Skip to content
type1.science

Dexcom G7

Dexcom

What it is

An all-in-one, factory-calibrated CGM with 10-day wear and a 12-hour grace period. US labeling covers ages 2+; named pump integrations and water-resistant sensor wear offer practical flexibility, with country, software and meter-backup requirements.

Editorial review: .

Most recent recorded citation date: 2026-06-24. Only explicit date metadata is included; an undated citation may be newer. This does not mean every claim was reviewed on that date.

Trial status, labels and access can change between reviews. How we review the evidence · How to read the evidence

Full sources ↓Full discussion ↓Report an issue →

Available nowRegulator-approvedcgm

Evidence behind this assessment

Key evidence notes. Study results, product eligibility and access answer different questions.

Who was studied?
Study populations and analysis groups vary. Product age limits alone do not describe who was studied.
Benefit or performance
Accuracy: Arm MARD was 8.2% in the Dexcom-funded adult accuracy study; this is a study average, not an individual-reading guarantee or a validated conversion to this editorial score.[1]
Important harms and treatment burden
Read the safety discussion and original sources. A missing summary does not establish safety.
Approval and country access
Country-specific approval and access are not summarized in this record.Approval, trial recruitment, local supply and funding are separate. Check the cited label or access source.
Follow-up and remaining uncertainty
Read the full discussion and original sources for follow-up duration and study limitations.

This record cites company or conference reports. Those sources may describe interim findings; their source type is shown in the source list.

Editorial score: calculation and evidence

A weighted editorial judgment on a 0–100 scale, not a probability of success or a measured treatment effect. Higher criterion scores mean more favorable assessments.

Score limitations. Scores are editorial judgments, not values validated by the cited studies. Standard G7 and G7 15 Day were studied separately; small differences in their accuracy, lag or calibration scores do not establish a measured advantage. Commercial and DIY compatibility are version- and country-specific.

Default calculation: 86 × 20 + 65 × 15 + 92 × 15 + 72 × 10 + 95 × 10 + 82 × 10 + 85 × 10 + 0 × 5 + 62 × 5 = 7725; divide by total weight 100. Unrounded weighted result: 77.25.

Accuracy86

Arm MARD was 8.2% in the Dexcom-funded adult accuracy study; this is a study average, not an individual-reading guarantee or a validated conversion to this editorial score.[1]

Low lag65

The adult study estimated mean system lag of 3.5 minutes by time-shifting readings to minimize MARD. It does not support a fixed 10–15-minute delay for every person.[1]

Interoperability92

Current US manufacturer listings include Tandem t:slim X2, Tandem Mobi, Omnipod 5 and iLet. Exact software, sensor packaging and country availability matter.[7]

Sensor lifespan72

US instructions specify 10 days of wear plus a 12-hour grace period; a sensor can still end early.[5]

No calibration95

No routine calibration fingersticks. Use a meter for treatment decisions if readings do not match symptoms or expectations, during warm-up, or if the glucose number or trend arrow is missing.[5]

Alerts & prediction82

Configurable alerts include Urgent Low Soon, predicting glucose reaching 55 mg/dL within 20 minutes. Supported display devices and alert settings matter.[5]

Form factor85

All-in-one sensor with warm-up under 30 minutes. The sensor-only water rating is 2.4 metres for 24 hours; it does not apply to the receiver or phone.[6]

Ketone sensing0

Measures glucose, not ketones.[5]

Access & cost62

Current US pharmacy and coverage pathways are documented, including conditional Medicare eligibility. Out-of-pocket cost depends on coverage and individual eligibility.[12]

The full picture

The Dexcom G7 combines a glucose sensor and transmitter in one wearable. The US label covers people with diabetes aged 2 and older, with upper-arm placement and an upper-buttock option at ages 2–6. Insertion sites and connected-device features can differ by country. FDA cleared the original G7 on 7 December 2022.54

