Episode 37 — Dexcom G7 and ONE+: Inside the CGM Behind the AID Revolution

Why this episode exists

Dexcom is one of only two CGM systems in the UK with iCGM approval — the highest international accuracy bar. But accuracy alone doesn’t explain why it’s become the device of choice for most AID users and the majority of paediatric centres. This episode exists to explain the full picture: the features that matter in daily life, the two products designed for different patient groups, and what Dexcom’s roadmap actually looks like beyond the Dexcom G7.

It also addresses something that doesn’t get discussed clearly enough: the risk of treating CGM as a commodity. For anyone making insulin decisions, accuracy is not interchangeable — and the evidence base behind a sensor matters as much as the sensor itself.

In this episode

Adam Dawes joined John having previously hosted John on Dexcom’s own podcast — so this time the tables are turned. Adam brings a clinical perspective that most medical affairs roles can’t offer: he trained and worked as a paediatric diabetes nurse specialist before joining Dexcom, and he’s been in post long enough to have watched the Dexcom G7 launch, navigate the early issues, and see the product land in a very different place four years on.

The conversation moves through the Dexcom G7 and ONE+ in practical terms — what they share, where they diverge, and which patients belong on which — before opening into the wider Dexcom roadmap, the role of CGM in type 2 diabetes, the FDA manufacturing story, and a candid discussion about why the CGM market will never be a true commodity market for insulin users.

What this episode explores

• The Dexcom G7 and ONE+: same hardware platform, different software — and why that distinction matters clinically
• Urgent low soon and Delay 1st Alert: how smarter alert design reduces alarm fatigue without reducing safety
• AID integrations: Tandem T:slim X2, Omnipod 5, and what’s coming with mylife and Tandem Mobi
• Why 30-minute warmup and 12-hour grace periods have transformed sensor management in school and paediatric settings
• Dexcom ONE+: designed for primary care, type 2 diabetes, and MDI users without hypoglycaemia risk
• CGM for type 2 diabetes — the evidence base, the biofeedback argument, and why immediate feedback changes behaviour in ways three-month HbA1c reviews cannot
• Stelo: over-the-counter CGM in the US, what it signals for the future, and the role of generative AI in personalised insights
• The 15-day Dexcom G7 sensor (US, adults): what changed in the algorithm, why paediatric data comes first, and the interoperability challenge
• G8 next generation: filament redesign, multiple analytes, and 25 years of learning built into a new platform
• Why insulin dosing is not a commodities market — and what that means when newer, cheaper CGMs enter the UK
• The FDA manufacturing findings: what happened, how Dexcom responded, and the Ireland factory built from first principles
• EPIC integration and population health: how Clarity data will flow directly into electronic health records

Watch or listen

Episode chapters

• 00:00 — Introduction: the CGM Series and why Dexcom is next
• 00:53 — Adam Dawes: from paediatric DSN to Dexcom UK Medical Affairs
• 03:18 — Why John’s unit runs 280 of 300 patients on Dexcom
• 05:52 — Dexcom G7 features: urgent low soon and the predictive alert evidence base
• 08:28 — Delay 1st Alert: the most underutilised alert in CGM
• 14:51 — AID integrations: Tandem T:slim X2, Omnipod 5, and what’s coming
• 17:28 — Form factor: 30-minute warmup, 12-hour grace period, and how families use them
• 20:31 — Dexcom ONE+: same hardware, different patient profile
• 22:07 — CGM for type 2 diabetes: the behaviour change argument
• 26:32 — Stelo, over-the-counter CGM, and personalised biofeedback
• 29:22 — Clarity moving in-app: data ownership simplified
• 30:42 — ATTD roadmap: 15-day sensor, G8, EPIC integration, and generative AI
• 43:45 — Why insulin dosing is not a commodity market
• 47:15 — FDA manufacturing: what happened, how Dexcom responded, and the Ireland factory
• 51:52 — Closing: measuring organisations by how they respond when things go wrong

Key themes

1. Same hardware, different software — and why that distinction matters

The Dexcom G7 and ONE+ are built on identical hardware: the same sensor filament, the same transmitter, the same all-in-one form factor. What separates them is software — specifically the alert suite and the patient profile each system is designed for. The Dexcom G7 retains urgent low soon and Delay 1st Alert, and carries the full AID integration stack. The ONE+ removes urgent low soon (appropriate for patients without significant hypoglycaemia risk) and simplifies the experience for primary care, type 2 insulin users, and MDI patients who don’t need the full alarm architecture. Understanding this distinction helps clinicians match the right system to the right patient rather than defaulting to one device for everyone.

