What Patients Are Asking AI, and How Health Systems Can Help

 

img_linkedIn-amanda-todorovich

Much of the AI conversation in healthcare has focused on visibility. Amanda Todorovich of Cleveland Clinic sees an even bigger opportunity in helping patients make care decisions.
At a conference where AI came up in nearly every session, most of the talk centered on how health systems use the technology and whether they show up in AI-generated answers. Amanda Todorovich, enterprise executive director of digital marketing at Cleveland Clinic, wants leaders to look at the other side of the screen.

“I don’t think we have enough conversation about how patients are using it,” she said in an interview after the event. She’s focused on the behavior change AI is driving. She pointed to research showing that millions of people now make decisions about their appointments based on what an AI tool tells them.

For Amanda, that makes care guidance and navigation, meaning where to go and what to do in a moment of uncertainty, the issue health systems most need to strategize around.

 

The Missing Middle of the Patient Journey

Patients are asking AI practical questions that healthcare content hasn’t traditionally covered: What do I do now? How long can I wait? What if I don’t go?

Provider content, Amanda notes, covers diseases, conditions, procedures, who treats them and which services are available. It rarely addresses the decision in between. “For years, as marketers, we’ve focused on preference, brand and choice, but not on the process of choosing,” she said.

Orthopedics makes the gap concrete. Patients almost never go straight from an appointment to surgery. Physical therapy, X-rays, MRIs and insurance authorizations come first. Explaining that path, Amanda noted, is a natural next step for hospital websites, since those are exactly the questions people bring to AI tools.

The questions people actually bring to AI tools are personal and practical. How much money is in my account? How much gas is in my car? I have work tomorrow. Can I skip this? Urgent care or the ER? What will it cost? Answering them well, she said, calls for a new way of thinking that goes beyond traditional keyword ranking.

An Opening to Build Trust

Amanda sees this shift as an opening for providers. Patients are bringing deeply personal questions to AI, and health systems are well placed to help answer them.

Part of AI’s appeal is that it feels personal: the tools remember you, know you and speak in an upbeat tone. Health systems can pair that accessibility with something patients need just as much: credible, expert guidance they can rely on.

It’s also a chance to deepen the relationship between patients and providers. Honest, practical answers about cost, timing and next steps show patients that their health system is there to help them through each decision.

Providers also bring something AI can’t: clinical expertise. The average healthcare consumer reads at a fifth-grade level, has no clinical background and often uses the wrong terms, and a chatbot simply follows their lead. “It’s not a clinical decision tree or a diagnostic workup,” Amanda said. Clear content that helps patients describe what’s happening and ask better questions can close that gap.

Your Website Is a Source, Not Just a Destination

Amanda sees decision-oriented content as the next frontier for the industry. Today, most health system content explains services and treatments; the opportunity is to also support the decision to get care.

Building it means rethinking what a website is for. It is no longer only a place patients visit. It is a source that AI tools pull from, reshape and redistribute, rarely in the form you intended. The job is to make your main points as clear, direct and consistent as possible.

Structure and governance matter more than ever. How pages are built tells models what information you hold. When the same information appears in many places, Amanda noted, keeping it consistent helps AI tools give patients accurate answers.

That calls for content that talks openly about insurance, lays out the real steps to getting care and names the right place to start.

Rethinking the Health System Chatbot

Many digital platforms are launching their own AI chat. Amanda encourages health systems to think carefully about the role a chatbot would play, whether they build in-house or work with partners.

Health systems operate under privacy laws and clinical standards that shape what any tool on their website can do. They can’t offer direct medical guidance online the way consumer AI tools do. “We’re not allowed to remember you, follow you or give you the nurturing some people want,” she said. Those safeguards protect patients, and they also mean a provider’s chatbot will serve a different purpose.

As more AI tools connect to electronic health records, Amanda also sees an important role for providers in championing patient safety and privacy. Patients trust the health systems that hold their records, and that trust is worth protecting as these tools grow more widely used.
Patients’ habits matter too. People comparing health systems tend to use general-purpose AI tools rather than visit each hospital’s chatbot in turn. For Amanda, that makes clear, consistent content that informs those tools a better investment than one more bot.

Measure Decisions, Not Citations

The biggest opportunity, Amanda says, is going beyond visibility. “This isn’t just about being in a citation,” she said. The real question is whether health systems influence the care people seek, and whether patients reach the right care at the right moment.

Because no two people ask the same question the same way, she sees prompt-level visibility as only part of the picture. Business impact is the measure that counts most.

Here’s what Amanda says Cleveland Clinic is seeing:

  • Visibility: the system appears in AI answers more than any other in healthcare right now.
  • Organic traffic: a very slight decline, close to flat, with more than a billion sessions expected this year.
  • LLM referrals: referral traffic is coming in from AI tools, which she reads as a sign that patients feel some uncertainty and want to go deeper.
  • Rich media: rising engagement with video, audio and visual experiences, not just written Q&A.

That last trend points to an advantage AI tools can’t easily copy. A chat interface puts the burden on patients to know what to ask. A well-built website navigates for them, with a clear next click or a deep-dive page on a condition. “Asking questions and getting responses through a conversational interface is just one modality,” Amanda said.

What Health Systems Can Do Now

Amanda’s advice for leaders comes down to three moves.

  1. Audit your content. Know what you have, what it actually says and whether it answers the questions patients are asking. “You own your content,” she said, so keep evolving it.
  2. Write for the decision, not just the diagnosis. Cover insurance, the real steps to care and where to start. Answer “What should I do now?” alongside “How do we treat this?”
  3. Teach patients to use AI well. People will keep using these tools, and many don’t know what to ask. “So why don’t we tell them?” Amanda said. She sees room for health systems to work together against health misinformation and share practical tips for patients.

That work will matter even more in the year ahead. Amanda expects 2027 to bring added complexity in insurance coverage, in what is and isn’t covered, and across hospital, outpatient and out-of-hospital care. She sees a clear role for health systems in making those choices as easy as possible for patients.

Her closing charge is empathy. Patients are trying to stay healthy while keeping their jobs, affording appointments and getting decisions right with their insurers. Health systems that help them find the next best step will earn a place in the decision, not just the answer.

 

About Amanda Todorovich

Amanda Todorovich, FACHE, is Executive Director of Digital Marketing at Cleveland Clinic, where she has spent more than a decade building and leading one of healthcare’s most recognized digital programs. She leads a team of nearly 100 across web, content, social media, video, email, SEO, analytics and AI, reaching millions of consumers each month. Previously, Amanda co-founded MedCity News, a digital media company covering healthcare innovation, which was successfully acquired. She was named Content Marketer of the Year and inducted into the Healthcare Internet Hall of Fame for her contributions to the industry and continues to help shape the future of digital healthcare.

Contact Us Today!

Partner with CARAVAN to level up your content strategies.