Patient access company Luma Health debuts generative AI platform

The first two products are an automated fax processing tool and a patient-facing voice AI concierge

Luma Health is an engagement platform for patients to help them better connect with their doctor both before and after an appointment. It does that by offering, among other things, smart scheduling capabilities, so that patients can self-schedule 24/7 or confirm, reschedule, or join a wait list from a text message, as well as referral management, meaning that it will automatically prompt referred patients to self-schedule. Patients also receive confirmation via their communication method of choice, and referred providers can track their status. 

Now the company is ready to make its platform even more efficient as it announced the release of its new EHR-integrated AI on Tuesday, called Spark, which is being woven into the Luma platform, enabling it to transform challenging parts of patient access. 

The company also revealed the first two Spark-enabled products: Fax Transform, which automates fax processing, and Navigator, which provides a patient-facing voice AI concierge. The idea behind both products is to help save staff time and improve patient access to care.

"When we launched the company, it was simple: needing healthcare can be hard, getting healthcare should be easy. So, how do we really unify and automate the entire patient journey? That means everything from access, communications, and readiness, which is how a patient actually gets prepared for care, to all of those different touch points along the patient journey. That's what we really have been focused on, and now 10 years into our journey, we're able to do that at scale," Adnan Iqbal, co-founder and CEO of Luma Health, said in an interview. 

"Spark isn't just incremental improvement, or little modifications. It's really, how do we embed workflow automation into every single one of these products and do this at scale? You have core systems of record, what we call the health IT stack: your EHR, your call center, your CRM, which almost always means Salesforce, Google, you have a revenue cycle and payments capabilities, and so everyone has these systems of record," he said.

"What's really missing, and what the market's pulling for is a true healthcare consumer or patient-based system of action, and that's us."

Navigator and Fax Transform

The phone is still very often the front door of care, with 76% of patients calling their doctor's office; for example, if they need a prescription refill, if they need to schedule an appointment, if they need to change my appointment, if they need to get a better result in my lab results, or they have some questions, this almost all done on the phone, which is in efficient and takes a lot of time, both for the patient and staff.

What Navigator provides is real time, automated, context aware assistance in the patient's native language. So, instead of having one-on-one conversations between the patient and the care team, Navigator allows patients can call about their own appointment, then follow up on their child's appointment and request a medication refill in the same interaction. They can also ask questions about how to prep for an upcoming procedure, then get a link with directions to the clinic. Or they can receive a text message back if their call is dropped, allowing them to continue where they left off.

It's also omni-channel, meaning the patient can start on their phone over text and then continue over voice, or can go log in from a computer.

"It's meeting the consumer where they are, and it's having a context aware set of conversations that's not just limited to a single topic. You can start with scheduling and quickly go to your diagnoses or go to questions you have, talk about how you're going to pay for it. It's fully intelligent," said Iqbal.

"That's a lot of power, because today, truth be told, the real front door for healthcare is still very much the call center. Now you can do it in a very sophisticated, thoughtful, intelligent way that automates all those different touch points."

The first academic medical center using EPIC that launched Navigator was the University of Arkansas for Medical Sciences, which has a cancellation line where people can call in to cancel; this typically happens over the weekend, because that's when people remember and have some time to call. That means, on Mondays, when the team comes in, they face three to four hours of voicemails that they would have to listen to and then mark as canceled in EPIC.

With, Navigator,  the conversation with the patient over text or the phone is automated, so it will make sure that that patient gets that appointment canceled; then, instead of that being a dead end, it then helps, through a conversation, to make sure that patient is rescheduled, meaning it automates a tremendous amount of staff work, because all that for staff would ever be multiple systems, clicking, listening, marking, and then calling the patient back.

Fax Transform, meanwhile, helps automate scheduling: when people need to get higher end care from a specialist, most of that is done via fax, which can literally mean having humans who's job it is to print out faxes, type things in, and call patients to schedule them.

"Faxes are all noisy, they all look different, and so this is a really, really good use case, a real problem that AI can actually help meaningfully transform. It took about five to 10 minutes of manual staff time per fax, with a lot of human error. Also, there's a tremendous amount of backlog. It just takes a lot of time, not just days or weeks, but often months. Any leading health system at scale will have a massive call center that's processing months and months and months of referral orders," Iqbal explained. 

Fax Transform uses Luma's Spark AI capabilities to automatically parse structured data from faxes including referrals, DME requests, and prescription refills. DENT Neurologic Institute saw a 3x faster fax processing and 70% time savings on fax workflows using the product. 

