Below is a recap of the key ideas from the workshop.
When Irving Stackpole, President of Stackpole & Associates, took the virtual stage for this HTDC/MedTourConnect educational workshop, he opened with fair warning: this would be different from the presentations that came before it. What followed was a candid, expert-level walkthrough of how digital technology and artificial intelligence are reshaping cross-border medical trade — not as a distant trend, but as a shift already underway.
The moderator framed the stakes clearly in her introduction: as healthcare organizations look to expand across borders, the industry needs exactly this kind of strategic, forward-looking thinking. Stackpole delivered it, drawing on decades of work inside the vertical markets that make up what most people simply call “medical tourism.”
Medical Tourism Isn’t One Industry — It’s Many
The workshop’s first reframe was foundational: medical tourism isn’t a sector. It’s a collection of highly fragmented, highly regionalized vertical markets. Treating it as one undifferentiated industry is, in Stackpole’s view, a strategic error. Organizations that specialize — that tailor their service offering and demand management to a specific acuity level and region — are the ones positioned to succeed. Generalist positioning, he suggested, tends to dilute rather than expand opportunity.
Cross-Border Care Has Gone Multimodal
For years, “medical tourism” has meant one thing: a patient crosses a border, receives care, and returns home. Stackpole pointed to the World Trade Organization’s General Agreement on Trade in Services (GATS) framework to explain why that picture is now incomplete. Cross-border trade in services happens across four distinct modes, and all four are now active in healthcare:
- Mode 1 — Cross-border information exchange. The oldest example is teleradiology; today it includes teleconsults, second opinions, remote monitoring, and even tele-ICU and robotic surgery support.
- Mode 2 — Consumption abroad. The classic model: the patient travels to the destination and returns home.
- Modes 3 and 4 — Commercial presence and movement of people. Clinicians traveling to teach or treat, hospital networks operating across borders, digital brokers, and AI-enabled navigation teams.
The throughline: digital services cross borders far more easily than patients do. As Stackpole put it, it’s easier to get data across a border than it is to get a body across one — patients are constrained by visas, transportation infrastructure, and geopolitics in ways that information simply isn’t.
Acuity Changes Everything
A recurring theme was the sharp divide between low-acuity and high-acuity care — and why conflating the two leads to bad strategy.
Low-acuity, elective procedures (cosmetic dentistry, hair transplants, aesthetic surgery) are the easiest services to export digitally. They’re voluntary, scalable, and well-suited to marketing-driven, self-service consumer journeys — a website and strong social presence can do much of the work of attracting and converting patients. Margins here come from repetition: doing high volumes of the same procedure, well.
High-acuity care — a child’s cerebral aneurysm, a solid mass tumor, complex cardiology — is a different market entirely. It requires physical infrastructure, multidisciplinary teams, and far more governance. Volume is inherently lower, and the value of digital tools shifts from consumer acquisition to triage, coordination, and managing outcomes.
The strategic implication: the digital and marketing model built for hair transplants cannot be the model used for hip replacements, and vice versa.
The Real Gap: What Happens After Care Ends
One of the workshop’s more pointed observations was about the industry’s collective blind spot: post-discharge follow-up. Providers and facilitators are, in Stackpole’s experience, very good at attracting patients and guiding them through a procedure — and much less consistent about what happens once treatment is complete and the patient goes home.
He cited OECD data showing average teleconsultations per patient rose from 0.5 in 2019 to a pandemic-driven peak of 1.3, before settling back to around 1.0 — with continued resistance among older physicians in developed markets, an advantage he noted for younger physician populations in some Islamic-majority countries. Availability of digital health services, meanwhile, climbed to 82% in 2024, up from 79% the prior year.
The takeaway: patients increasingly expect to reach a doctor by video or WhatsApp, and to get real answers about recovery — and closing this follow-up gap is, in Stackpole’s assessment, the clearest opportunity for building lasting patient and provider relationships.
Where AI Actually Fits Today
Stackpole was direct about cutting through hype: most commentary on AI is, in his words, opportunistic scaremongering. The more useful distinction is between what AI can do now and what it may do later.
Today, most healthcare AI is large language models — tools like Claude, Copilot, and Perplexity — trained on publicly available information. Current, practical applications include:
- High-quality translation, now dramatically improved
- Patient matching (blood, tissue, and similar data-driven comparisons)
- Pathway navigation and remote monitoring
- Administrative support and case packaging for payers, ministries, and international patient centers
He also flagged an important caution: much of the underlying data driving these tools skews toward Western populations, and organizations need to be scrupulous about that bias when applying AI-driven insights to diverse patient populations.
On the horizon — agentic AI, where systems make decisions rather than simply surfacing information — is not yet here in any meaningful clinical sense, but Stackpole pointed to several likely next steps: intelligent referral routing, automated case packaging, digital twins for treatment modeling, and predictive post-discharge monitoring.
Guardrails Aren’t Optional
Perhaps the most important throughline of the session was accountability. As AI takes on more of the administrative and even clinical support burden, human accountability for high-risk decisions has to remain central — not just as a principle, but as an operational requirement: credentialing, certification, and oversight of everyone involved in arranging cross-border care.
Stackpole was candid that errors happen even in well-run systems — by his estimate, 10%-20% of more complex procedures result in some untoward outcome — which makes oversight, not just optimism about technology, the real safeguard.
He also pointed to interoperable health data as a major unresolved challenge. Health systems frequently don’t talk to one another — not just across borders, but within the same country, between public and private systems, or between hospitals and clinics. Harmonizing this data, along with procurement and regulatory processes, is a substantial undertaking already attracting serious attention — and one that will reshape how embassies, payers, and third-party administrators screen and select partners going forward.
Three Strategic Messages
Stackpole closed with a clear set of takeaways for organizations operating in this space:
- Telemedicine is the enzyme, not a side product. It’s the catalytic element that runs through every vertical of cross-border trade, not a bolt-on feature.
- Build separate models by acuity. The approach that works for elective, low-acuity services will not work for complex, high-acuity care — and shouldn’t be forced to.
- Invest in AI where the channel demands it. The value AI adds depends entirely on where and how an organization operates — there’s no one-size-fits-all AI strategy for cross-border healthcare.
——————————
Have a question for Irving that wasn’t covered in the session? He welcomes follow-up — reach out directly at istackpole@stackpoleassociates.com.