Continuous Glucose Monitoring (CGM)

Scope, method, and source rules

This deep-dive replaces the legacy CGM note’s coverage gap flags: no_deep_research_21d, source_types_lt_3, stale_research, and tech_no_buyer. It adds named buyer/customer classes from trial inclusion criteria and FDA-cleared over-the-counter use, and it applies the highest available source tiers. No arXiv preprint is used; the clinical references below are peer-reviewed journal articles. Any figure that could not be independently verified against a primary source is labeled needs primary source validation.

Source tiers used here:

• Tier 1 for technical/performance/IP claims: granted patents, peer-reviewed papers, standards-body documents, regulator technical/press documents.
• Tier 2 for adoption/deployment context: company product pages and company press releases, marked self-reported.
• SEC regulatory filing tier for risk-factor cross-reference: SEC EDGAR company filing indexes, used only for adoption/risk disclosure and not for technical performance claims.

All URLs were accessed 2026-04-26 unless otherwise stated. A granted patent is evidence of recorded claims, not proof of performance, current commercial viability, or current patent exclusivity.

Technology architecture and core measurement stack

1.1 Electrochemical amperometric sensing

Modern CGM devices share a core architecture: a subcutaneous sensor measures interstitial glucose, a transmitter or integrated electronics sends raw electrochemical signals to a receiver/app, and an algorithm converts those signals into glucose estimates and trend arrows.

The sensor chemistry is electrochemical and amperometric. The oldest continuously relevant granted patent used in this note is US 6,809,653 B1, titled Telemetered characteristic monitor system and method of using the same, granted 2004-10-26 and assigned to Medtronic MiniMed, retrieved from Google Patents:

Source: US6809653B1 — Telemetered characteristic monitor system and method of using the same, granted 2004-10-26, accessed 2026-04-26.

The patent describes a monitor that uses a characteristic sensor—for example, a glucose oxidase-based electrochemical sensor—and telemetered data transmission to a receiver. This is an early granted patent in the CGM lineage, but it is not evidence that any current product uses this exact claim set, nor is it evidence of current patent exclusivity.

Abbott’s FreeStyle Libre 3 and Dexcom’s G7 product materials describe the consumer-facing result of this architecture, but the company pages are Tier 2 self-reported product documentation, not independent technical verification.

1.2 Transmitter, algorithms, and wear logistics

The wear period and replacement cadence are important because CGM economics are driven by recurring sensor replacement.

• Dexcom G7: advertised 10-day sensor wear plus a 12-hour grace period, 30-minute warm-up, and smartphone connectivity. This is self-reported product data from Dexcom’s product page.
• Abbott FreeStyle Libre 3: 14-day sensor wear, no separate transmitter for the Libre family, and optional real-time alarms. This is self-reported product data from Abbott’s product page.

Source: Dexcom G7 product page, self-reported, accessed 2026-04-26.

Source: Abbott FreeStyle Libre 3 product page, self-reported, accessed 2026-04-26.

The wear durations are product specifications, not independent performance validation. They are useful for the recurring-revenue calculation below.

Patent/IP deep-dive

2.1 Granted patent used here

• Patent number: US 6,809,653 B1
• Title: Telemetered characteristic monitor system and method of using the same
• Grant date: 2004-10-26
• Assignee: Medtronic MiniMed, Inc.
• Document URL: [patents.google.com
• Access date: 2026-04-26

This is a granted patent, not an application. It supports the claim that telemetered characteristic monitoring, including glucose monitoring, was claimed in US patent records by 2004. It does not support any claim that a specific current Abbott, Dexcom, Medtronic, or Senseonics CGM is protected by this patent, because patent claims, current products, and validity all require separate analysis.

2.2 What a granted patent does and does not prove

A granted patent is evidence of recorded claims; it is not proof of performance, safety, accuracy, commercial viability, or market exclusivity. In this note, the patent is therefore used for the architectural/IP claim that “a telemetered characteristic monitor for glucose was claimed in a granted US patent by 2004,” not for any clinical or commercial performance claim.

This distinction matters for moat analysis. A 2004 Medtronic MiniMed patent may be expired or substantially narrowed by later prosecution and litigation; without a current claim chart and maintenance-fee status, it should not be treated as a current moat.

Clinical performance evidence

3.1 DIAMOND and GOLD randomized trials

Two peer-reviewed randomized clinical trials are used here to establish the clinical benefit of CGM in adults with type 1 diabetes treated with multiple daily insulin injections. Both are Tier 1 peer-reviewed journal articles.

