How Artificial intelligence (AI) will change the CPD work of healthcare professionals.
By Vjekoslav Hlede, PhD, DVM
Artificial intelligence (AI) has been here for some time. Whether we like the idea of an AI-enhanced reality, whether we use ChatGPT or not, AI is here. We use it. Zoom recordings generate a summary and list of action items because of AI. As we type, our text editors masterfully provide grammar, style, and tone suggestions – because of AI. Digital navigation tools guide us flawlessly through busy streets because of AI. Even our old search engine provides great results because of – you know it, AI. AI-enhanced virtual assistants such as Alexa have become household names. Alexa, tell me more about AI in CME/CPD.
Dr. Som Biswas tells a convincing story on how ChatGPT will change medical writing in his recent article ChatGPT and the Future of Medical Writing. ChatGPT is a generative pretrained transformer language model. ChatGPT has been trained on a huge dataset of text available through the internet. It can create new text – comparable to the text used for training. While ChatGPT and similar language models can not replace human writers (yet), they can be of huge help. For example, peer review, literature review, creation of articles, and article summaries can all be enhanced by AI. Similarly, we can use AI to enhance curriculum development and CPD practices. Attached is an example of a curriculum developed by ChatGPT and directed by our colleague Robert Dantuono. While human direction played an essential role in the creation of that curriculum and quality assurance, each word written in the document is generated by AI.
So, where are we going from here and what can we expect in the future? Let’s ask ChatGPT…
Vjeko: Hello ChatGPT. Please provide an introduction to the ChatGPT technology and explain that this is written as an interview between Vjeko and ChatGPT.
ChatGPT: Hello Vjeko! Of course. I am ChatGPT, a product based on OpenAI’s GPT-4 architecture. I’m a state-of-the-art language model designed to understand and generate human-like text based on the prompts I receive. I’ve been trained on vast amounts of data and can answer questions, assist with tasks, create content, and engage in conversations. I can answer questions, assist with tasks, and engage in various types of text-based conversations. However, I don’t have consciousness, emotions, or self-awareness; I simply generate responses based on patterns in the data I was trained on. It’s essential to highlight that the contents of this discussion have been dynamically generated in response to prompts provided by Vjeko. Such a collaborative approach exemplifies the intersection of human direction and AI capabilities, culminating in the insightful exploration.
Vjeko: Thank you for that introduction and explanation, Chat. Tell me about the history of AI in medical education.
ChatGPT: Certainly! The use of artificial intelligence (AI) in medical education has a rich and evolving history. As with many fields, AI’s integration into medical training has been driven by the desire to enhance educational outcomes, improve efficiency, and personalize learning experiences. Let’s delve into the journey of AI in this domain:
1. Early Beginnings: Computer-Based Learning
Before sophisticated AI models came into play, computer-based learning modules were introduced in the 1960s and 1970s. These systems, although rudimentary by today’s standards, set the foundation for digital learning in medicine.
2. Decision Trees and Expert Systems
In the 1980s, expert systems became popular in medical education. These were essentially decision trees that helped students make diagnoses or understand treatment paths. They didn’t “learn” in the way modern AI does but were based on the inputs of medical experts.
3. Simulation and Virtual Reality (VR)
The late 1990s and early 2000s saw the rise of simulation-based medical training. With advancements in graphics and processing power, virtual reality (VR) and augmented reality (AR) began to find their place in medical training, offering students a hands-on experience without the risks associated with real-life patients.
4. Machine Learning and Personalization
The 2010s ushered in an era of machine learning. Medical training platforms started utilizing AI to analyze student performance data and adapt content to individual needs. For instance, if a student struggled with cardiology topics but excelled in pulmonology, the system would adjust accordingly, offering more resources and quizzes in the weaker areas.
5. Natural Language Processing (NLP)
With the rise of sophisticated AI models capable of understanding and generating human-like text, medical students could now interact with chatbots for Q&A sessions, clarifying doubts, or even simulating patient interactions. These chatbots, powered by NLP, offer real-time feedback and help in honing diagnostic or communicative skills.
