Can GenerativeAI really revolutionize the healthcare industry and improve patient outcomes? In our latest webinar, Decimal.health brought together healthcare and industry leaders to discuss real-world use cases of where GenAI can improve patient access and beyond, and talk candidly about what boundaries and governance are needed to make sure it’s used appropriately in a health system.
Finding the right unmet needs for GenAI
While GenAI is being hailed for its massive potential to transform healthcare, we need to make sure any technology is applied to the right problem in the right way in healthcare. To do this, our panelists all emphasized the importance of aligning the needs of healthcare systems with GenAI being developed in partnership with industry.
The panelists outlined several key ways they are identifying suitable unmet healthcare needs for GenAI and making sure they can safely test models:
- Health systems should co-create, and not innovate in a silo: Using participatory engagement methods to engage a range of internal and external stakeholders, including healthcare providers and industry, in the decision-making process for identifying unmet healthcare needs, assessing opportunities for GenAI and the design and development of any resulting technology.
“I think one thing that health systems struggle with is aligning the right group for the conversation about using AI. The conversations have to be participatory and leverage different expertise. The other thing that is very important is the whole concept of co-creation with potential users of the solution.”
- Start with patient-oriented thinking: Identifying suitable use cases through design based thinking approaches that put the end user (i.e., patient, healthcare provider) at the heart of the problem identification process
“We are taking a problem-based approach using design thinking. What are the personas? And what are the major challenges those personas face? Can we line up the right use cases for GenAI?”
- Testing in the proof-of-concept stage is an essential building block right now for GenAI: Conducting low-risk, time-bound testing of use cases in sandbox environments. Using sandboxes creates a safe environment for testing different models, protecting confidential data and organizational knowledge, and exploring the risks of operationalizing models before pushing the model live.
“One of the things that we’ve done when we advise and help build roadmaps and use cases is to get your sandbox stood up. Your sandbox is going to protect whatever you are doing and protect your confidential data.”
- Do no harm: Setting up suitable governance structures and guardrails for the introduction of GenAI into health systems and patient care. The panelists highlighted that healthcare systems need appropriate and proportionate governance to ensure GenAI is responsibly used and that all players are placing patient safety and confidentiality at the heart of all decision making.
“There is such a big journey to actually bring GenAI to our health systems. And the first part starts with governance. Do we have the right governance to be able to evaluate the right technology and trust the right technology?”
Real-world use cases for GenAI
Our panelists shared some use cases for GenAI in healthcare that they have been exploring, from using it as a digital ‘front door’ for patient access to enhancing patient engagement to helping with clinical care.
Our panelists saw a real opportunity in using GenAI to progress what, in some health systems, conversational bots are already doing to help guide patients to the right point of care and back out again. This may be using a GenAI enhanced bot to help patients navigate to a specialist, enable them to understand their own health data, and answer common logistical questions.
“[Healthcare organizations have used] conversational AI chatbots on their website to guide you to finding the right doctor and making selections… There are some organizations that are starting to implement GenAI in this space – where can we use it, for example, for personalized content generation or drive better patient engagement by providing tailored information back based on their questions.”
Another area of interest was exploring what GenAI can do to better enable data analysis and actionable insights from patients’ electronic medical records (EMRs). While there is a wealth of data within EMRs the panelists were cognisant of the need to set-up robust security and confidentially approaches when training GenAI on this type of sensitive data.
There was consensus across the panelists that GenAI could show real impact and value in health systems by supporting logistical and operational tasks such as appointment scheduling, tracking medical equipment, and staff roster management.
While the allure of clinical use cases may be strong, starting with non-clinical back-office tasks can lead to significant efficiency gains and cost savings for health systems. Additionally, focusing on less glamorous, but essential tasks, allows healthcare organizations to build, test, and learn with GenAI safely while demonstrating the value add it can have to both healthcare leadership, clinical staff, and patients.
Additionally, the panelists highlighted the importance of working with medical bodies, such as the American Medical Association (AMA), to develop or use existing frameworks to measure the impact of GenAI in certain use-cases:
“We need a model to prioritize – there are so many use cases – working with the AMA… [for example] on how we measure and holistically look at the impact on patient outcomes, quality metrics and return on investment.”
