The CX universe takes on healthcare with agentic AI

MRIs, Vegas style

Where on earth can you get an MRI after sundown? Las Vegas, of course. Agentic AI enables it.

"We're a 24-hour town," said Rachel Papka, chief innovation officer at Steinberg Diagnostic Medical Imaging (SDMI). "We offer MRIs up to 11 p.m., and our schedule fills up."

Providing such extended service involves solving a complex data puzzle. Obtaining an MRI requires a physician referral, which must be verified against insurance coverage and the availability of the scanners. Yet patients who encounter delays in getting their MRIs can endure anxiety and potential setbacks in care.

Agentic AI hosted by Genesys has helped SDMI support 4,000 more calls per month than its contact center could previously handle. Call abandonment rates dropped from 10% before the agent to less than 3% currently. The company's virtual agent can access patient data and book appointments 24 hours a day, seven days a week, over the phone. During business hours, it takes pressure off the humans in the contact center by answering basic questions that agents otherwise would have fielded.

"We have a [lot of older patients]," Papka said. "Maybe they're calling in to make sure that they have the details right, or 'I need to cancel my appointment' or 'I need to see if you have my referral' -- just true status checks that, to me, don't need human intervention."

Tara Mahoney, vice president of Genesys' healthcare practice, said that in her experience, healthcare providers are driven by results. They care about leakage -- when patients go outside of their networks to seek care -- and eliminating no-shows, improving physician utilization, closing care gaps and other issues that generally boil down to workflow problems. And they don't necessarily care about which technology fixes them.

"In healthcare, tech starts and ends with workflows," Mahoney said. "No nurse ever wakes up in the morning and goes, 'You know what? I need an agentic AI platform to help me.' But if you talk about post-care workflows, if you talk about triage workflows, they are very much interested in how to make those workflows better."

Personalizing care, sales processes

One of the root causes of the current government shutdown is a dispute over healthcare funding and Affordable Care Act (ACA) subsidies. One of the companies on the front line of coordinating healthcare for patients on Medicare, Medicaid and ACA plans is Adobe Population Health, based in Phoenix.

The company not only connects patients and caregivers but also addresses social determinants of health (SDOH), how the healthcare industry describes a host of issues -- such as lack of transportation and food insecurity -- that can derail good healthcare outcomes.

"If you can't buy food, you're probably not going to buy your medications," said Alexander Waddell, CIO at Adobe Population Health. "We find that if we help people with those things, then they actually can take steps to improve their health and their quality of life."

Yet SDOH support is woefully underfunded through Medicaid, state and local government programs and private entities. Companies such as Adobe Population Health hope that agentic AI can create efficiencies that stretch that funding further. A Salesforce-based agent is being developed to support social workers, nurse practitioners and registered nurses as they conduct detailed, hour-long risk assessment interviews with patients. The agent will handle clinical documentation during the interviews, allowing field clinicians to focus on the patient and design more personalized care plans than before.

"Putting data into a system isn't going to solve somebody's problem," Waddell said. "It's going to be that human touch that's going to make a big difference in people's lives."

While customers are building their own AI agents, Salesforce, too, is working on agents for both its Healthcare Cloud and Life Sciences Cloud that automate processes common to those industries.

"A year ago or so, when Agentforce was starting to come to light, we took a step back and asked ourselves, 'What if we could give digital labor industry expertise?'" said Jeff Amann, executive vice president and general manager of Salesforce Industries, the company’s unit that creates custom clouds for verticals. "We have tons of it wrapped up in our industry applications."

Healthcare and pharma are two industries where agentic AI can be effectively applied, Amman said, adding that the average hospital healthcare system has 78 different data systems -- an oft-cited statistic based on a University of Pittsburgh Medical Center case study that might be understating the complexity of U.S. health IT. He also referenced research that 36% of the world's data is healthcare-related, with some estimates that 80% to 90% of it is unstructured.

Agentic AI can be a valuable tool for tackling data sprawl when strategically applied to bottom-line business goals. Haleon, a British consumer packaged goods company encompassing various brands such as Sensodyne toothpaste, Advil pain reliever, and Centrum vitamins, was spun off from GSK and Pfizer in 2022. The Salesforce customer plans to deploy agentic AI to better inform salespeople about what products their individual pharmacies and dental customers should carry based on their clientele's needs.

Currently, that process requires sales reps to perform time-consuming research ahead of sales calls, said Kelly Ellis, head of commercial tech at Haleon. Generative AI will not only do it faster but also give them richer data for more personalized customer plans.

