Table 4.
Multivariate binary logistic regression analysis of healthcare workers’ variables and their expectation of ChatGPT potential improvement of the healthcare future (N = 1057).
| Variable? | Multivariate Adjusted Odds Ratio | OR 95% C.I. | p-Value | |
|---|---|---|---|---|
| Lower | Upper | |||
| Sex | 0.772 | 0.536 | 1.112 | 0.164 |
| Age | 0.960 | 0.822 | 1.122 | 0.609 |
| Clinical role | 1.048 | 0.858 | 1.281 | 0.646 |
| Trust * | 0.804 | 0.627 | 1.031 | 0.085 |
| History of ChatGPT use at the time of the survey | 1.902 | 1.226 | 2.950 | 0.004 |
| Familiarity with ChatGPT | 2.023 | 1.508 | 2.714 | <0.001 |
| Medical decisions @ | 1.463 | 0.994 | 2.154 | 0.054 |
| Comfort level Ψ | 2.327 | 1.650 | 3.281 | <0.001 |
| Patients’ outcomes Σ | 7.927 | 5.046 | 12.452 | <0.001 |
| Constant | 0.006 | <0.001 | ||
Dependent variable = Participants’ expectation of potential healthcare future improvement by ChatGPT. * Trust level in AI’s ability to provide medical decisions for healthcare providers; @ Belief in ChatGPT’s ability to provide valuable medical decisions; Ψ Using ChatGPT for medical purposes/practice comfort level; Σ Belief of ChatGPT patients’ outcomes improvement.