CGM FAQ 2 – The in-depth look

Where is CGM regulation today?

Three main pathways currently shape the global CGM landscape:

1. FDA Class III medical device approval (USA)

• The highest regulatory bar for medical devices.
• Demonstrates safety and effectiveness.
• Transparent: approval documentation is public in the FDA database.
• Devices passing Class III are highly trusted for insulin dosing.

2. iCGM (USA, since 2018)

• The only CGM-specific quality standard ever developed.
• Introduced three study design “special controls”: pragmatic, loosely defined, but achievable for the market leaders of 2018 (Dexcom G6).
• Built on Class III approvals, it codified what safe, effective CGMs could achieve and accelerated approval via the 510(k) pathway.
• Crucially, iCGM enabled interoperability — allowing CGMs to link with third-party apps, bolus calculators, and automated insulin delivery (AID) systems.
• In 2018, this was transformative: it sped up approvals, drove safe innovation, and fuelled global adoption.

3. CE Marking (Europe)

• CE is market authorisation, not a quality standard.
• Approval can be based on equivalence to existing devices, a representative sample of data “on file”, often without published peer-reviewed data.
• Critically opaque: the evidence remains between the manufacturer and the notified body.
• Contrast with the USA: FDA Class III and iCGM approvals are fully transparent, and data can be reviewed by anyone via the online database.
• CE approval does not mean poor quality:
  • Abbott, Dexcom, and Medtronic devices hold CE marking.
  • Roche also launched its newest CGM Accu-Chek SmartGuide with strong accuracy data based on a robust study design (48 subjects and 15,000 data points, compared with the market leaders with 80–200 subjects and 10,000–30,000 data points) – Mader et al. (2024).
  • The CareSens i-SENS also launched with a strong study design and accurate data (30 subjects, 7,000 data points) – Jendrike et al. (2025).
  • But you only know by reading publications; CE marking alone is not enough.
  • The DSN Forum chart, which gives study design a score out of 5 (where a score of at least 4, ideally 5, is currently the best available tool to assess the study design of CE-marked products).

4. Clinician-led consensus: eCGM, Latin American, and BSI

• eCGM and Latin American = clinician consensus recommendations aiming to bring in iCGM special controls.
• BSI = draft UK guidance to codify iCGM.
• Both essentially try to globalise iCGM principles.
  • I support the ideology behind these initiatives — they are well-intentioned attempts to close important gaps.
  • But the execution is missing: they are not recognised standards, they carry no legal mandate in any jurisdiction, and as splinter efforts, they risk fragmentation, monopolisation of the market, and slowing innovation by locking into outdated 2018 pragmatic rules, not 2025 science.
  • Notified bodies (around 50 in Europe) can only enforce an international harmonised standard.

5. Other markets (e.g. TGA in Australia)

• Often follow CE marking, extending its global influence.

Bottom line

• iCGM and FDA Class III are quality standards and were excellent in 2018; they are still quality standards, but they are outdated by 2025 science.
• CE marking is not a quality standard; it provides market access but lacks consistency, robust evaluation, and most importantly, transparency.
• eCGM and BSI are recommendations.
• The only binding path is a harmonised international standard.

👉 See: Pemberton et al., Diabetes Obes Metab, 2023
👉 Explore our CGM Education Series
👉 Listen: Podcast with Guido Freckmann on CGM Regulation

---

Is the 2018 iCGM and FDA Class III outdated for CGM regulation?

iCGM raised the bar in 2018, and Class III FDA approval is the highest bar for CGM regulation.

Unfortunately, the bar is uneven — and at certain points, limited study designs can slip under the lowest part. In other words, the bar is no longer high enough.

• 2018 approach: loosely defined study design, but practical and based on what market leaders could achieve.
• 2025 science: validated procedures now expose discrepancies that iCGM cannot resolve.
• Evidence: in the 2025 Diabetes Care head-to-head study, the same patients wore three market-leading sensors — two iCGM-approved CGMs and one FDA Class III–approved CGM. Results showed an 8% difference in time-in-range and a 13% difference in time-in-tight-range. A clinically meaningful difference is just 5%.

Remember, AID 5 is driven by CGM readings lower than venous glucose, so you need a 5–10% higher TIR to get the same HbA1c.

AID System – See the 2025 analysis	End-of-trial TIR % (95% CI)	End-of-trial HbA1c % (95% CI)
AID 1	72.4 (57.8–87.0)	7.1 (6.4–7.7)
AID 2	71.8 (69.4–74.2)	7.0 (6.9–7.1)
AID 3	67.9 (66.2–69.6)	7.3 (7.1–7.5)*
AID 4	69.0 (67.5–70.5)	7.1 (7.0–7.2)*
AID 5	74.4 (69.7–79.1)	7.1 (6.8–7.4)
AID 6	63.1 (59.4–66.8)	7.2 (7.1–7.3)

This means:

• iCGM was essential for accelerating innovation and interoperability.
• But it now risks holding us back if not updated in line with validated procedures.

