Integrating ChatGPT Into Clinical Laboratory Science Education: A Survey and Pilot Study

INTRODUCTION

ChatGPT, a powerful language model developed by OpenAI (San Francisco, CA, USA) and released in late 2022, has sparked extensive discussions about its transformative potential across various fields, including higher education.^1-3 Generative artificial intelligence (AI) tools such as ChatGPT excel in natural language communication, demonstrating the ability to generate high-quality, human-like text across diverse formats, from poetry to programming code.^4,5 These tools are particularly effective in summarizing information, comparing viewpoints, answering questions, and completing a wide range of language-based tasks. Such capabilities offer significant opportunities to transform teaching and learning in higher education, particularly within health professions education.^6-8

This potential has led educators to examine the implications of tools such as ChatGPT, particularly in academic writing and other educational applications.^1,3 Although enthusiasm for these tools is evident, concerns about ethical implications have prompted debates about restricting their use.⁶ However, prohibiting AI tools is neither practical nor conducive to progress. A more constructive and realistic approach is to integrate these technologies responsibly into educational practices.^3,9 For instance, ChatGPT could function as a virtual tutor, supporting learners during lectures, answering course-related questions, and fostering collaboration between teachers and students, thereby enhancing educational efficiency.³

Educators in health professions have a responsibility to ensure that AI technologies are used responsibly to enhance learning while preserving academic integrity.^3,10 Integrating AI into clinical laboratory science (CLS) education can help learners navigate complex subjects and enhance engagement. However, addressing the limitations and challenges of AI tools is essential to optimize their use.¹¹ By doing so, educators can refine their instructional strategies, address potential misconceptions, and uphold ethical standards, leveraging AI’s potential to improve learning outcomes while maintaining rigor and integrity in CLS education.

Entry-level CLS education centers on the practical aspects of clinical laboratory testing, with a strong emphasis on hands-on skills such as specimen processing, analytical techniques, and result interpretation. This foundational training equips learners with the technical expertise required to perform routine laboratory procedures and contribute effectively to patient care. In contrast, advanced CLS education, such as a Doctorate in Clinical Laboratory Science (DCLS), focuses on elevating professional competencies beyond technical proficiency. DCLS curricula emphasize advanced expertise in laboratory management, clinical research, evidence-based practice, and leadership. This advanced training prepares graduates for diverse roles, including consultancy, academic positions, and leadership in clinical laboratory management. By integrating technical proficiency with advanced professional skills, DCLS education addresses the evolving demands of modern health care and fosters leadership within the laboratory profession.

The potential applications of AI tools in CLS education, research, and clinical practice have generated considerable research interest in this rapidly evolving field. This study investigates the knowledge, attitudes, and use of ChatGPT among CLS and DCLS learners and explores its practical applications in enhancing CLS/DCLS education.

MATERIALS AND METHODS

RESULTS

Survey Results

Participant Demographics

A total of 13 CLS learners and 10 DCLS learners responded to the questionnaires, representing a 76% response rate for CLS learners and a 100% response rate for DCLS learners. The age range for CLS respondents was 22–25 years, representing senior college students, whereas DCLS respondents ranged from 27 to 40 years, representing postcollege professionals. Among the DCLS respondents, 2 (20%) were male, and 8 (80%) were female. In the CLS group. 85% of the respondents were female.

All CLS learners were in their senior year and lacked prior MLS work experience. In contrast, all DCLS learners held either a Bachelor’s or Master’s degree, were certified as MLSs by the American Society for Clinical Pathology (ASCP), and had at least 5 years of professional experience in the clinical laboratory field. These distinct demographic differences between CLS and DCLS participants contribute to notable variations in their knowledge, attitudes, and use of ChatGPT.

Scoring of Participants’ Responses

Participants’ responses regarding their knowledge, attitudes, and use of ChatGPT were summarized by program. Descriptive statistics, including frequency counts, percentages, means, and standard deviations, were calculated for the knowledge, attitude, and use domains for both CLS and DCLS respondents. Knowledge was assessed using 4 questions, each scored based on the correctness of responses (range: 0–1). Correct answers earned 1 point, whereas incorrect answers received 0 points. Familiarity with ChatGPT, which did not have a definitive correct answer, was evaluated on an ordinal scale (range: 0–4). For the attitude and use domains, responses were scored as follows: A “Yes” response earned 1 point, whereas a “No” response scored 0 points.

Participants’ General Knowledge About ChatGPT

Participants’ general knowledge about ChatGPT varied across respondents. Correct answer rates for the 4 ChatGPT knowledge questions ranged from 10% to 80% for both CLS and DCLS participants. On average, CLS learners achieved a correct answer rate of 39%, whereas DCLS learners scored higher at 55% (Table 1).

View this table:

• View inline
• View popup

Table 1.