2. Delay 1st Alert: alarm design that actually reduces harm

Alarm fatigue is one of the least-discussed harms in CGM use — but it’s real. When high alerts fire every time a patient eats, patients switch them off or set thresholds so high they no longer provide useful signal. Delay 1st Alert solves this by only alerting once glucose has been above a set threshold for a defined period (typically two hours), by which point the alert is actionable: it’s time to check ketones, take a correction, consider changing a set, or get moving. For paediatric settings this is particularly significant: it removes the alarm that tells a child they have diabetes during a school lesson without giving them anything to do about it. The alert, when it fires, is a signal to act.

3. AID integration: the Dexcom G7 as a connected platform, not just a sensor

The Dexcom G7 currently integrates with Tandem T:slim X2 and Omnipod 5 in the UK, with mylife integration announced for piloting in other markets. Tandem Mobi is on the horizon. This makes the Dexcom G7 the widest AID integration stack available in the UK right now. But integration isn’t just about hardware compatibility — it creates interoperability obligations. Every firmware update to any element of the stack requires re-validation across all connected systems, which is why new sensors, new algorithms, and new pumps take time to reach patients even once the underlying technology is ready. The 15-day sensor, for example, will need to be validated separately for each connected AID system before it can be used in those configurations.

4. CGM in type 2 diabetes: the biofeedback argument

The evidence base for CGM in type 2 diabetes has shifted from a trickle to overwhelming in five to ten years. The mechanism is less about alarm thresholds and more about immediate feedback. For a person with type 2 managing through lifestyle, diet, and possibly basal insulin, a CGM provides something no three-monthly HbA1c can: a direct, visible connection between a specific choice and a glucose outcome — within hours. John describes this as gamification with a biological foundation: when reducing evening carbohydrate intake produces a visible improvement in overnight glucose, motivation is sustained in a way that abstract future outcomes cannot achieve. For clinicians, this changes the educational toolkit.

5. 25 years of data: why longevity in the CGM market compounds

Newer CGM companies entering the market face a challenge that goes beyond manufacturing: algorithm training. A CGM algorithm is only as good as the data it has learned from, and that data accretes over years and iterations. Dexcom’s shift from enzyme and hardware development cycles to algorithm-led improvement — explicitly discussed in relation to the 15-day Dexcom G7 in the US — reflects 25 years of learning about where the failure modes are, what the edge cases look like, and how to iterate without introducing new risk. For patients and clinicians evaluating newer entrants, this is a legitimate question to ask: how many sensor-years of real-world data has trained this algorithm?

6. The 15-day sensor and the G8: what’s actually changing

The 15-day Dexcom G7 in the US is not simply the same sensor with a longer lifespan — the extension has come alongside a significant algorithm update that Dexcom is limited in discussing publicly due to regulatory constraints. What Adam confirms: it’s algorithm-driven, not hardware-driven, meaning the filament and transmission hardware are the same. The G8, however, represents a full platform generation: a redesigned filament, the possibility of additional analytes (ketones have been discussed; John advocates for lactate as the more meaningful metric for pre-diabetes and exercise), and a smaller on-body footprint. Both the Dexcom G7 15-day and the G8 are coming to Europe, but adults-first in the US means paediatric studies are likely underway concurrently.

7. EPIC integration: from retrospective reviews to proactive population health

Dexcom’s EPIC integration means Clarity data — AGPs, time in range, time below range — can flow directly into an NHS electronic health record for sites that have made the switch. For diabetes centres, the clinical value is significant: it removes the need to toggle between Clarity, LibreView, Glooko, and Tandem Source in separate browser tabs during a clinic consultation, and it creates a dataset that matches CGM metrics to HbA1c records in one place. For researchers, this enables studies comparing GMI to HbA1c, examining sensor-specific reporting patterns, and identifying the patients between clinic visits who most need proactive intervention — rather than waiting for the next scheduled appointment.

