Finding the right mix of staff and technology that keeps patients coming back

Before health care providers hand any part of the patient relationship to artificial intelligence (AI), they should make sure their basic digital tools work as promised, according to Accenture's new patient loyalty research.

Just 55% of patients said their digital interactions with a provider were “very easy.” Comfort with AI is even more limited: Fewer than 30% of patients, even those who trust their provider, are comfortable with that provider using AI to answer health questions or help plan treatment, according to the survey.

Sarah Sinha, a managing director at Accenture, is the lead author of the report “Turning patient uncertainty into loyalty,” based on a survey of 10,000 U.S. patients, and leads the company's patient access and experience work in North America. Here she continues her discussion about digital quality, trust and how practices, especially smaller ones with limited budgets, should think about investing in new technology, including AI.

This transcript has been edited for length and clarity.

When patients rate their experience as “very easy,” loyalty is strong, with almost 90% saying they'll stay with their provider. But that loyalty drops sharply even at just “easy.” What's the way to make things “very easy” for patients?

Sarah Sinha: That was one of the more surprising findings, particularly when we looked at the specific words we used in the survey: Was it very easy, or was it easy? You'd probably think easy is good, and you should feel great about that. But you see a precipitous drop-off. I think it's partly because there are so many different points of interaction, and so many points of potential friction, that make an aspect of the journey cumbersome. One point in the journey may have been exceptional, and then another point falls off, and in total that makes the journey more cumbersome.

It's a lot to manage, and this influences trust a lot too. Sometimes providers get overly focused on specific interactions or tools and don't zoom out to look at the broader journey. You can't have one point work really well and everything around it not work, because the patient will experience all of those things across an episode of care, and all of it will influence the outcome. Some of it was great, some of it not so good, and maybe that averages out to that middle bucket of experience.

Beyond the physician, what's the role of staff in making that journey as easy as possible for patients?

Sarah Sinha: It's pretty significant, because almost everything that lives around the actual care episode will be handled by, if not a digital channel, then front desk staff, clerical staff or other support staff. All of those pieces together have a huge role in how patients experience the whole journey. One thing that's been clear in the survey for many years, so much so that we almost forget to mention it, is that when patients switch providers because they're unhappy, meaning not for usual churn like moving to a different city, those are overwhelmingly experience challenges, not a feeling that they didn't get good clinical care. That's been true for a lot of years in our survey, and it still surprises people. Experience drivers are really the most important thing beyond a terrible clinical care experience, and those aspects of the journey are handled by front desk staff and everyone else involved. They're all playing pretty critical roles.

You've advised organizations to improve their existing digital tools before adding new ones. Can you explain that?

Sarah Sinha: We're in an experimental phase of very rapidly evolving technology, with a lot to figure out, both with the evolution of AI and with dramatic improvements in automation and basic digital capabilities. The point we're trying to make is that rather than experimenting with pieces and parts of four or five digital channels, providers should have one foundational digital channel, their web and portal, and maximize that first. Go for full enablement of everything a patient could do by picking up the phone and calling, and bring that primary digital channel as close to parity with the live experience as possible.

What we often see with web and portal capabilities is two big gaps between that and the live experience. One is pure enablement: Physicians may have opted in or out, so when a patient goes to that digital channel, they may not find their provider at all, or only for a subset of their visit types. There's usually a large gap between what's enabled digitally and what you'd find if you picked up the phone or showed up in person. Beyond that, there's a gap between enablement and conversion. Digital adoption numbers are typically extremely low in the industry today, which speaks to whether, once a patient started something online, they could actually finish it. When patients run into friction and have to abandon that channel, that's the gap to conversion.

AI-enabled tools for scheduling and messaging are increasingly marketed to physicians, yet fewer than 30% of patients, even those who trust their provider, are comfortable with that provider using AI to answer health questions or help plan treatment. How do you advise practices to integrate AI, and for what tasks?

Sarah Sinha: Really thoughtfully. That was a surprisingly low number. If you also think about how many patients are using AI on their own to navigate health care, the way they might have previously used a Google search, they're using it, but they're holding their providers to a higher bar of accountability. What was really important to us in that finding is that providers be thoughtful about how they roll it out. We have to get past the point of experimenting with AI to make sure it works and that it's accurate, and find a way to sequence and scale those capabilities more safely and reliably.

The first use case probably shouldn't be direct interaction with patients. The earlier use cases should probably be operational support in a call center, for example, things like agent assist, getting staff to knowledge articles, job aids and workflow rules faster, so the agent is more efficient on the call. Analytics and operations support let you practice and scale the internal support functions before moving to direct patient support, while measuring accuracy over a large volume before increasing the complexity of the use cases you're supporting.

What should smaller, independent practices with limited budgets look at first when it comes to technology investment?

Sarah Sinha: For smaller groups with limits on the types of capabilities they can invest in, I'd recap two things. One is to maximize the technology you already have before expanding to something new; we tend to find those capabilities to be suboptimized or underutilized. The other is the operational streamlining that makes technology work better, which helps front desk staff be more efficient, which is lower cost, and it also increases the patient experience.

One challenge for a smaller practice is that, on the plus side, they have lower complexity and tighter control over what happens, but they're dealing with a smaller complement of staffing, so sometimes someone will answer the phone in 10 seconds and other times it will take five minutes, depending on callouts or peak times. The more simplified and streamlined you can make your workflows, the more that also allows for adopting automation capabilities that are likely to be rolled out in your electronic medical record (EMR), so you may not have to invest in new technology, but can adopt those capabilities as they come along if your workflows are ready for it.

What didn't I ask about that you'd like to emphasize from the report and its findings?

Sarah Sinha: I'd emphasize that experience factors are critically important, probably more important than you'd think, and that patients will make decisions based on them as the care delivery model shifts. There are more settings, or other types of options, where a patient can receive routine care, and that makes the whole landscape more competitive, and makes these experience factors even more important. If I'm deciding between a doctor's office, an urgent care, a telehealth visit and other settings that might meet my need, the one that can serve my experience needs, the things that are most important to me, first or best, might be the one that wins my loyalty over the long run.