Governance for existing AI techniques, such as Natural Language Processing, is still a work in progress in a lot of healthcare organizations. Plus national regulation from the Food and Drug Administration (FDA) for the use of GenAI in healthcare is forthcoming. Therefore, the growing interest in GenAI poses a big governance challenge for healthcare leadership.
Our panelists have all played a part in deciding on governance structures for AI in healthcare and are already turning their attention to how to apply that to GenAI. The areas of governance at the top of mind for the panelists include:
- Setting up advisory boards representing a diverse range of views that are part of the GenAI use case identification and product design, development and implementation (participatory engagement techniques)
- Acceptable use policies for where GenAI (or other AI techniques) can be used in their healthcare organization and with patient data / patients themselves
- Educating healthcare providers and staff within the health system on what GenAI is, how and when it should be used, and what the risks are
- Updating internal information security, compliance and data handling policies with considerations for GenAI (or other AI techniques)
- Designing an ethical framework to ensure GenAI and the data it is built on is ethically responsible and does not introduce bias or potential harm to the patient
- Hiring into the healthcare system senior leadership a “Chief AI advisor” or similar who is on the pulse of this new technology. This role should provide an informed viewpoint on senior decisions as a formal part of the organizational structure, as well as foster a culture of transformation and responsible AI adoption.
Generative AI holds immense promise for revolutionizing healthcare, and our webinar provided valuable insights into harnessing its potential responsibly. By developing AI solutions in line with healthcare needs, identifying the right use cases, and implementing robust governance measures, healthcare organizations can usher in a new era of patient-centric care and operational efficiency. As regulatory environments continue to evolve, it is crucial for healthcare organizations to embrace AI thoughtfully, using it as a tool to drive positive healthcare outcomes for healthcare systems and patients.
Thank you to our panelists
Thank you to all of our panelists for sharing their experience of and expectations for GenAI in healthcare. Learn more about our experts below.
Kassandra Karpathakis (moderator) is an Engagement Manager at Decimal.health. She was previously Head of AI Strategy for the English National Health Service (NHS) where she spearheaded strategic direction and international engagement on the use of AI in health systems.
Lisa Esch is the Head of Industry Solutions for the Healthcare Provider business at NTT DATA Services. She has over 30 years of experience working in the healthcare services industry with an extensive background in digital innovation, transformation, product commercialization, and sales and marketing.
Chris Boyer is a results-driven digital strategist with a focus on enhancing patient experience through innovative digital solutions. He has worked for Beth Israel Lahey Health, M Health Fairview and Northwell Health and is also a nationally recognized digital thought leader, speaker, and author.
Joyce Oh is the Vice President and Chief Information Officer (CIO) at Moffitt Cancer Center. She is responsible for leading Moffitt’s information technology teams and contributing to the organization’s digital transformation efforts. Joyce brings more than 25 years of technology experience to Moffitt. She joined the cancer center in 2022. Before joining Moffitt, Oh served as the divisional chief information officer for Beaumont Health in Michigan.
Ashish Atreja, M.D., oversees UC Davis Health’s expansion of its digital relationships with patients and with other hospitals, bridging the gap between IT, academia, research, and innovation as “Digital Davis” becomes a global hub for digital health. Prior to his UC Davis Health appointment, Atreja, a gastroenterologist, served as the Chief Innovation Officer for the Department of Medicine at Mount Sinai Hospital.
Work with Decimal.health
Want to enter the health AI space? Reach out to a Decimal.health to learn more about Decimal.health’s expertise and how we can help you develop a go-to-market strategy and grow your business.
Decimal.health provides end-to-end solutions that help companies navigate the complicated healthcare ecosystem and become a market leader rather than a market follower. The team brings several decades of deep experience in digital health, provider, and payer systems to help digital health businesses develop strategies that accelerate the launch and growth of their business and maximize their impact. Decimal.health combines relevant clinical and care delivery experience with the technology needs of our care providers and patients to create useful, usable, and enjoyable product experiences.
There’s no doubt that AI is making waves all around the world, including in healthcare. At a special roundtable discussion at ViVE23 held by Wysa, Decimal.health, and Hop Labs, health tech leaders came together to discuss the future of ChatGPT in healthcare. We invited participants from health.
Digital health is becoming more mainstream within health systems, especially after the COVID-19 pandemic and now staring down the impending labor shortage crisis. However, the adoption, implementation, and scale achieved by these technologies across use cases and types of health systems (and even.