Dear readers,

It’s great to have you join me once again as we browse through my Digital Health Notes together.

The other night, at half past midnight, I found myself sitting in front of my laptop with a sore throat and a mild fever, facing an existential question: Should I google my symptoms? With the risk that after twelve minutes I’d be convinced I was either about to die or at least patient zero of a new pandemic.

According to a recent report by OpenAI, around 230 million people ask ChatGPT health-related questions every week. And now the company is planning a dedicated product: ChatGPT Health.

What’s happening here is bigger than just a more convenient symptom checker. AI-driven systems are making medical knowledge available around the clock – accessible, language-adapted, without waiting rooms and without appointments.

For people with limited access to healthcare, this is revolutionary. For overstretched health systems, it may even offer relief. But for all of us, it is also a double-edged sword.

Because one thing is clear: ChatGPT Health does not replace a doctor. It simulates one.

The system does not understand disease. It recognizes patterns in text. It does not know what lungs sound like. It cannot smell ketones on the breath. It cannot see a rash. It only sees words.

And in medicine, words are notoriously unreliable.

“I feel dizzy” could mean dehydration.
Or a stroke.

Between empowerment and overconfidence

Supporters argue that informed patients are better patients. Those who assess their symptoms in advance enter conversations with doctors more structured. Those who understand side effects better understand their therapies. And those who have medical terminology translated into everyday language feel taken seriously.

All of that is true.

But AI also creates a false sense of certainty. The answers are calm, logical and linguistically confident. That is precisely what makes them so convincing – and potentially so dangerous.

We tend to confuse well-formulated responses with factual accuracy.

If a machine says, “This sounds more like a harmless viral infection,” that can be reassuring. Perhaps too reassuring.

Medicine thrives on uncertainty, probabilities and clinical experience. AI thrives on data patterns. These are not the same thing.

The underestimated risk: our most intimate data

Health is one of the most sensitive categories of data we have.

Anyone feeding ChatGPT Health with symptoms, diagnoses, medication plans or mental health concerns is sharing more than just information – they are sharing vulnerability.

OpenAI promises encrypted storage and separation from other chats. That sounds reassuring. But even the highest security standards cannot eliminate real risks: data breaches, cyberattacks and technical failures are part of digital reality.

And unlike a forgotten password, health data cannot simply be reset.

A mental health diagnosis or genetic risk does not disappear just because we click “log out.”

The crucial question therefore is not only: How good is the medical advice?

It is also: Where does my data end up – and what happens if it does not stay there?

The fact that GPT Health has not yet been announced for Europe is probably no coincidence. The GDPR and European medical device regulations are not exactly friendly playgrounds for rapid product launches.

Who carries the responsibility?

If a doctor makes a mistake, there are liability rules, clinical guidelines and professional regulations.

If ChatGPT Health makes a mistake, there is an update.

Responsibility is politely handed back to the user: “Please consult a healthcare professional in case of serious symptoms.” Legally sound. Socially unsatisfying.

Because realistically, people will not only use AI as a source of information – they will use it to help make decisions.

Especially when they are uncertain.

The real potential

And yet the potential is enormous.

AI can translate medical information into understandable language, structure therapy plans, remind patients to take medication in chronic conditions, answer questions people may hesitate to ask their doctor, or help interpret a second opinion.

Used correctly, ChatGPT Health could become a navigation system through a complex healthcare system.

Not a replacement for physicians – but an intelligent guide.

The key question is therefore not: Should AI be allowed to answer medical questions?

The real question is: How do we integrate it responsibly into existing healthcare structures?

Conclusion: between digital family doctor and dangerous illusion

ChatGPT in healthcare is neither a miracle cure nor a threat. It is a tool.

The real problem is not the AI itself. The problem is our tendency to mistake eloquence for competence.

Perhaps the doctors of the future will not compete with AI, but interpret it.

Perhaps medical expertise will increasingly include knowing when to trust the machine – and when not to.

And until that day arrives, it might not be such a bad idea that when I feel sick at half past midnight, I am not sitting in front of my laptop but lying in bed, recovering the old-fashioned way.

In that spirit: stay healthy. And stay critically curious.

Until next time,
Torsten Christann

Who’s shaping the future of diabetes technology?

For the fifth time, Digital Oxygen presents its comprehensive overview of the most relevant players in advanced diabetes technology.

The Advanced Diabetes Technology Landscape 2026 maps more than 250 companies across the rapidly evolving diabetes ecosystem – from Continuous Glucose Monitoring (CGM) and smart insulin delivery systems to AI-driven software platforms and non-invasive glucose measurement technologies.