👉 Read: Diabetes Care, 2025 – three-sensor study
👉 See commentaries: Pemberton et al., Diabetes Care, 2025; Beck, Diabetes Care, 2025; Waldenmaier et al., Diabetes Care, 2025

---

Why does the comparator matter (device bias, venous vs capillary samples)?

Important note: Since the 2023 Guidelines and Recommendations for Laboratory Analysis in the Diagnosis and Management of Diabetes Mellitus, it has been clear that glucose measurements must exhibit a bias of <2.1% to be considered effective as reference methods.

Recent data show that even “lab-grade” analysers can exceed the accepted glucose testing bias threshold of <2.1%. Only with retrospective correction can handheld blood glucose meters (that you and I use) or ‘lab-grade’ analysers consistently meet the <2.1% bias threshold. Independent thought leaders have recommended a shift to using capillary as the comparator (reference) for these reasons:

• Capillary glucose reflects lived reality: people with diabetes have always checked CGM accuracy — and historically corrected insulin doses — using capillary blood glucose. Every manufacturer’s Instructions for Use (IFU) clearly state that if there is doubt about CGM accuracy, confirm with a capillary blood glucose meter.
• Capturing peaks: capillary is at least 10–30% higher than venous glucose postprandial spikes, and after-meal highs, rather than fasting levels, correlate more strongly with HbA1c and, therefore, future complications, in a systematic review of 14 quality studies.
• Bias: data clearly demonstrate that with retrospective calibration to higher-order reference materials, capillary SMBG comparators can achieve <2.1% bias.

However, the key point here is standardisation. If the decision is to use venous glucose as the reference, then CGM systems must demonstrate a positive bias consistent with the datasets that established the associations with complications. If the decision is to use capillary glucose as the reference, then CGM systems must instead demonstrate a slight negative bias aligned with those same datasets.

Either approach can work — what matters is that all systems are standardised to the same reference and maintain a bias of less than 2.1%.

---

What are eCGM/Latin American/BSI CGM recommendations (and why is it not enough)?

The eCGM is a clinician consensus document being promoted in Europe, and there is a counterpart Latin American version. There is also the British Standards Institution (BSI) standard for CGM. The issues:

• It has no regulatory force — it is guidance only.
• Both are built on 2018 iCGM science.
• Not harmonised international standards.
• They carry no legal mandate.
• They risk creating fragmentation if adopted in parallel to CE marking.

---

Are we at a pivotal point?

Yes. We now face two choices:

Option A: stay with CE marking + iCGM and introduce fragmentation with consensus efforts

• Pros: convenient, easy, fits current inertia, will result in short-term stability and consistency in quality.
• Cons:
  • Monopolisation of the market by two to three manufacturers
    • Ruling out robustly assessed CGM systems, such as the Accu-Chek SmartGuide and one other as described above.
  • Fragmentation between regulatory systems.
• Tools like the DSN Forum chart and the FIND initiative backed by the IDF can help clinicians assess risk. But they cannot:
  • Standardise the comparator.
  • Correct for comparator bias.
  • Deliver cross-device comparability.
• Bottom line: we will end up with patchwork tools, not a true standard.

Option B: build an international harmonised standard

• Backed by IFCC 2025 methodology:

• Capillary comparator aligned to patient experience, or venous comparator with a requirement for a positive bias for the CGM device.
• Bias correction (<2.1%).
• Validated procedures (meal and insulin challenges on the same visit).
• Dynamic Glucose Region Plot (DGR) to map sensor accuracy across the zones where real-world decisions happen:
  • Low and going lower → hypoglycaemia treatment.
  • Low but recovering → alarm reliability.
  • High and going higher → correction doses.
  • High and sustained → prolonged hyperglycaemia detection.

• Challenges:
  • Manufacturers may face pressure for dual approvals unless the CDRH of the FDA comes to the table.
  • Multicentre sites must standardise methodology, which requires certification of study centres.
  • Cost — who will fund this?
    • Manufacturers, philanthropic foundations, independent research grants, crowd-based or pooled contributions.

Bottom line: CGM needs an international standard to protect its future.

---

What will it take for global regulators — especially the FDA — to act?

• Regulators listen when patient communities speak with one voice.
• Organisations (charities, professional societies, advocacy groups) amplify that voice.
• Without visible demand, regulators assume the status quo is “good enough”.