Knowledge and Attitudes Regarding Use of Chat GPT in Dentistry Among Dental Students and Dental Professionals

Introduction

ChatGPT, an artificial intelligence (AI) program that generates text in response to written instructions has become quite popular as indicated by its availability on the web through open AI. AI has the potential to improve precision and accuracy while reducing the amount of time needed for various system components, its use in healthcare systems is essential [1]. Dental AI applications have been proven to be helpful in the analysis of medical imaging, especially in the identification and diagnosis of dental caries, periodontitis, and implantitis. They have also been found to help with surgical planning in oral and maxillofacial surgery [2]. Advanced AI-based tools could have a substantial impact on healthcare education, such as Generative Pre-trained Transformer (GPT)-based tools. Using the same techniques as the InstructGPT model, ChatGPT was taught through reinforcement learning with human input. ChatGPT has received a lot of interest since it went public on November 30, 2022, particularly in the field of education [3]. Proximal Policy Optimisation, a reinforcement learning technique, was used to train InstructGPT by maximizing the preferences of human annotators for particular prompts. Contrarily, ChatGPT has been expressly developed and instructed using dialogue-like prompts for enhancing dialogue-based output [4]. ChatGPT is an advanced language model that uses deep learning methods to generate answers to inputs that resemble those of humans. The widely used and accessible ChatGPT is an example of the wide-ranging interest and scrutiny received in academia and among medical practitioners [5]. It can handle a variety of concepts because it is a versatile conversational agent, making it suitable for chatbots, customer support, and other applications [1]. Recently, the effectiveness of ChatGPT in healthcare education has been studied and it has been shown to enhance the personalized learning experience, owing to its ability to improve communication skills and students involvement in the learning process. The prospective applications of ChatGPT in the medical field include helping experts with clinical and laboratory diagnostics as well as generating relevant research ideas. However, these programs also have limitations and ethical issues like credibility and plagiarism [3]. ChatGPT's novelty and potential for use in dental education & research require an awareness of the elements that influence its acceptability. The revolutionary potential of ChatGPT in self-learning, feedback, and problem-solving requires research for a successful integration. Assessing the attitudes of dental professionals provides insight into benefits and technology preparedness & also assists in addressing ethical concerns about using ChatGPT responsibly. The purpose of this study was to assess the knowledge and attitude of dental professionals in the adoption and use of ChatGPT.

Materials and Methods

Study participants

The participants were recruited by convenience sampling through the contacts of the authors. The survey link was sent through WhatsApp and Facebook groups targeted to dental students & dental professionals. The survey was open from 06 September 2023 and was closed on 30 September 2023. A total of 354 respondents took part in this study. Participation was voluntary and anonymous without any incentive given for participation. The survey consisted of 28 closed-ended questions divided into sections of demographic details, ChatGPT in Dental education, ChatGPT in Dental Research, ChatGPT in Dental practice, and conclusive questions. This study was approved by the Institutional Review Board (IRB), with the approval number 164, granted on 05 September 2023.

The survey development process involved a literature review followed by questionnaire development, validation, and pilot testing to ensure clarity. The results were analyzed and tabulated. Frequency and percentage were calculated. To assess the knowledge and attitude of dental professionals towards ChatGPT, statistical analysis using the Chi-square test was done.

Results and Discussion

The respondents predominantly belonged to the age group of 17-30 Years (82.8%) followed by >45 years. The majority were females (73.2%). Around 63% belonged to the undergraduate category followed by faculty (16.1%) and then postgraduates (12.4%). More than 87% said that they like using technology like ChatGPT for dental practice, learning, and research & 90% believed that it has the potential to enhance efficiency and productivity in dentistry. As high as (89.3%) of respondents had heard of ChatGPT among whom 80% were familiar with its applications (Table 1).

Table 1. Characteristics of the study respondents (n=354)

With regards to questions related to ChatGPT in dental education, 82% were aware that it could have a role in enhancing education in dentistry, and around 94.6% rated ChatGPT as having high to moderate potential to enhance education. Around 50% said that they would consider integrating ChatGPT into the dental education curriculum. Most respondents said that ChatGPT was beneficial for providing instant answers to student queries (63%), explaining complex dental concepts (60%), generating practice exam questions (43%), as a tool for group learning and discussion (40%) and for assisting in case-based learning (39%) and repetitive tasks (29%). The possible limitations mentioned included factual inaccuracies in case scenarios (54%), less reliance on teachers for learning (45%), issues regarding plagiarism (42%), and essay assignments becoming obsolete (41%) (Figures 1a and 1b).

With regards to the role of ChatGPT in dental research, around 89% of the respondents felt that it could have a role in enhancing research-related tasks in dentistry. When asked about tasks ChatGPT would help in, most respondents said that it would assist in data collection and analysis (58%), followed by a literature review (53%) & writing research summaries (51%) as well as for assistance in manuscript writing (41%) & formulation of research hypothesis (39%). Around 6% felt that it could not help in any research-related task. Around 47% felt that relying on ChatGPT would prevent them from acquiring critical thinking skills & 48% had concerns about using ChatGPT in reference to plagiarism and 40% felt it may lead to plagiarism or copyright infringement accusations. The common limitations that respondents said were ethical concerns for ghostwriting and plagiarism (65%), factual inaccuracies (51%) followed by redundancy in the text (31%), and bias based on the dataset used for training (45%) (Figures 1c and 1d).

The features of ChatGPT that could help in dental practice as per respondents, included a quick response to patient queries (61%), language translation for multilingual patients (58%), detailed procedural explanations (47%), case-specific treatment suggestions (41%), oral hygiene and post-procedural instructions (40.1%). The limitations include lack of emotional support and patient interaction (66%), lack of thoughtful reasoning (56%), medicolegal issues  (55%) & concerns regarding privacy (44.4%) & inaccurate information (40%) (Figure 2).

Around 82.4% said that they would want to integrate ChatGPT into their dental practice for tasks like patient education and appointment scheduling. Only 16% said that they would use ChatGPT to stay updated with the latest advances in dentistry. However, 84% said that they would use ChatGPT to obtain quick answers for information in practice. Around 77% did not have reservations about ChatGPT’s capacity to offer accurate information, while 30% believed that it may not be accurate (Table 2).

Table 2. Component Analysis

Around 55% thought that ChatGPT has the potential to advance dentistry if used ethically and responsibly. Most respondents believed that ChatGPT is most beneficial for dental education (72%), dental research (63%), and dental practice (26%). 191 respondents said that they would consider using ChatGPT for dental education, research, and practice-related tasks (Table 3).

Table 3. Conclusive Questions

The knowledge of ChatGPT was maximal in respondents in the age group of 31-45 Years (80%) followed by 17-30 years (65%). Males and females demonstrated almost similar knowledge while faculty members & undergraduates followed by practitioners showed good knowledge. The undergraduate students regardless of the year, had good knowledge regarding ChatGPT. None of the results were statistically significant (p<0.05) (Table 4).

Table 4. Knowledge regarding ChatGPT

The attitude to use ChatGPT was maximum in 31-45 years (73.1%) followed by >45 years (75%) which corresponded to faculty members (70.1%) and practitioners (77.4%). It was almost similar in both males and females. The high attitude to use ChatGPT was presently seen in 3^rd & 4^th BDS students. None of the results were statistically significant (p<0.05) (Table 5).

Table 5. Attitude regarding ChatGPT