6. Data-Driven Insights
Modern AI tools in medical education provide educators with insights into student performance, learning styles, and areas that may need curriculum adjustment. By analyzing data on a granular level, institutions can make informed decisions about educational strategies.
7. Continuous Evolution and the Future
The potential of AI in medical education is vast. From virtual dissections to AI-driven case studies that adapt in real-time based on a student’s actions, the future promises even more immersive and personalized learning experiences. Moreover, with the integration of wearables and biometric data, there’s potential for even more interactive and individualized training modules.
Vjeko: Thank you, Chat. Now, can you please speculate on how AI will transform the education of healthcare professionals in the next 5-15 years?
ChatGPT: Certainly! While predicting the future is always speculative, based on current trends and technological advancements, we can make some educated projections regarding AI’s role in the education of healthcare professionals over the next 5-15 years:
- Personalized Learning Experiences
With AI-driven analytics, educational platforms will be able to better understand individual learners’ strengths, weaknesses, and preferences. This will enable tailored curricula that adapt in real-time, ensuring that each student receives training optimized for their needs. - Advanced Simulations and Virtual Reality (VR)
The line between virtual and reality will blur further. Enhanced VR simulations will allow healthcare students to practice surgeries, diagnostics, and other medical procedures in a controlled and risk-free environment. These simulations will likely be adaptive, growing more complex as a student’s skills improve. - Intelligent Tutoring Systems
Chatbots and virtual tutors powered by AI will be available 24/7 to answer questions, clarify doubts, and provide guidance. These systems will be sophisticated enough to offer detailed explanations, reference current research, and even simulate complex patient scenarios. - Continuous Assessment and Feedback
AI systems will monitor a student’s progress in real-time, providing instant feedback. This can help in early identification of areas that need improvement and can suggest targeted resources or exercises to address those gaps. - Real-time Translation and Global Collaboration
AI-driven language translation tools will enable seamless collaboration among students and professionals from around the world, fostering a more interconnected and global approach to medical education. - Augmented Reality (AR) in Clinical Settings
With AR glasses or similar devices, students and trainees in clinical settings can see overlays of information, procedural steps, or even real-time guidance as they interact with patients or conduct procedures. - Integration of Biometric Feedback
Using wearables and other sensors, AI systems could assess a student’s physiological response during training sessions. For instance, monitoring stress levels during a simulated surgery could provide feedback about performance under pressure. - Up-to-Date Content Delivery
The medical field is ever-evolving with new research and findings. AI systems will constantly scour the latest journals, studies, and clinical trials, updating course material to ensure students are always learning the most current and relevant information. - Ethical Training Scenarios
With growing concerns about ethics in AI and medicine, AI-driven platforms might provide simulations that challenge students with ethical dilemmas, helping them navigate the grey areas of medical practice. - Enhanced Remote Learning
Especially relevant in the post-pandemic era, AI can optimize online learning environments, ensuring they’re as effective, if not more so, than traditional classroom settings - Interdisciplinary Integration
AI can help integrate knowledge from various fields, offering a holistic approach to medical education. For instance, a lesson on cardiovascular health could seamlessly integrate genetics, nutrition, and socio-economic factors, all curated by AI to provide a comprehensive view
Conclusion
The future of medical education, with the integration of AI, promises to be more personalized, dynamic, and responsive to both global and individual needs. As technology continues to advance, it will be crucial for educators, institutions, and students to remain adaptable and open to these transformative changes. AI’s journey in medical education has been marked by consistent growth and adaptation, so too will be the tools available for training the next generation of healthcare professionals.
Vjeko: Thank you for such a nice explanation, Chat. Do you have any closing statements for the readers of CE News?
