Creating impactful AI features in healthcare requires thoughtful integration, user-centric design, and a clear vision of how AI can empower users through automation and augmentation.
In several health tech projects we’ve been working on lately, we’ve observed a keen interest in incorporating AI features. However, there was often little consideration for how these additions needed to merge into the journeys and contexts of patients or physicians.
This can cause a disconnection between the addition and actual users’ contexts, and to counter this, we know we can apply our well-known user journey mapping activities. Journey mapping keeps the user at the center by being mindful of their activities in a more holistic way. It’s also intuitive for product and design roles and can help with integrating physicians or patients into the discovery phase.
And here is a small twist for the case of Transforming Healthcare with AI. We do three simple things that enhance this journey mapping:
- The first one is pretty straightforward: we add ML experts to the definition team. For each step in the journey, they ask questions about data needs and ownership. They can also make suggestions about data collection and feedback loops.
- We introduce beforehand the concepts of Automation and Augmentation and the differences between them.
- In a second round over the Journey, we do a mapping of opportunities for integrating AI in each step of the journey, making all of the participants think of this simple categorization about how AI can affect the user journey.
This way, they can start thinking not only about where AI can enhance an experience but also about different ways of affecting the experience. Let’s explore these concepts.
Automation and Augmentation
In simple terms, automation is about reducing manual effort, while augmentation is about enhancing human capabilities.
Automation focuses on taking over repetitive (and sometimes tedious and time-consuming), data-intensive tasks. This transformation is ideal for processes with high volume and frequency, where manual effort can be reduced without compromising care quality.
For instance, the paperwork needed to ask for a lab test and the EHR integration can be automated after a simple prompt from the physician. This is a moment of the journey where AI can streamline workflows, saving time and resources. But, is efficiency the only metric we should consider?
Augmentation is about empowering users with AI as a supportive tool. It involves AI applications that assist in decision-making processes, where human judgment is irreplaceable. Because of the nature of Health tech products, we will often face this situation when human control is necessary and even mandatory.
A good example of this can be tools aiding the physician to make treatment decisions: an AI can augment the information present in a medical image, by highlighting areas where it infers there can be a problem, and even present the level of confidence. This empowers the doctor, but his more holistic view, and his human judgment, are needed to go beyond what’s been detected.
Augmentation doesn’t replace the human element; instead, it enhances it, offering insights and support that lead to better patient outcomes. As you can imagine, augmenting is also more intensive in terms of discovery and design: the key to any effective AI integration lies in a deep understanding of end-user needs, goals, preferences, constraints, and the context in which they operate. But in the case of augmentation, you also need to think about how the ‘augmenting’ experience happens in the context of a user interface, altogether with short-term and long-term feedback loops for the ML model.
There are also more mixed scenarios. For example, there can be cases where automation can help you to solve 90% of the task but you need augmentation to cover the rest. The decision between automation and augmentation should be guided by the nature of the task, its impact on patient care, and the value it adds to the user experience.
Using this process to avoid fragmentation in Health Tech
As this discovery process means getting into the user journey and actual constraints, it could be key to preventing new AI applications from bringing more fragmentation to the already complex Health Tech space. We can guide AI’s integration into healthcare in a way that truly enhances, rather than disrupts, the workflows of healthcare professionals and staff members.
For instance, we are now working on a project to assist Social Workers and Care Managers in keeping track of patients’ SDOHs. During the discovery phase, we are also attentive to which tools from the practice they use to organize and track their work, asking direct questions. The main moments in their journeys are: assessing a patient’s needs, proposing interventions, documenting them, and following up. Integration with EHRs is key to keeping everybody in the loop, and we have to think about how each of these steps helps them reduce re-work with the EHR, communication with pairs and physicians, communication with the patient, and SDOHS actions reporting.
Sometimes this can even mean adding side-sessions to get into the details of reviewing an API and understanding what’s possible. But this is valuable, as it’s how we can ensure that the new wave of AI applications in health tech adds value, rather than clutter, to an already intricate domain.
Conclusion
The choice between automating and augmenting with AI in healthcare is not just a technological consideration but a strategic one. It’s about understanding and aligning AI capabilities with the real-world needs of those at the frontlines of healthcare. It’s about making AI more human-centered. And as decision-makers, it’s crucial to reflect on how AI can not just replace but enrich the human aspects of healthcare, leading to innovations that are both meaningful and impactful.
Want more in-depth information about our expertise? Ready to transform your health tech project? Reach out to us at hello@arionkoder.com, and let’s push the boundaries of what’s possible together!