Artificial intelligence (AI) and similar technologies are already having a significant effect on healthcare for international patients and will go further, the World of America Convention has heard.
Hospitals and insurers are using hundreds of these systems to save clinician time, improve diagnosis and treatment and also better read notes from other countries and languages.
It also has the potential to allow insurers to better monitor patient outcomes from facilities.
Speaking at the World of America Convention, Northwestern Medicine operations vice president Susan Helfrich said AI was “very transformative” and explained how the hospital group was looking to increase efficiencies in hospital systems.
“We face a shortage of physicians and all clinical workers in the future,” she said.
“We’re getting a lot more patients coming in, people are living longer and we need to deal with how we’re going to accommodate and handle all these patients coming in for healthcare.
“What we’re trying to do is decrease the administrative burden on physicians more than anything – we really want to focus on helping accommodate people coming into Northwestern for healthcare.”
Saving 11 hours a month
Helfrich (pictured) highlighted some key practical examples of how using AI had improved care and been more efficient.
“One example is when patients are in an exam room there’s some ambient listening, and that dictates a chart for a physician,” she continued.
“While it’s pretty accurate, there are a few tweaks that need to be made for the final chart before it goes into the record, but that’s really helped significantly.
“It’s saved primary care physicians about nine to 11 hours a month and with that additional time, we can accommodate more patients coming through our healthcare system.”
There is also AI being in radiology, specific to x-ray and CT scans.
“That’s made some significant efficiencies where we have about 40% savings on x-rays and about 20% on CT scans, so that helps with the workforce and improving that efficiency and throughput of patients.”
She emphasised this was you’re not replacing people but giving them technology to be able to work more efficiently and faster – to enhance their effort and overall use.
300 AI applications
Mayo Clinic senior director for contracting and payer relations Jon Vineyard agreed that AI had myriad possibilities for the healthcare sector and was already being used in hundreds of ways.
“In 2025 we had well over 300 different applications of AI, both clinically and administratively, and I see that accelerating to well over a thousand,” he said.
“There will be much more of a pantry of AI, whether it be incorporation of large language models (LLMs), generative AI applications, or accelerated validation of data.
“This is going to be one of the things that will help to be the catalyst for accelerated clinical innovation and research that will help to bring to clinical practice new innovation that will benefit patients more and more.
“I also see there will be some significant innovation in adaptability between providers across a number of different countries; that it will be much more seamless than it has been in the past.
“I hope there will be more of that adoption, but certainly I think that we’re going to see a parabolic rise in the application of AI.”
Removing the language burden
For international private medical insurance (IPMI) customers who are likely to have treatment in multiple countries and potentially multiple languages, application of AI here could be game-changing and improve their treatment pathways.
It could also relieve a notable burden for themselves and help them to understand their treatment and medical records better.
“One of the challenges has been that receiving medical records in a different language can create unnecessary time spent translating that and can be clunky,” Vineyard continued.
“There’s not a whole lot of reliability on getting the right translator to do as quickly as possible utilising some of these new lenses that translation can be almost instantaneous.
“That’s minimising the administrative burden on the patient too and having that taken away from the entire ecosystem of international healthcare is crucially important.
“Then when a patient does receive care at any institution within the United States having the ability to see that medical record for themselves and sharing it back with the home physician translated instantaneously is a remarkable and a big innovation that will ultimately benefit patients,” he added.
However, Vineyard highlighted that ultimately, enhancing patient trust was a critical baseline to all these developments.
“At the core of this, AI needs to be something that facilitates and promotes trust between the provider and the patient,” he said.
“If AI applications are not equitable, if they are not available to all, if they are not something that promotes that trust, you’ve missed the mark.
“One of the things that we need to ensure as providers is that it’s very patient-centric – which could mean very different things for very different populations; people in different generations may see the utilisation of AI very differently from those of younger generations.
“Some are more than happy to interact with a bot or something that’s a little bit more personalised, but others want to hear the human voice on the other end, so we need to make sure that however we begin to implement and incorporate AI into clinical practice, that it is patient-appropriate and that it’s patient-centred,” he added.
‘Consulting every encyclopaedia related to that condition’
Part of this patient-centred approach is ensuring the diagnosis and treatment are as accurate and tailored as possible.
It can also mean making sure that doctors are as well-informed as possible about their patients and the conditions they are diagnosing and treating.
Here, Vumi Group vice president and chief medical officer Dr Juan Carlos Velazquez argued AI would be able to play a major role.
“From a medical standpoint, artificial intelligence definitely makes responses regarding diagnosis, severity of the condition, and regarding prognosis and treatment much more accurate,” he said.
“Why? Because in normal life, what happens is one gets treated by a doctor, and that doctor has read 15 books or 25, and has five or 25 years of experience.
“However, when we submit that same case through a process where artificial intelligence acts as support for that doctor we are simply consulting every encyclopaedia related to that condition and that treatment.
“Furthermore, since in medical practice not all of us have the same level of experience regarding a specific condition, it has the ability to evaluate how that condition is going to behave in that person, at that age, with that sex, with that lifestyle, in that country, and with that diet.
“So, definitely, from a medical point of view, we are much closer to a more appropriate answer, and treatments can be adjusted in a much more individualised way.”
‘Fundamental tool’ for evaluating hospitals
Importantly, Dr Velazquez highlighted that AI systems could help insurers to monitor the quality of care and patient outcomes being delivered by hospitals.
With quality outcomes data difficult to receive from facilities in some locations, these tools could become highly valuable in assessing which healthcare providers to use.
“I believe that artificial intelligence will also help us, above all, to evaluate hospitals based on outcomes and not on marketing or individual experiences – and that is where the industry should be moving,” he continued.
“When we talk about outcomes, we are talking about costs, about customer experience, about what the diagnosis and the evolution of the illness have been, the severity of the illness, the number of complications; this is part of the hospital quality evaluation that we insurers have to do.
“Our definition of networks should be based more on outcomes than on geographic location, so meaning a fundamental tool for evaluating this is going to be artificial intelligence,” he concluded.