This year’s edition visualizes the landscape in the form of the chemical structure of glucose – the molecule at the very center of diabetes care. The structure highlights how different technologies and companies are increasingly connected within one complex innovation ecosystem.

Key observations from this year’s landscape include:

• Automated Insulin Delivery (AID) has become the new standard in tech-supported diabetes therapy.
• Non-invasive glucose monitoring remains a hotbed of innovation, with numerous companies pursuing different technological approaches.
• CGM continues to be the fastest-growing segment, with many new entrants entering a market once dominated by only a few players.
• Diabetes software platforms are expanding rapidly on both the patient and healthcare provider side.

As the ecosystem grows more complex, patients benefit from smarter tools and better therapy support. At the same time, healthcare professionals face increasing demands for training, guidance, and digital literacy.

As Managing Partner Torsten Christann explains:

“The combination of affordable CGM systems and widely available AID software may finally unlock the full potential of smart insulin pens – bringing next-generation diabetes care to millions of people with type 2 diabetes.”

The 2026 Landscape also groups the Non-Invasive Measurement segment into four major technology approaches currently under development:

• Optical
• Electro-magnetic
• Thermal-mechanical
• Bio-chemical

Which of these technologies will prevail remains one of the most exciting questions in diabetes innovation.

The Advanced Diabetes Technology Landscape 2026, is structured into 11 key categories, including:

• Insulin Pumps (traditional, patch and tubeless – with and without AID)
• Smart Pens, Caps and Clips
• Continuous Glucose Monitoring (CGM)
• Implantable CGM
• Non-Invasive Measurement
• AID Software
• Point-of-Care Software
• Diary & Titration Apps
• Nutrition & Lifestyle Apps
• Patient and Therapy Support
• DiGA (Digital Health Applications)

Download the full landscape and explore the companies shaping the future of diabetes technology.

Dear Readers,

There is something peculiar about the first days of a new year. The calendar reliably fills up again with appointments, meetings, and deadlines—and yet, there is sometimes that brief moment where one genuinely believes things can change. New routines. Better habits. Perhaps even a more conscious approach to one’s own health. Between New Year’s resolutions and the first mandatory appointments, I ask myself every year: What am I truly taking with me into the new year—and what is better left where it belongs: in the last one?

It was with exactly this mixture of curiosity and healthy skepticism that I looked at the digital health innovations of CES 2026. One wants to remain open to new things—while simultaneously wondering which of them will still be relevant by Easter and which will have quietly vanished from one’s app overview long before then.

The good news first: In 2026, Digital Health feels significantly more “grown-up” than it did just a few years ago. Wearables, diagnostic solutions, and health platforms have become quieter. And that is precisely their greatest progress. Apple, Samsung, and Google largely moved away from grand health “showstoppers,” focusing instead on incremental improvements: more precise sleep analysis, more stable stress tracking, and better integration into existing platforms. Unspectacular? Yes. But probably exactly what health needs. It is no longer being revolutionarily reinvented—it is being quietly and continuously improved.

From Data Fetishism to Meaningful Trends

As always, a look at the specialized providers was exciting. Oura showcased a new generation of rings with a stronger focus on metabolic health and long-term trends. Garmin pushed cardiovascular monitoring and load management closer to clinical-grade questions. The market is visibly moving away from a daily “data fetish” toward longer timeframes and more meaningful benchmarks and analyses.

This represents a massive opportunity, especially for prevention. At the same time, the risk of “silent overwhelm” is growing. The more we measure continuously, the greater the need for explanation. Data without context is of little help, no matter how elegantly it is visualized.

Another focus of CES 2026 was once again Home Diagnostics. Providers like Vivoo presented advanced tests that combine multiple biomarkers and enrich their results—naturally—with AI-supported recommendations. Added to this were new blood analysis concepts for home use, sitting somewhere between lifestyle optimization and medical ambition. The line is thinning—and with it, the responsibility of the manufacturers is growing.

Particularly interesting are solutions like the evolved health mirror “Omnia” by Withings. Here, health literally gains a physical place in the room—no longer just one app among many, but a daily point of reference in a location that is an integral part of one’s daily routine. Strategically clever. At the same time, the uncomfortable question remains: Do such integrated health hubs really reach the general population—or primarily a well-meaning, wealthy, tech-savvy audience?

The Reality Check for AI

And then, of course, there is Artificial Intelligence.

Hardly a product or platform at CES 2026 could do without it. AI analyzes, predicts, prioritizes, reassures—and sometimes, it simply seems to be marketing. Whether it’s a wearable, a home test, or a health dashboard: AI is inside everything, often so naturally that one almost forgets to ask what it actually makes better.