Accuracy: what the studies actually measured

The Dexcom-funded adult study analyzed 316 people with diabetes, 619 sensors and 77,774 in-range sensor–laboratory pairs. MARD—the average absolute percentage difference from the reference measurement—was 8.2% on the arm and 9.1% on the abdomen. The study deliberately varied glucose, used blinded/postprocessed sensor signals and excluded sensors that failed before the first clinic session. Arm MARD was 11.9% on day one, so the overall average does not describe every day or reading. This was an accuracy study, not a trial demonstrating an HbA1c or time-in-range benefit.1

In the separate Dexcom-funded pediatric study, 127 analyzed participants aged 7–17 contributed 15,437 laboratory-reference pairs: arm MARD was 8.1%, abdomen 9.0%. Among 28 analyzed children aged 2–6, the pooled result was 9.3% across only 343 capillary meter-reference pairs from several insertion sites. The younger result is not an arm-only estimate and is not directly interchangeable with the older children’s laboratory comparison.2

A concurrent comparison of commercial standard G7 and Libre 3 sensors found MARD 13.6% versus 8.9% in 55 participants with laboratory-reference data. Abbott funded, designed, analyzed and interpreted that study. Unlike the pivotal study, it did not deliberately manipulate glucose. It used 3,640 G7 and 4,020 Libre 3 laboratory pairs; only 35 per device were below 70 mg/dL and one per device below 54 mg/dL. Its accuracy result is relevant, but the sparse low-glucose data do not establish superior hypoglycemia safety, and it did not test G7 15 Day.3

Wear, water and everyday use

US instructions specify 10 days of wear plus a 12-hour grace period, with warm-up under 30 minutes. Factory calibration removes routine calibration fingersticks. A blood glucose meter is still needed for treatment decisions if readings do not match symptoms or expectations, during warm-up, or if the glucose number or trend arrow is missing.5

The sensor is water-resistant to 2.4 metres for 24 hours. This rating does not apply to the receiver, phone or watch. Direct to Watch can display sensor readings without the phone nearby, but a compatible phone is still needed to pair each new sensor; supported hardware, software and regional feature availability matter.65

Customizable high/low alerts include Urgent Low Soon, which predicts glucose reaching 55 mg/dL within 20 minutes. A prediction is not a guarantee that every low will be detected. Follow the current instructions for the display device and alert settings in use.5

The adult study estimated mean system lag of 3.5 minutes, calculated by shifting sensor and reference readings in time to minimize MARD. This is a study estimate, not a fixed delay for every person or a claim that glucose changes always reach the sensor in 3.5 minutes. It does not establish the lag of G7 15 Day.1

Pump, phone and open-source compatibility

Current US manufacturer listings include Tandem t:slim X2, Tandem Mobi, Omnipod 5 and iLet. Check the specific pump, software and compatible sensor packaging in your country; a US partnership listing does not establish every regional configuration.7

Original project documentation lists standard G7 support in Loop, Trio and AndroidAPS. AndroidAPS currently warns that older BYODA and new DiaKEM installations have update/authentication limitations; its documentation describes alternatives. These are project-specific compatibility statements, not a promise that every open-source loop or installation works, and not regulatory clearance of DIY use.8910

The G7 measures glucose, not ketones. Its readings do not replace a ketone measurement.5 Current US manufacturer guidance permits use during pregnancy for type 1, type 2 or gestational diabetes; that permission is distinct from proof of pregnancy-specific clinical benefit.11

Access and transition from G6

Commercial access is documented in the US, UK, Germany, Canada, Australia and Japan. The EU marker here includes verified German access, not a claim about every EU country. US costs depend on insurance and eligibility; Medicare coverage is conditional. UK self-funded purchase, Canadian pharmacy/private or provincial pathways, and Australia’s NDSS subsidy from 1 March 2025 each have their own conditions.121314151617

Dexcom’s official transition information gives 1 July 2026 as the G6 manufacturing deadline, while remaining stock and app support can continue. The global transition is phased. It does not mean that all G6 supplies vanished on that date or that G7 is the company’s only prescription CGM platform everywhere.242526