8. Why insulin dosing is not a commodities market

As more CGM manufacturers enter the global market and attempt to compete on price, there is a structural pressure towards commoditisation — the argument that CGM readings are interchangeable and that the cheaper option is equivalent. John’s counterargument is precise: for any patient making insulin decisions based on CGM data, the 40/40 threshold (the maximum acceptable error band for iCGM-approved sensors) defines the floor, not a ceiling. A reading that sits outside that band on 1% of occasions will, over a sensor lifetime, produce incorrect insulin doses. The risk isn’t theoretical. This analysis changes for non-insulin CGM users (behaviour change, lifestyle monitoring, pre-diabetes) where the consequences of an occasional inaccurate reading are different — but for type 1 diabetes and insulin-using type 2, device quality is a safety question, not a cost question.

9. Manufacturing transparency: the FDA as a system of accountability

The FDA’s finding that some Dexcom manufacturing processes didn’t match design specifications is described by Adam with an unusual degree of directness: it was an opportunity, not just a crisis. The new Ireland factory — Dexcom’s first in Europe — is being built from those lessons ground up. More broadly, John makes a point worth noting: FDA post-market surveillance and manufacturing audit are not replicated at the same level under CE marking. For patients and clinicians evaluating any CGM, the question of whether a device operates under FDA oversight is not just a geographic technicality — it’s an indicator of the accountability structure around that product’s quality control.

10. The Dexcom G7 at four years: a different product from launch

Adam is candid that the UK and Ireland were the first markets to launch the Dexcom G7 — and that, in retrospect, being first came with risk. Early performance issues led John’s paediatric centre to pause migration from G6 and restart it in January 2026. The experience since has been different. For patients who tried the Dexcom G7 early and moved back to the G6, the message from this episode is specific: the sensor you experienced then may not be the sensor available now. Algorithm and software updates throughout a device’s lifespan mean that the Dexcom G7 at launch and the Dexcom G7 at four years are, in meaningful ways, different products.

Practical exploration checklist

For people with type 1 diabetes (and their families):

• If you’re using the Dexcom G7, explore the Delay 1st Alert setting — it can remove an alarm that fires after meals when there’s nothing actionable to do
• If you tried the Dexcom G7 early and went back to the G6, it may be worth revisiting — the sensor has been through several algorithm updates since launch
• If you have 10 or more followers set up on your CGM, consider whether all of them are necessary — John’s clinical view is that surveillance beyond five is usually unhelpful for the person wearing the sensor
• Explore when your 12-hour grace period activates and how your family uses it — particularly for school mornings when changing a sensor at 7am and waiting two hours is not realistic
• Understand that all sensors, including the most accurate, will occasionally produce a reading outside the 40/40 error band — always have a finger prick meter available

For clinicians and educators:

• When migrating patients from G6 to Dexcom G7, consider Delay 1st Alert as part of the setup conversation — not as an optional extra
• Consider which patients in your cohort are appropriately placed on a Dexcom G7 versus ONE+ — the two products are designed for different risk profiles, not different accuracy needs
• If your centre has moved to EPIC, explore the Dexcom Clarity EPIC integration — it removes the multi-tab clinic workflow and creates a matched CGM/HbA1c dataset for audit and research
• For patients with type 2 diabetes on insulin where behaviour change is the primary goal, the biofeedback evidence base is now substantial — CGM provides immediate feedback that retrospective clinic reviews cannot replicate
• When discussing CGM choice with patients, be explicit about what iCGM approval means, what the 40/40 threshold represents, and why insulin dosing decisions require that level of accuracy standard

About the guest

Adam Dawes is Senior Medical Affairs Manager for Dexcom UK and Ireland. Before joining industry, Adam trained and worked as a paediatric diabetes nurse specialist, working in London and then at Cambridge’s Addenbrooke’s Hospital. He joined Dexcom over four years ago and leads medical education and evidence dissemination across UK and Ireland, including the expansion of CGM access into primary care and type 2 diabetes. His clinical background gives him an unusual perspective for a medical affairs role: he has prescribed and supported the devices he now works on, and he has worked with the patient populations — children, families, school teams — whose daily experience of CGM the product decisions ultimately affect.

Related reading on GNL

• CGM Series Ep 35 — Prof Othmar Moser on CGM accuracy and study design
• CGM Series Ep 36 — DSN Forum: CGM accuracy in clinical reality
• The GNL CGM Series — all episodes
• GNL AID Algorithm Optimiser — explore how AID algorithms behave
• 10, 20, 30 Minutes Walking to Lower Highs — glucose responses to different activity types
• Hypo & Hyper Treatment — understand the thresholds and what drives them
• All GNL Explorers