"Agentic AI, as part of [our] platform, is going to help us really condense all of that down so that it's a much more seamless experience for the reps so they're much more efficient," Ellis said.

Embedding agentic AI in the patient experience

Tucked among the headlines from massive companies investing billions in data centers as they race to commoditize generative AI -- or create artificial general intelligence, depending on who you're talking about -- was a bombshell healthcare acquisition, albeit on a smaller scale: Qualtrics bought Press Ganey Forsta for nearly $7 billion. The number was substantial, considering SAP acquired Qualtrics itself for $8 billion in 2018, spun it off in 2020 and sold its ownership stake for $7.7 billion in 2023.

Healthcare performance improvement company Press Ganey had built a longstanding business on patient quality surveys, including the Hospital Consumer Assessment of Healthcare Providers and Systems (HCAHPS), to which both government and private insurers tie healthcare provider and hospital reimbursements. The resulting data analytics inform healthcare providers on how they can improve their care and patient outcomes. Forsta, which Press Ganey acquired in 2022, is a customer and employee experience survey and data analytics platform.

Qualtrics remains a customer experience measurement platform that has recently embraced agentic AI orchestration and expanded into marketing AI and A/B testing. Acquiring Press Ganey Forsta not only is an instant acquisition of healthcare customers, but also a wealth of healthcare data that makes it an instant player in healthcare.

Press Ganey Forsta's data set can power agentic AI use, said Nick Apeland, executive vice president of regulated Industries at Qualtrics Labs.

"If we can build a longitudinal view of the patient -- not just that specific measurement you get from a paper perspective, but from birth to death," Apeland said, "…We can start to be more predictive, take action, which makes it better for caregivers, makes it better for patients and our customers."

Apeland envisions agents pulling patient experience data to help customize future episodes of care, i.e., maybe a patient has expressed a need for help to be met at their vehicle in the parking lot and walked into the doctor's office. AI agents, in theory, can uncover and pass on such details before an appointment.

Qualtrics is currently working with Stanford Health Care to determine agentic AI use cases in the healthcare setting, such as language support for care, eliminating conflicting care instructions and others. In this way, data can help ease stress on patients as well as uncover solve problems in the existing information that they receive.

"AI allows us to pre-configure all of that throughout the whole journey, automate it and update it," Apeland said. "Rather than us having to ask you every time, let's build a model that knows what the patient's preferences are. Then let's automate the actions we should take so we're being proactive -- versus reactive -- using agentic capability."

Matching patients to financial help

If you pay attention to most prescription drug ads on television, you'll typically hear a line at some point that says, "If you cannot afford your prescription, you may qualify for assistance."

It turns out that a significant amount of back-end IT support is involved in matching patients to those assistance programs, enrolling them, sorting out insurance benefits, and -- maybe most importantly -- ensuring they adhere to the prescription regimen to derive the benefits of the medicine.

Agentic AI can make these processes much more efficient, said Sherilyn Tuthill, partner and life sciences patient engagement practice lead at Deloitte Consulting. A lot of manual work goes into finding patients who have been prescribed a particular drug. Refreshing data from insurers and healthcare providers is one area where generative AI can process much faster than previous systems. Agents can also match patients to pharmacies that are not only in their area but also have the inventory on hand.

Saving time in these cases is not just a convenience for the patient but can drive healthier patient outcomes.

"[Drug] manufacturers are really trying to make sure that they get patients onto therapy as fast as possible, to have the best possible outcomes," Tuthill said. "We're seeing a lot of applications of agentic in this area of patient services and patient support within these manufacturer programs."

Salesforce has long been active in the life sciences sector, Tuthill said, as pharmaceutical companies have taken their CRM for patient services in-house in the last decade. The company's push into agentic AI has been a catalyst for the industry to explore new methods to deliver medications to patients more quickly after a prescription has been written.

Last month, Deloitte began adding prebuilt agents to the Salesforce AppExchange, which will work within Salesforce's Life Sciences Cloud, as part of a $2 billion investment that includes a suite of agents, with additional agents to be released over the next six months. While some customers will jump in with those ready-made agents, others will take a longer-tail approach.

"Some are taking more of a pilot approach: 'Let's start with use cases, see how the adoption goes,'" Tuthill said. "Different organizations are taking different approaches -- one or two that I know of are having more of an enterprise-level, 'We want to review and approve everywhere AI is getting used,' versus others are letting it be more bottom-up with more of an approval process on the fly."

Don Fluckinger is a senior news writer for Informa TechTarget. He covers customer experience, digital experience management and end-user computing. Got a tip? Email him.