Descriptive analysis of knowledge about ChatGPT across various educational programs

Significant differences were observed between CLS and DCLS participants in their familiarity with ChatGPT and understanding its capabilities. CLS learners reported significantly greater familiarity with ChatGPT compared with DCLS learners (P < .05; Table 1), with the distribution of response choices (score range: 0–4) shown in Figure 1. However, when questioned about ChatGPT’s capabilities, CLS learners were more likely to provide incorrect answers (P < .05; Table 1). Both groups demonstrated some uncertainty regarding ChatGPT’s limitations, including common misconceptions such as believing it could provide consultation services or reference recent real-world events.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

The difference in familiarity with ChatGPT among CLS and DCLS learners. Clinical laboratory science (CLS), n =13; Doctorate in Clinical Laboratory Science (DCLS), n = 10.

Participants’ Attitudes Toward ChatGPT

Participants’ attitudes toward ChatGPT reflected a range of perspectives, highlighting both optimism and concerns regarding its potential applications. The “Yes” response rates for 2 questions ranged from 17% to 60% and are summarized in Table 2. Regarding ChatGPT’s integration into educational curricula, significantly more DCLS respondents supported its integration into DCLS education, viewing it as an emerging technology with valuable application in clinical laboratory settings (P < .05; Table 2). Additionally, over half of the respondents believed that undergraduate CLS education is currently the most impacted by ChatGPT.

View this table:

• View inline
• View popup

Table 2.

Descriptive analysis of attitudes toward ChatGPT across various educational programs

Table 3 highlights the primary concerns and challenges associated with integrating ChatGPT into CLS and DCLS education. Among DCLS respondents, academic integrity emerged as the top concern, whereas CLS respondents were most worried about the potential for inaccurate outputs. DCLS learners emphasized the need for technical expertise as a significant challenge, whereas CLS learners identified limited availability of training data as a major hurdle. Despite these differences, both groups agreed on the importance of faculty providing guidance on the critical and responsible use of AI technology.

View this table:

• View inline
• View popup

Table 3.

The primary concerns and major challenges associated with integrating ChatGPT in CLS/DCLS education

Use of ChatGPT Among CLS and DCLS Participants

CLS learners reported significantly greater familiarity with ChatGPT and more frequent use of it in their coursework compared with DCLS learners (P < .05; Table 4). Notably, none of the DCLS learners reported using ChatGPT in their studies at the time of the survey. Among CLS learners who had used ChatGPT, it was primarily used for assignments such as case studies, essay writing, and presentation. They typically used ChatGPT to generate ideas, answer specific questions, and assist in finding results for tasks.

View this table:

• View inline
• View popup

Table 4.

Descriptive analysis of ChatGPT use in CLS/DCLS education across various educational programs

The perceived utility of ChatGPT in future clinical practice, along with its potential applications in various clinical laboratory functions, is illustrated in Figure 2 and Table 5. Nearly 60% of CLS and DCLS respondents foresee that ChatGPT/AI will be moderately to very useful in future clinical laboratories. CLS learners highlighted its potential for recognizing trends in quality control data, whereas DCLS learners emphasized its value in developing patient education materials.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

The difference in the usefulness of ChatGPT for future careers between CLS and DCLS learners. Clinical laboratory science (CLS), n = 11; Doctorate in Clinical Laboratory Science (DCLS), n = 10.

View this table:

• View inline
• View popup

Table 5.

What MLS/DCLS functions in the clinical laboratory could ChatGPT or similar software be most useful for?

DISCUSSION

The emergence of ChatGPT has introduced a new dimension to higher education, offering significant potential to reshape teaching and learning. Despite concerns persist within the educational community regarding risks, such as cheating, plagiarism, and inaccurate output, several positive aspects of ChatGPT in teaching and learning have been identified. These include enhancing learner engagement, providing learning support, reducing teachers’ workload, increasing time efficiency, and overcoming teaching barriers.^7,8

This study represents the first survey specifically conducted among CLS learners to explore their knowledge, perspectives, and usage of ChatGPT in the study and practice of the laboratory profession. Although the sample size accounted for only half of the total learners in the programs, the survey included both undergraduate CLS and graduate DCLS participants. Although both programs provide essential training in CLS, DCLS education emphasizes advanced clinical, research, and management skills necessary for leadership roles and advanced practice. Consequently, participants from CLS and DCLS programs exhibit differing responses regarding their familiarity with ChatGPT, its use, and its integration into professional study and practice. These differences reflect variations in age, educational level, training processes, and clinical experiences.

From the knowledge domain analysis, we identified several misconceptions among the respondents, particularly regarding the limitations and capabilities of ChatGPT. For instance, ChatGPT cannot provide consultations, critiques, or cite third-party sources or reference recent real-world events or published material.⁶ Additionally, although ChatGPT excels in essay writing and natural language processing, it lacks training in critical thinking, problem-solving, and reasoning skills.⁶ Reflective tasks, such as explaining thought processes, developing concept diagrams, and creating mind maps, are not well suited for AI tools at present.

