Generative AI: Revolutionizing Healthcare and Advancing the Quintuple Aim
In this article
*Warning: This article was authored by a carbon-based, human life form.
In 1950, medical knowledge doubled every 50 years. In 2020, it doubled in 73 days…! Think you can keep up?
My 19-year-old daughter was home on break over the holidays in 2022, and she said, "I just used this thing called ChatGPT and asked it to create a high-protein meal plan for me. It did it in less than two seconds." That began my exposure to the art of the possible in healthcare. "Wait, could this thing write meaningful clinical encounter notes?"
The global healthcare sector has been evolving rapidly in recent years, and a significant part of this evolution is the advent of advanced technologies like artificial intelligence (AI), more specifically, large language models (LLM) or generative AI. This technology, particularly models such as ChatGPT, is creating tremendous excitement in the healthcare sector, with experts underscoring its potential to transform the industry and promote a more equitable system. This transformation promises to significantly impact the quintuple aim of healthcare, encompassing enhancements in patient experience, population health, cost reduction, the work-life of healthcare providers, and health equity.
In the context of patient experience, generative AI is primed to revolutionize healthcare delivery. Conversational AI, which includes AI-powered chatbots and customer service tools, can deliver quick and accurate responses, significantly improving patient interaction and overall satisfaction.
Furthermore, companies like Apple are now using generative AI to provide personalized health coaching. By integrating AI into wearable devices, they offer individualized health assessments that include sleep analysis, dietary suggestions, and even AI-generated music for stress management.
The upshot is a tailored healthcare experience that addresses each patient's unique needs. Generative AI has the power to transfigure the entire consumer healthcare journey. It would start by knowing one's specific homeostasis, environment, healthcare history, lifestyle choices, and such, course correcting through coachable empathetic conversation and communicating with the care team, care circle, and the patient for proactive intervention and health maintenance. Then it would assure the patient has a clear understanding of their next steps, access, referrals, appointments, medications, labs and studies, and the cost of care all in a language and experience level they can appreciate.
Generative AI has vast potential to enhance population health. LLMs like Google's PaLM-2, trained to answer medical queries, will increasingly refine their accuracy and utility, promoting health education and awareness at a population scale. Similarly, platforms like January.ai employ generative AI to predict individuals' glucose responses, empowering them to make informed dietary choices.
By facilitating these health-conscious decisions, generative AI can contribute to overall population health. LLMs may drive better outcomes for health systems as well. If trained correctly, these models can impact the entire population healthcare systems care for. For example, generative AI's ability to analyze large datasets and identify patterns can be crucial in identifying health trends within a population. This can enable healthcare providers to preemptively address health issues before they become widespread, further enhancing population health.
Healthcare is notorious for its escalating costs. Generative AI can significantly reduce this painful reality, particularly by addressing the challenge of escalating labor costs in the healthcare industry (their highest line item). AI can streamline and automate tasks typically handled by nurses, such as documentation, care coordination, and other administrative tasks. This creates efficiencies and allows clinical staff to work top-of-license and have more empathetic patient connections. All-in-all, happier staff, less turnover and happier patients all impact revenue.
Additionally, generative AI has the capacity to scale our most precious resources in healthcare, our clinicians. As it stands, we will always need more physicians, nurses and supportive clinical staff. We are in a significant deficit, and the system cannot produce and train them fast enough to match the growth rate of our aging population. By 2030, all baby boomers will be 65 or older; 20% of our people will need more chronic care management than ever before. LLMs will be able to circumvent this shortfall by bringing proactive and predictive medicine to the attention of various care teams so that they can prioritize their patients appropriately.
Additionally, LLM promise to make considerable strides in precision and personalized medicine, using patient data to generate curated, personalized treatment plans considering a patient's genetics, lifestyle and social determinants of health. This can improve treatment costs as well as patient satisfaction. Moreover, in the realm of drug discovery, a traditionally lengthy and expensive process, generative AI is proving to be a valuable asset. Companies like Evotec already employ generative AI in drug discovery, leading to promising clinical trials. These advancements hint at substantial cost reductions in the future.
Generative AI is not just beneficial to patients. It can also be an answer for healthcare providers. Emerging as indispensable co-pilots, generative AI tools automate tasks such as inbox responses, clinical note writing, identifying healthcare gaps and misses, diagnosis assistance, high-yield report generation, billing code automation, referrals, and prior authorizations.
This automation allows healthcare providers to focus more on direct patient care, improving job satisfaction, and reducing burnout. For example, DAX Express, Microsoft and Nuance's conversational, ambient and generative AI solution can listen to conversations and complete tedious clinical workflow documentation to alleviate much of the administrative tasks causing clinician strain.
One of the ultimate goals of healthcare is to achieve health equity, ensuring that everyone has a fair opportunity to attain their full health potential. Generative AI can democratize access to healthcare information and tools, breaking down language and accessibility barriers to make healthcare genuinely universal. For instance, in areas where healthcare professionals are scarce or access is a barrier, ChatGPT-4 could be the critical guidance. It can be the resource that shares the same information with a Mayo Clinic physician as a villager in Cuchillo Parado, MX — yet curated at a level each audience can appreciate.
As with any innovative technology, generative AI is accompanied by a set of challenges. Data privacy, bias in AI models, and establishing trust in AI utilization are vital concerns that must be addressed. As AI ethicist Stefan Harrer proposed, we require a robust ethical framework that includes principles for human oversight, data transparency, privacy protection, and accountability in training data and AI-generated content.
Despite the substantial potential benefits of AI and LLMs, regulatory challenges exist. The debate over regulating AI technologies continues globally, with the primary concern being who will take the lead in this process. Problems like privacy, bias and national security must also be addressed. In the U.S., there's competition among Congress, the Biden Administration and federal agencies to establish regulations, but progress could be faster. Meanwhile, Europe has taken the lead with the approval of the AI Act. These regulatory developments will significantly determine the extent and speed of AI and LLM integration into healthcare.
The successful integration of generative AI in healthcare must also address the issue of job displacement. It is crucial to emphasize that AI is an assistive tool rather than a replacement for human healthcare providers. We must approach this transition as an opportunity for role redefinition rather than job elimination. AI will not replace clinicians. However, those who choose not to embrace AI stand to be replaced.
Generative AI holds incredible promise for the healthcare industry, offering the potential to achieve the quintuple aim in healthcare. By harnessing its potential, we can enhance the patient experience, improve population health, reduce costs, better the work-life of healthcare providers, and move toward health equity. However, successfully realizing this vision requires great care to navigate ethical considerations, privacy concerns, and workforce implications.
As generative AI continues to evolve, we must ensure it is integrated responsibly into our healthcare system, maximizing its potential benefits while minimizing potential risks.
It reminds me of a cautionary tale — A Twilight Zone episode of an alien species that arrives on Earth with superior intelligence and benevolent motives. The humans saw them as altruistic, providing world leaders with advanced technologies to eliminate war and famine and solutions for free clean energy. They also left a book for government officials in their alien language. Cryptographers could only translate the title of the book, To Serve Man. After trust between the two species was established, humans were excited to board the UFOs to visit their planet. In true Twilight Zone fashion, the plot twist comes eerily precise: the book was not meant to help humans but to prepare them as a meal. It was a cookbook!