Meanwhile, the University of Arkansas for Medical Sciences (UAMS) saw 98 manual call center staff hours saved/eliminated, 95% of phone calls completely automated, 82% patient verification success rate, and 1,200 cancellations processed without any human interaction by using Navigator. 

Spark has really been able to drive ROI in three ways, Iqbal explained, with the first being that it's far more effective at delivering an easy to use healthcare consumer experience that drives up every version of scheduling, which he calculates as resulting in a 25% lift from embedding Spark into access solutions.

The second part of ROI is time saved in terms of staff productivity: the company is able to save four to five hours per staff member, so an 80% increase in staff productivity. Then the third part is the differentiated consumer experience.

"Everyone is operating in a competitive environment and if everyone uses the same core foundation, they all have Epic as their EHR, Cisco as their call center, and Salesforce as their CRM, if everyone looks and feels the same, they're not able to differentiate. And really, that's where we've seen people really move the needle and deliver just true consumer delight which hasn't been a theme in healthcare," Iqbal said.

"We're just making this so easy, where someone is able to get the care they need when they need it, how they need it, with just the channels that they use, really meeting the consumer where they are, like we enjoy un every other part of our life."

The future of AI in healthcare

Artificial intelligence is growing quickly in every industry, and healthcare is no exception: nearly half of all startups that have raised funding are using AI in some capacity, and VCs are changing their strategy because of it. 

At the same time, there's a debate over how to regulate it, with California Governor Newsom recently vetoing a bill that would have created AI safeguards and regulations. 

"AI is obviously a very powerful technology. It's a new technology, so there's a lot of uncertainty in certain pockets, hesitancy around it. So, the government being involved and being thoughtful isn't a bad thing: for the innovators, for the technology leaders, for the standard setters, to be aligned on regulation earlier in AI's journey is a good thing. The fact that we're talking about this now, as opposed to many years from now, when it's almost too late, is a good thing," said Iqbal.

"AI is a great technology, it's going to have many, many, many applications. We're in the first wave of how you can apply it. It's going to have a much more incremental impact in every part of our lives. Everyone realizes that it's going to be a part of our lives. But, today, the power of AI is really, hey, are there real business problems, real human problems, real challenges, that AI can be applied to in a thoughtful, secure way that really moves the needle?"

In healthcare, the technology has already been used to help transform revenue cycle management, meaning insurance claims: when people have to type in codes, the error rates of those codes became a real business problem that had staff impacts, patient impacts, and downstream insurance company impacts. It's also helped in workflow automation, where Luma plays, and there's a lot of power of applying AI there.

The next place where AI has a big impact on health, Iqbal predicts, will be around clinical applications.

"How do we make our providers' lives better? How do we ensure that there's not false positives, false negatives, hallucinations, all these things? How do we equip our providers to have the AI copilot? The ambient listeners, that's good technology, it's already deployed. That's workflow automation, in my mind, that's not really super clinical, because before you used to have note transcribers," he said.

"Now you're doing that in a better, thoughtful way. But when we can actually start to move the needle in terms of diagnosis, therapy recommendations, context aware therapy recommendations that help a provider make better decisions? That's going to be a lot of power," he said.

In that landscape, Luma will continue to its current mission of driving workflow automation through the use of artificial intelligence; Navigator is an example of that, as it improves the workflow for patients and for staff, while Fax Transform really improves the workflow for staff, while making it so much easier for patients to schedule and get to the specialty care they need.

Going forward, Luma will also do more for other stakeholders in the healthcare ecosystem, such as the marketing department within health systems, automate their workflow, making them more thoughtful.

"Can we automate more and more call center workflows? Because the phone is still the primary front door for healthcare, and so Navigator can be applied to refills, scheduling, rescheduling, and questions. Can we go deeper and deeper and deeper to free up time for staff to not spend 90 seconds with 1,000 people, but to spend 20 minutes with the elderly patient who really needs to be walked through all the different things? That provides meaning to that individual staff member but then, of course, it’s really impactful for that patient and their business. So, we'll lean into the cost center workflows," he explained.

Another future area for Luma will also be the patient's financial experience, especially as healthcare becomes more expensive. The company will be able to help patients know what the cost of care is before they ever step into the office. 

"There's so much opportunity to automate. That's where we're looking to deliver, and we're close to it being a true shop, click, pay experience for healthcare consumers. That's what ultimately we all want is for patients to say, 'it's easy, I know what it costs, I'm going to get it. I'm not waiting months. I'm able to get in within days.' When that happens, they can think about how to stick to their care journey and being a good patient, as opposed to all the work it means just to get our foot in the door. So that's really where we continue to lean in here at Luma."