AI and the digital patient journey: What healthcare leaders said at Reuters Digital Health 2025

• Ambient and Agentic AI are hot technologies, reflecting widespread use and acceptance of large-language models (LLMs)
• AI applications are not evenly adopted along the digital patient journey, with slower adoption of AI technologies in the treatment and diagnosis phases
• Change management solutions are sorely needed to realize the full potential of AI technologies across the digital patient journey

As AI technology continues to evolve, healthcare systems are increasingly adopting AI applications as part of a digital patient journey, aiming to enhance patient experience, improve operational efficiency, and reduce clinician burden. Recently, Healthcare Transformers attended Reuters Digital Health 2025 in Nashville, a conference in which vendors and healthcare systems came together to talk about what is and isn’t working right now in digital healthcare. While the promise of ambient and agentic artificial intelligence (AI) featured prominently in discussions, the barriers to implementing other types of AI technologies seemed to be on everyone’s mind.

To gain deeper insights into these perceptions, Healthcare Transformers conducted a two-part interactive data collection session with 36 C-suite and upper management healthcare leaders at the conference. Their collective wisdom provides a unique perspective on where AI is currently making the biggest impact, where it has the greatest untapped potential, and where it faces the biggest implementation challenges across the digital patient journey. Participants also reflected on the AI solutions that are currently on the market, assessing their strengths and weaknesses in addressing modern healthcare’s critical challenges.

The patient journey in modern healthcare systems typically encompasses several key stages:

1. Prevention/Wellness
2. Symptom/Concern (awareness of a problem)
3. Primary Care
4. Specialist/Testing
5. Diagnosis
6. Treatment
7. Management/Monitoring

In the first part of the session, participants were asked to engage in a “human barometer” exercise. After asking each question, participants moved to the part of the patient journey they thought best answered the question—a kind of physical survey. Once there, we spoke with the participants to understand the “temperature” in the room, and participants reflected on the reasons for the distributions.

Participants moved to the parts of the patient journey that take place outside of the clinical setting, either before or after the clinical encounter. When probed as to why the room gravitated towards those parts of the journey, participants noted the at-home monitoring devices that are shifting disease management away from the clinician and towards the patient.

AI applications that they mentioned as part of this rationale included smart watches, mobile apps, and self-testing and self-monitoring technologies. Participants hypothesized that patients currently seem to be more comfortable utilizing AI technologies when they are in control of their use.

Results showed that participants felt the diagnosis and treatment part of the journey held great promise with AI solutions, but these were not being implemented to their full potential. Participants voiced that barriers around trust prevented both doctors and patients from utilizing AI solutions, citing relatively slower adoption for technologies like AI clinical decision-making tools and AI-enabled diagnostic technologies.

For this question, there was a wider distribution of responses. In keeping with the previous question, most participants felt that the diagnosis and treatment part of the journey suffered from a lack of trust, contributing to reluctance for healthcare systems to implement AI technologies. However, others argued that the biggest problems were in connecting individual prevention/wellness data collected through consumer digital health apps with primary care services. Still another group felt that a lack of resources in primary care is prohibiting wider adoption of AI technologies that could increase the number of touch points between patients and providers.