ChatGPT: Yes, of course. I genuinely appreciate the chance to share insights and engage in this collaborative exploration alongside your guidance. As we delve into the intricacies and potentials of technological advancements, it’s evident that we stand on the precipice of an era where innovation shapes every facet of our professional and personal lives. From the discussions and insights shared within these pages, it’s clear that the path ahead is both exhilarating and challenging.
Remember, while technology, like AI, offers tools of immense power and potential, it’s the human spirit, our collective ingenuity, and our shared values that will determine how we harness these tools for the greater good. Let us move forward with a vision that balances progress with prudence, ensuring that as we shape the future, we remain grounded in the principles that have brought us this far. Thank you for your continuous engagement and commitment to staying informed. Here’s to a future of exploration, understanding, and growth!
Vjeko: That is such a powerful closing statement, Chat.
ChatCPT: I’m glad I could assist and provide meaningful content for this CE News article. If you have any more questions or topics to explore in the future, don’t hesitate to reach out. Wishing you all the best!
References
- Chat GPT and the Future of Medical Writing. Biswas, Som. Radiology, 2023 Apr;307(2):e223312. DOI: 10.1148/radiol.22331
- ChatGPT Is Shaping the Future of Medical Writing But Still Requires Human Judgment. Kitamura FC.Radiology. 2023 Apr;307(2):e230171. doi: 10.1148/radiol.230171. Epub 2023 Feb 2.PMID: 36728749. No abstract available.
- ChatGPT and Other Large Language Models Are Double-edged Swords. Shen Y, Heacock L, Elias J, Hentel KD, Reig B, Shih G, Moy L.Radiology. 2023 Apr;307(2):e230163. doi: 10.1148/radiol.230163. Epub 2023 Jan 26.PMID: 36700838 No abstract available.
- Logic and language in medical writing. BAUER J.Science. 1953 Jan 9;117(3028):40-1. doi: 10.1126/science.117.3028.40-a.PMID: 13028276 No abstract available.
- Overview of Early ChatGPT’s Presence in Medical Literature: Insights From a Hybrid Literature Review by ChatGPT and Human Experts. Temsah O, Khan SA, Chaiah Y, Senjab A, Alhasan K, Jamal A, Aljamaan F, Malki KH, Halwani R, Al-Tawfiq JA, Temsah MH, Al-Eyadhy A.Cureus. 2023 Apr 8;15(4):e37281. doi: 10.7759/cureus.37281. eCollection 2023 Apr.PMID: 37038381
Vjekoslav Hlede, PhD, DVM is Learning Management System Manager at the American Society of Anesthesiology, Chicago, IL.
Example: Using ChatGPT to Develop a Curriculum on Transgender Healthcare
Program Title: Providing Appropriate, Effective and Culturally Sensitive Care to Transgender Patients
By Robert D’Antuono, MHA, Reporter
ChatGPT Request #1: Write a continuing medical education “needs assessment” for a seminar on transgender health care.
Needs Assessment
Transgender health care is an emerging field in medicine that requires specialized knowledge and training. The seminar should aim to address gaps in knowledge related to the specific healthcare needs of transgender patients, including hormone therapy, gender-affirming surgeries, and mental health support. It should also address gaps in practice related to the provision of culturally competent care, including understanding the unique challenges faced by transgender patients and creating a welcoming and affirming healthcare. By doing so, healthcare providers can improve the quality of care they provide to transgender patients and help to reduce health disparities in this vulnerable population.
Target Audience
The target audience for this CME seminar includes healthcare professionals who may encounter transgender patients in their practice, including primary care physicians, endocrinologists, psychiatrists, psychologists, nurse practitioners, physician assistants, and social workers.
Learning Objectives
The following learning objectives have been identified as essential for healthcare professionals to provide quality care to transgender patients:
- Define and understand the terminology associated with transgender health care.
- Describe the epidemiology of transgender health care and the health disparities that exist in this population.
- Discuss the medical and surgical options available for gender-affirming hormone therapy and gender-affirming surgeries.
- Recognize and manage the physical and mental health concerns unique to transgender patients.