The oft-cited AI bubble was certainly hard to miss. While some providers are seriously working on explainable, medically resilient models—for example, based on platforms like NVIDIA Clara—others content themselves with high-sounding promises and little transparency. In 2026, AI is a requirement, not a differentiator. It’s becoming harder to truly impress anyone with it.

Perhaps that is the actual signal of this CES. The industry seems to sense that the “AI magic” won’t last forever. The closer digital solutions move to medical decision-making, the less a “built with AI” sticker suffices. Explainability, liability, and responsibility are moving to the forefront—entirely without the buzzwords.

Conclusion: Less “Smart,” More Meaningful

CES 2026 shows a Digital Health world that has arrived and must now prove its worth. The opportunities clearly lie in networked prevention, continuous support, and new care models. The risks are equally clear: overwhelm, lack of transparency, and an AI euphoria that promises more than it can deliver.

Perhaps that fits quite well with January. Fewer resolutions, more realism. Less “smart,” more meaningful.

My wish for the new year: That in the digital health sector, we don’t just want to measure more and more, but that we want to—and will—understand better. That technology is not an end in itself but creates real added value in daily (healthcare) life. And that those solutions prevail which are allowed to stay because they work—not because they were particularly loud at CES.

And if, while reading this, you feel the urge to critically review your own health apps and consider which ones really deserve to stay—then you feel just like I do.

Until next time—and with best wishes for a healthy New Year 2026.

Yours,

Torsten Christann

Dear readers of PM-Online,

It’s a pleasure to have you peruse my Digital Health Notes once again, where I delve into current developments in the expansive field of digital health.

Speaking of perspectives: when two individuals look at the same object—one sees “black,” the other “white”—it often warrants a second glance. This was my experience while reading the two recent DiGA reports: one from the GKV-Spitzenverband, the other from the SVDGV (German Association for Digital Health). Both examine the same market, both rely on the same data foundation—yet they seem to present entirely different pictures.

Two Truths, Many Numbers – A System in Motion

Fact is: the DiGA landscape has evolved remarkably in a short time. From 24 DiGAs at the end of 2021 to 68 by the end of 2024, with 59 actively listed. That’s nearly a 200% growth over three years—in a healthcare system typically characterized more by persistence than agility.

Over one million prescriptions have been issued, with approximately 81% activated by patients, according to the GKV report—a sign that DiGAs are not only prescribed but also utilized. In the last reporting period alone, about 420,000 activations occurred—a 100% increase compared to the previous year.

Financially, the sector is also gaining momentum: the annual market volume now stands at approximately €110 million, a 72% increase over the previous year. Since the introduction of the DiGA system, total expenditures by statutory health insurance have reached about €234 million—a significant volume for a care sector that has existed for only four years.

Content-wise, clear focal points emerge: the majority of applications address mental health conditions, followed by metabolic and musculoskeletal disorders. Notably, a single DiGA (Oviva Direkt) accounts for nearly 28% of all code activations—a strong indicator of the potential for targeted applications in major indication areas.

User demographics also paint a clear picture: the average age is 47, with the highest usage among those aged 55 to 60. Women comprise 73% of users—men, younger, and older individuals are currently underrepresented.

Approximately 20% of prescriptions are repeat prescriptions—suggesting that many DiGAs are not yet used long-term or repeatedly. Whether this is due to efficacy, user-friendliness, or structural barriers remains unclear—but it indicates that integration into standard care is still ongoing.

The GKV’s Perspective: Expensive, Weakly Substantiated, Not Yet Integrated

The GKV-Spitzenverband’s report reads more critically. The tone: many DiGAs are costly—but deliver little. Only 18% of applications demonstrated a positive healthcare effect upon inclusion in the directory. The vast majority—56 applications—were initially included on a trial basis.

And the success? According to the GKV, mixed. As of the reporting date, 21 DiGAs were still under evaluation. Of the remaining 35, nine were fully removed, and seven partially removed by the BfArM due to insufficient evidence of benefit. This means: only about half of trial DiGAs fully deliver on their promises.

Economic concerns also arise: manufacturer prices averaged €580 in 2024, with the highest exceeding €2,000. Negotiated prices, however, average €226 per quarter. The GKV-Spitzenverband refers to this difference as “startup financing without return.” This totals around €96 million—€41 million in the first year, where the manufacturer sets the price freely and the GKV covers it fully, and another €55 million in the second year, as many DiGAs remain in trial longer. While there is a general obligation to repay—insolvencies, as already filed by seven DiGA manufacturers, negate this guarantee. Thus, the GKV provides significant liquidity over two years—with high economic risk and no repayment assurance.

