	1.63	.003	1.54	.014	1.41	.051
	Hispanic	1.33	.097	1.24	.228	1.19	.349
	Other	1.94	.015	2.05	.020	1.83	.081
Education	Less than High school (Reference)
	High school graduate or equivalent	0.74	.195	0.71	.157	0.72	.172
	Some college/Associate degree	0.72	.165	0.65	.086	0.62	.052
	Bachelor's degree	0.82	.420	0.81	.443	0.83	.489
	Post grad study/Professional degree	0.63	.070	0.54	.034	0.54	.034
Health literacy	High (Reference)
	Low	1.18	.260	1.16	.350	0.99	.937

OR: Odds Ratio. Bold values indicate P < .05.