DIAMOND trial

Citation: Beck RW, et al. Effect of Continuous Glucose Monitoring on Glycemic Control in Adults With Type 1 Diabetes Using Insulin Injections: The DIAMOND Randomized Clinical Trial. JAMA. 2017;317(4):371-378.

DOI: 10.1001/jama.2016.19975

Published: 2017-01-24

Accessed: 2026-04-26

The DIAMOND trial reported a larger A1C reduction in the CGM arm than in the control arm. In the abstract-level reporting used here, the CGM arm had a mean A1C change of approximately −1.0 percentage point, while the control arm changed by approximately −0.4 percentage points, producing a between-group difference of −0.6 percentage points with a 95% CI of approximately −0.9 to −0.3. This is the primary quantitative input for the worked calculation below.

GOLD trial

Citation: Lind M, et al. Continuous Glucose Monitoring vs Conventional Therapy for Glycemic Control in Adults With Type 1 Diabetes Treated With Multiple Daily Insulin Injections: The GOLD Randomized Clinical Trial. JAMA. 2017;317(4):379-387.

DOI: 10.1001/jama.2016.19976

Published: 2017-01-24

Accessed: 2026-04-26

GOLD reported a between-group A1C difference of approximately −0.4 percentage points (95% CI approximately −0.57 to −0.23). The GOLD and DIAMOND results are from different trial groups and institutions, so the direction and approximate magnitude of A1C improvement in type 1 diabetes are cross-verified. The exact numerical values are presented as reported in the abstracts; the reader should query the full texts for patient-level granularity.

3.2 Worked calculation: A1C treatment effect and IFCC conversion

This calculation is quantitative, formula-based, and source-bound. It is not a market-size or margin calculation; it demonstrates the clinical effect size used throughout this note.

Formula

`

ΔA1C_between = ΔA1C_CGM − ΔA1C_control

`

Inputs

Calculation

`

ΔA1C_between = (−1.0) − (−0.4) = −0.6 percentage points

`

Sensitivity

The reported 95% CI for the between-group difference is approximately −0.9 to −0.3 percentage points. If the true difference were as small as −0.3 percentage points, the clinical effect would still be directionally favorable but materially smaller; if it were −0.9 percentage points, the effect would be larger.

IFCC conversion sensitivity

Using the NGSP–IFCC master equation for A1C conversion:

`

IFCC (mmol/mol) ≈ (NGSP% − 2.15) × 10.929

`

A 0.6-percentage-point NGSP difference is approximately:

`

0.6 × 10.929 ≈ 6.6 mmol/mol

`

Sensitivity range using the 95% CI:

• Lower bound: 0.3 × 10.929 ≈ 3.3 mmol/mol
• Upper bound: 0.9 × 10.929 ≈ 9.8 mmol/mol

Source for conversion factor: NGSP IFCC/NGSP conversion page, accessed 2026-04-26. NGSP is a harmonization authority, not a Tier 1 standards body, so this conversion factor is classified as Tier 3 authoritative reference in the parameter table.

3.3 Sensor accuracy (MARD) and self-reported claims

Dexcom’s G7 marketing materials state a mean absolute relative difference (MARD) of approximately 8.2% in adults, and a 30-minute warm-up, but these are self-reported, single-source performance claims. MARD is a derived accuracy metric:

`

MARD (%) = (1/N) × Σ |CGM − YSI| / YSI × 100

`

where YSI is the reference laboratory glucose measurement.

The MARD figure is not independently re-derived in this note. It should be treated as needs primary source validation until independently confirmed against a Dexcom SEC filing, FDA summary document, or peer-reviewed accuracy study. It is not used as a load-bearing claim in this note.

Source: Dexcom G7 product page, self-reported, accessed 2026-04-26.

Standards context: ISO 15197:2013 defines accuracy requirements for self-monitoring blood-glucose systems, but it is not a direct MARD standard for CGM. The standard is cited here to show that regulatory accuracy terminology is standards-bound, not a marketing construct.

Source: ISO 15197:2013 — In vitro diagnostic test systems — Requirements for blood-glucose monitoring systems for self-testing in managing diabetes mellitus, accessed 2026-04-26.

Company filings cross-reference and marketing discrepancies

SEC filings are used here only for adoption/deployment and risk-factor context; they are not used for technical performance claims.

4.1 DexCom: marketing “most accurate” vs. filed risk factors

Dexcom markets the G7 as an accurate, low-warm-up CGM. The product page uses accuracy claims in marketing language.

Source: Dexcom G7 product page, self-reported, accessed 2026-04-26.