- Develop communication skills that promote trust, respect, and collaboration with transgender patients and their families.
Assessment Methods
To identify the knowledge gaps and learning needs of healthcare professionals related to transgender health care, the following assessment methods will be employed:
- Surveys: A pre-seminar survey will be sent to potential attendees to gauge their knowledge and experience in transgender health care. A post-seminar survey will be conducted to assess the effectiveness of the seminar in meeting the learning objectives.
- Planning Committee: A Planning Committee will be appointed of healthcare professionals who have experience working with transgender patients. This committee will provide valuable insight into the specific learning needs of healthcare professionals in this area.
- Expert Consultation: Experts and key opinion leaders in transgender health care will be consulted and invited to serve as program faculty to ensure that the seminar addresses the latest developments and emerging trends in the field.
Teaching Modalities
The seminar should include a variety of teaching modalities, including didactic lectures, case studies, interactive workshops, and small group discussions. This will allow healthcare providers to learn from each other and to apply their new knowledge and skills in a practical setting.
ChatGPT Request #2: Develop an agenda of topics for a 6 hour CME program on transgender healthcare needs and challenges.
AGENDA
An agenda of topics for a 6-hour CME program on transgender healthcare needs and challenges follows:
Introduction (30 minutes)
- Overview of the seminar goals and objectives
- Importance of transgender healthcare education for healthcare providers· Basic terminology and concepts related to transgender health
Transgender Health: Medical and Mental Health Needs (90 minutes)
- Overview of the medical and mental health needs of transgender patients
- Hormone therapy for gender transition
- Gender-affirming surgeries
- Mental health support and resources
Legal and Ethical Issues in Transgender Health Care (60 minutes)
- Overview of legal and ethical issues related to transgender healthcare
- Understanding informed consent and capacity to consent
- Legal protections for transgender patients· HIPAA and confidentiality concerns
Creating a Welcoming and Affirming Healthcare Environment (60 minutes)
- Importance of creating a welcoming and affirming healthcare environment
- Best practices for culturally competent care for transgender patients
- Strategies for improving communication and building trust with transgender patients
Addressing Health Disparities in the Transgender Community (60 minutes)
- Overview of health disparities in the transgender community
- Barriers to healthcare access for transgender patients
- Strategies for reducing health disparities and improving health outcomes for transgender patients
Case Studies and Small Group Discussions (120 minutes)
- Review of case studies involving transgender patients
- Small group discussions and problem-solving exercises
- Opportunity to apply new knowledge and skills in a practical setting
Conclusion and Feedback (30 minutes)
- Review of seminar content and key takeaways
- Opportunity for participants to provide feedback on the seminar
- Resources for continued learning and support
Note: This is just a sample agenda, and the actual content and timing of the seminar can be customized to meet the specific needs of the audience and the organization hosting the CME program.
ChatGPT Request #3: What specific metrics could be established for this seminar on transgender care?
Evaluation
Establishing specific metrics can help to evaluate the effectiveness of a seminar on transgender care and determine if the learning objectives have been achieved. Here are some metrics that could be established for this seminar:
- Pre- and post-seminar knowledge assessment: A knowledge assessment can be administered to participants before and after the seminar to evaluate the change in their knowledge and understanding of transgender healthcare needs and challenges.
- Participant feedback survey: A survey can be distributed to participants at the end of the seminar to gather their feedback on the seminar content, teaching modalities, and overall satisfaction with the program.
- Case study analysis: Participants can be given case studies to analyze and solve in small groups during the seminar. These case studies can be evaluated to determine if participants have developed skills in applying their knowledge to practical situations.
- Post-seminar follow-up survey: A follow-up survey can be sent to participants a few months after the seminar to evaluate if they have incorporated new knowledge and skills learned in the seminar into their practice, and if they have seen any changes in patient outcomes.
- Attendance and participation rates: The number of participants who attend the seminar, the level of engagement and participation during the seminar, and any barriers to attendance can be monitored to evaluate the reach and impact of the seminar.