The report also highlights deficits in evidence: many studies are methodologically weak, using retrospective designs, lacking control groups and blinded evaluations, and often relying on self-reports from users—subjective assessments of symptoms, mood, or behavior. Objective measures are frequently absent, significantly limiting the validity. Particularly critical are DiGAs that were removed without demonstrating benefit: nearly €4 million was spent in the first year alone—without repayment obligation and without proven healthcare effect. From the GKV’s perspective, this contradicts the principle of evidence-based medicine—and unnecessarily burdens the solidarity system.

The SVDGV’s Perspective: Young, Growing, Learning

The SVDGV report offers a contrasting viewpoint. Where the GKV-Spitzenverband sees risks, the SVDGV sees a system in development—with learning curves but also clear progress. The high number of trial inclusions is interpreted as a feature of a learning system. Over two-thirds of trial DiGAs have achieved permanent listing—now with study designs comparable to those required for drug approval. This demonstrates: the process works—when innovation is given space.

The SVDGV also positively notes the increasing establishment of DiGAs as a “third care sector” alongside inpatient and outpatient care. Nearly one in five DiGAs is re-prescribed—a sign of therapeutic trust and sustainable benefit.

Significant regulatory advancements, such as the integration of DiGAs into electronic patient records (ePA), a mandatory success verification (AbEM) from 2026, and the inclusion of risk class IIb products, indicate: DiGAs are no longer mere novelties but integral parts of the healthcare system. However, the association also calls for: less bureaucracy, more digital processes—and especially an end to the “analog activation code,” which still unnecessarily complicates access.

My Conclusion – Somewhere Between App Store Romanticism and Evidence Obligation

DiGAs are neither the sole future of medicine nor its downfall. They represent an attempt to digitize care—especially considering that we must treat more patients with fewer professionals. This is their structural value: not as a replacement for human interaction, but as a means to alleviate processes, ease access, and allocate resources more effectively.

What currently concerns me: the overly simplistic categorization into “works” or “doesn’t work.” It’s not that straightforward.

What we need instead is a nuanced perspective:

• Yes, stronger evidence is necessary.
• Yes, not every app belongs in the directory.
• But also: Yes, innovation requires space—and time to demonstrate impact.

And me? I remain curious.

I’ve tried several DiGAs—out of genuine interest. Not out of sheer tech enthusiasm, but from an earnest desire to understand how care can be different. Better, more accessible, more digital—where it fits.

What I hope for? That we judge innovation not by how flawlessly it starts, but by its willingness to learn. DiGAs are no substitute for doctors, no miracle cure. But they are a step—perhaps not perfect, but necessary.

And when my smartwatch vibrates again soon, I won’t just think of my step counter. I’ll remember that movement sometimes starts in the mind. Not everything is black or white. But it’s often in between where things get interesting.

Yours,
Torsten Christann

Who’s shaping the future of diabetes care?

For the fourth time, Digital Oxygen presents its comprehensive map of the most relevant players in advanced diabetes care. The 2025 edition of the Advanced Diabetes Technology Landscape provides a structured overview of more than 250 companies, covering the entire spectrum of innovative diabetes technologies – from CGM systems and smart pens to AID software and Non-Invasive Measurement.

This year’s edition highlights a rapidly evolving market:

• Automated Insulin Delivery (AID) has become the new standard in tech-supported diabetes therapy.
• Non-invasive glucose monitoring continues to be a hotbed of innovation – while products are still pending, R&D is booming.
• CGM remains the fastest-growing segment, with more than 20 companies entering a space previously dominated by just a few key players.
• Diabetes Software is becoming even more crowded on both ends, HCP and patient focused.

With the growing complexity of the technology landscape, patients benefit from smarter tools – but healthcare providers face new challenges. The need for training, guidance, and tech literacy is rising. This opens opportunities for tech companies to support HCPs more directly and strengthen their patient relationships.

More competition also means lower costs – and better access to advanced diabetes tech for broader patient populations. As Managing Partner Torsten Christann puts it:

“The combination of affordable CGM systems and widely available AID software may finally unlock the full potential of SMART Pens – bringing next-gen diabetes care to millions of type-2 patients.”

This year, the Non-Invasive Measurement space is grouped into the 4 most dominating technologies used to measure glucise levels: optical, electro-magnetic. thermal-mechanical, bio-chemical. Which of these technologies will prevail is not yet clear.

Explore the full Landscape 2025, structured into 11 major categories:

• Insulin Pumps (Traditional, Patch, Tubeless – with/without AID)
• Smart Pens, Caps, and Clips
• Continuous Glucose Monitoring (CGM)
• Implantable CGM
• Non-Invasive Measurement
• AID Software
• Point-of-Care Software
• Diary & Titration Apps
• Nutrition & Lifestyle Apps
• Patient and Therapy Support
• DiGA (Digital Health Applications)