DexCom’s Form 10-K risk factors, accessed through the SEC EDGAR filing index, disclose that product accuracy and reliability are material risks. The filing language is conditional: accuracy can vary with manufacturing tolerances, sensor placement, interfering substances, and physiologic state. The contrast between absolute marketing language and conditional risk-factor language is the relevant discrepancy.

Source: SEC EDGAR company filing index, DexCom, Inc., CIK 0001093557, Form 10-K, accessed 2026-04-26.

4.2 Abbott: Libre product claims vs. filed risk factors

Abbott markets FreeStyle Libre 3 as a small, 14-day sensor with optional real-time alarms.

Source: Abbott FreeStyle Libre 3 product page, self-reported, accessed 2026-04-26.

Abbott’s Form 10-K risk factors, accessed through the SEC EDGAR filing index, disclose that the diabetes care franchise faces intense competition, manufacturing/supply-chain concentration, and product acceptance risk. The marketing page communicates a product advantage; the filing communicates that advantages can erode. This is a framing discrepancy, not a factual contradiction.

Source: SEC EDGAR company filing index, Abbott Laboratories, CIK 0000001800, Form 10-K, accessed 2026-04-26.

4.3 Cross-reference result

No claim in this note relies on company marketing as proof of technical performance. Where marketing and filings conflict, the filing is treated as the more conservative evidence. The MARD figure is therefore single-source and self-reported, not a cross-verified technical claim.

Parameter table

Unit economics / business model

CGM monetization is recurring hardware/consumable revenue, not a one-time device sale. The value chain is:

1. Sensor manufacturers supply disposable wearable sensors. They monetize through sensor replacements every 10–14 days. DexCom, Abbott, Medtronic, and Senseonics all sit here.
2. Transmitter/receiver or integrated electronics are sold separately or integrated into the sensor. In Dexcom’s architecture, the transmitter historically is a separate durable component; in Abbott’s Libre architecture, the sensor contains the electronics.
3. Software and algorithm layer monetizes through device/app functionality and increasingly through subscription or connected services. This layer is increasingly critical but is not separately priced in the sources retrieved here.
4. DME and pharmacy channel captures distribution margin. In the U.S., CGM has shifted from durable medical equipment (DME) to pharmacy distribution for both Abbott and Dexcom, which affects channel economics but exact channel margins are not independently verified here.
5. Payers reimburse CGM for eligible patients, which creates the demand floor.

Margins: CGM consumables are widely described as high-margin recurring revenue, but specific gross-margin figures are not quoted in this note because they require direct verification against the exact current-year 10-K or earnings release. Any commonly cited margin figure should be treated as needs primary source validation.

Recurring-volume calculation:

`

annual_sensors = 365 days / wear_days_per_sensor

`

For Dexcom G7 at 10 days:

`

365 / 10 = 36.5 sensors per person-year

`

For Abbott FreeStyle Libre 3 at 14 days:

`

365 / 14 ≈ 26.1 sensors per person-year

`

This calculation uses the product-page wear durations cited above; the resulting volumes are the basis for recurring-revenue scale, not specific revenue forecasts.

Who captures value: The sensor supplier captures the largest fraction of the recurring consumable stream because the sensor is the rarefying/disposable component. Integrated pump manufacturers such as Insulet and Tandem capture system value by incorporating CGM into automated insulin delivery, but the CGM sensor revenue accrues primarily to the CGM supplier.

Disconfirming Evidence

The implicit thesis of this note is that CGM is a clinically validated, recurring-revenue sensing technology with durable adoption across insulin-treated diabetes and emerging over-the-counter wellness markets. The thesis would be weakened or invalidated by evidence such as:

1. A randomized controlled trial in insulin-treated type 2 diabetes or non-insulin-treated type 2 diabetes showing no A1C improvement or no time-in-range benefit, because that would undermine the mass-market clinical expansion case.
2. A post-market safety signal or manufacturer field correction involving systematic sensor inaccuracy, calibration drift, or interference from common medications, because the recurring-revenue model depends on safety and accuracy expectations.
3. A CMS or private-payer coverage contraction that removes or narrows reimbursement for CGM, because the purchased-through-reimbursement buyer class would shrink.
4. Successful patent invalidation or non-infringement findings that allow commoditized, lower-priced sensors to enter, because that would compress the recurring sensor gross margin that the model currently assumes.
5. A large-scale OTC consumer study showing the wellness/non-insulin buyer class does not retain after initial purchase, because the investor narrative relies on broadened buyer class volume beyond insulin-treated diabetes.