CLS education emphasizes these reflective and analytical tasks to bridge the gap between academic learning and workplace skills, providing learners authentic, hands-on experiences. This is particularly relevant for DCLS practitioners with whom critical thinking and innovative problem-solving are essential for clinical practice. DCLS professionals possess a unique combination of advanced clinical knowledge and complex decision-making abilities, which are developed through specialized education and clinical residencies—tasks that remain beyond the capabilities of AI. Moreover, DCLS practitioners bring expertise in patient care coordination, quality assurance, and interpreting laboratory results within a broader clinical context—responsibilities that AI is currently unable to fully perform or replicate.

In the attitude domain, respondents expressed diverse opinions regarding the integration of ChatGPT into the CLS/DCLS curricula, highlighting both opportunities and challenges associated with the technology. DCLS respondents generally demonstrated a more positive attitude toward integrating ChatGPT in DCLS education, as reflected by a significantly higher rate of “Yes” response. This suggests that advanced laboratory practitioners, such as DCLS learners, may feel more confident using AI tools because of their extensive professional training and clinical experience. The differences in educational focus and clinical practice between CLS and DCLS learners likely contribute to their varying perspectives on the benefits, concerns, and challenges of integrating ChatGPT into CLS/DCLS education.

Respondents recognized the impact of ChatGPT on CLS/DCLS teaching and learning, and perceived undergraduate CLS education currently as being more profoundly affected. Both groups agreed that ChatGPT is particularly useful for generating ideas to support academic work. To effectively integrate AI tools into education, educators and institutional leaders must collaborate to develop clear policies, guidelines, and updated curricula on the use of ChatGPT. These measures should aim to empower learners by helping them understand the appropriate uses of AI tools, their potential to enhance learning outcomes, and their inherent limitations.⁹

In the use domain analysis, significantly more CLS learners reported using ChatGPT in their coursework compared to DCLS learners. This difference may be attributed to the survey timeline: DCLS learners were surveyed between April and May of 2023, whereas CLS learners received the survey during the last quarter of 2023. The later survey date for CLS learners provided them with a longer period to explore and engage with the technology, potentially accounting for the higher reported use rates.

The potential utility of AI in health care field has been widely recognized, particularly in areas such as personalized medicine, drug discovery, and improving diagnosis and clinical decision-making through the analysis of large data sets.⁹ In health care research and clinical practice, AI has demonstrated benefits including enhanced scientific writing, increased research equity, more efficient data set analysis, streamlined workflows, cost savings, and improved health literacy.^2,9,13 Specifically within clinical laboratory practice, AI offers support in identifying abnormal patient results, optimizing test use, and developing testing algorithms. However, integrating AI into health professions education remains controversial because of the extensive breadth information and complex concepts learners must master. Despite this, AI presents several educational advantages, such as generating clinical vignettes of varying complexity, providing immediate, tailored feedback to learners, and supporting the development of communication skills.¹⁴

One of the major challenges associated with the use of ChatGPT and AI in health professions education is the concern over the quality and availability of the training data, which can lead to biased content or inaccurate information.¹⁵ Therefore, it is essential for educators to provide guidance to learners on how to use AI responsibly and effectively.

During the focus group discussion, CLS and DCLS participants shared their experiences after completing an AI-assisted learning project. Consistent with survey findings, CLS participants identified factual inaccuracies, potential bias, and the risk of disseminating misleading or misused information as significant drawbacks of using ChatGPT. These concerns highlight the importance of carefully reviewing and evaluating ChatGPT-generated content throughout the learning process.² Although ChatGPT may not always produce fully accurate or creative responses, it excels at generating text aligned with user intent.⁶ Its training on large data sets of human conversations enables advanced language processing capabilities, but its limitations become evident when addressing scientific accuracy, particularly in medical queries.³ Despite these challenges, ChatGPT represents a significant technological achievement, demonstrating the potential of deep machine learning.³ The integration of AI into health professions education offers promising opportunities to enhance personalized learning experiences while emphasizing the need for critical evaluation and responsible use.

CONCLUSIONS

As ChatGPT and similar AI tools continue to advance, they hold the potential to revolutionize health professions education. The study revealed that although CLS learners were more familiar with and used ChatGPT more frequently in their coursework, DCLS learners expressed greater optimism about its integration into professional education and practice. Participants highlighted concerns about academic integrity and the accuracy of AI-generated outputs, emphasizing the need for human oversight. To ensure ethical and effective use, the implementation of ChatGPT in CLS education must be approached cautiously, with clear policies and guidelines to address potential misuse and challenges. Although health professionals remain cautiously optimistic about AI’s role in clinical practice, they also recognize the irreplaceable value of human expertise in health care. A thorough evaluation of AI’s real-world impact is essential to mitigate potential risks and maximize its benefits.