BACKGROUND
As cancer incidence continues to rise, oncologists are experiencing escalating work pressure and challenges. In recent years, China has made notable progress in using Artificial Intelligence (AI) technologies, particularly in the realm of cancer diagnosis and treatment. However, the understanding, attitudes, and utilization of AI among oncologists in China have not been systematically studied.
OBJECTIVE
We conducted a nationwide cross-sectional questionnaire aimed to gain insights into the cognition, attitudes, and utilization of AI among oncologists in China.
METHODS
This survey initiative was carried out on behalf of the Chinese Society of Artificial Intelligence in Oncology, the China Anti-Cancer Association (CACA). The questionnaire design underwent approval after deliberations among a panel of 12 board members. Data collection took place from March to October 2023, resulting in the acquisition of 1,350 completed questionnaires. We performed statistical and quantitative analysis on the questionnaire data to uncover insights into the current landscape of oncologists' interactions with AI in China.
RESULTS
Among the 1,350 respondents, we found that 1,128 (84%) do not have a clear understanding of the definition of AI. The variances in doctors' perceptions of AI were primarily attributed to factors related to their personal and professional backgrounds (including age, education, clinical technical title, and work experience), and work environment (clinical department, university hospital, and hospital category). Economic region (p=0·15) and hospital level (p=0·27) did not have a significant impact. In the 5-point Likert scale of all 13 questions on participants’ attitude towards AI, 11 questions scored more than 3·5 scores on average. Oncologists considered that the top three positive impacts of AI were improving the accuracy of diagnosis and treatment (84%), reducing diagnostic errors (72%), and improving clinical and research capabilities (83%). However, they were also concerned that AI would lead to a decrease in capabilities (75%), misdiagnoses (65%), and data bias (62%). In terms of AI utilization, oncologists mostly hoped to use AI-assisted diagnostic tools (73%), clinical decision support systems (54%), and risk prediction models (46%). Regarding the experience of AI utilization, 68% of oncologists had been incorporating AI into their practice for less than three years.
CONCLUSIONS
Oncologists in China are under work pressures and burdens, and the potential of AI to assist them has already been demonstrated. Chinese oncologists need to improve their understanding of AI through professional education and further improve their attitudes towards AI to effectively leverage AI technologies in their practice.