Safety notices to check

App correction. FDA lists open Class II corrections Z-2163-2026 and Z-2164-2026 for G7 iOS and watchOS versions 2.9.0–2.11.2, initiated on 14 April and posted on 11 May 2026. The defect can replay older glucose values and trend arrows and delay alerts. Affected users are directed to update to the latest applicable Dexcom app. Both standard G7 and G7 15 Day are included; these notices concern the apps.1819

Receiver speaker recall. A separate FDA notice concerns certain G7 receivers whose speakers may fail to sound low/high alerts. Following Dexcom’s 9 June 2025 update, users should check their receiver at Dexcom’s serial-number checker, request a replacement if affected, and test the speaker whenever charging. A successful sound test does not rule out later failure. The notice lists affected models; it does not cover every G7 receiver.20

Stolen sensor lots. Dexcom’s announcement dated 26 May 2026, published by FDA on 29 May, identified scrapped G7 lots 1725204004 and 1725069002 that were stolen and resold. The first lot was not properly sterilised; the second had an elevated internal test-failure risk and may provide no readings. Follow the notice’s lot-check and replacement instructions. FDA’s Class II recall Z-2574-2026 was initiated on 27 May and posted on 24 June. These are lot-specific findings, not an assurance that other devices cannot fail.2122

Japanese Android notice. A separate Dexcom Japan notice dated September 2026 concerns G7 Android versions 2.9–2.11: under specified Focus Mode and app-setting conditions, vibration could be suppressed while sound and visual alerts still occurred. It recommends the latest Google Play update. This regional manufacturer notice should not be confused with the iOS/watchOS FDA correction above.23

Longer wear and future products

G7 15 Day is a separate version: its US label is 18+, with a 60-minute warm-up and 15 days plus a 12-hour grace period. Those specifications should not be substituted for standard G7 instructions.27

Dexcom’s May 2026 investor presentation describes a future G8 size-reduction target of 50% and expected late-2027/early-2028 timing. These are company plans, not released specifications. The presentation does not establish ketone sensing for G8.28

Sources

  1. [1]
    Accuracy and Safety of Dexcom G7 Continuous Glucose Monitoring in Adults with Diabetes · Peer-reviewed study · 2022-02-21 — Dexcom-funded study. Methods, Results and Tables 2–5: 316 analyzed adults, 619 sensors and 77,774 in-range pairs; site-specific accuracy and estimated lag. This is analytical accuracy, not a trial of clinical benefit.
  2. [2]
    Accuracy of a Seventh-Generation Continuous Glucose Monitoring System in Children and Adolescents With Type 1 Diabetes · Peer-reviewed study · 2022-04-25 — Dexcom-funded study. Methods and Tables 2–5 distinguish 127 analyzed participants aged 7–17 using YSI from 28 aged 2–6 using capillary reference measurements. Online date; journal issue July 2023.
  3. [3]
    Comparison of Point Accuracy Between Two Widely Used Continuous Glucose Monitoring Systems · Peer-reviewed study · 2024-01-08 — Abbott funded, designed, analyzed and interpreted the study. 55 participants contributed YSI data; commercial standard G7 and Libre 3, not G7 15 Day. Only 35 low-YSI pairs per device and one below 54 mg/dL limit hypoglycemia comparisons.
  4. [4]
    FDA clearance letter K213919 — Dexcom G7 · Regulatory decision · 2022-12-07 — Decision date. The manufacturer announcement followed on 8 December; neither date establishes an FDA endorsement of superiority over other CGMs.
  5. [5]
    Dexcom G7 US user guide · Manufacturer — Current US instructions: indications, warnings, treatment decisions, alerts, Direct to Watch and technical information. Component water limits and compatible-device requirements apply; country labeling can differ.
  6. [6]
    Dexcom US comparison of G7 and G7 15 Day · Manufacturer — US product specifications and footnotes: age, wear, warm-up, common sensor footprint and sensor-only water rating. Direct to Watch requires a compatible phone to pair each new sensor.
  7. [7]
    Dexcom US insulin pump and pen partners · Manufacturer — US partner listings support named commercial pairings, not universal compatibility across versions or countries.
  8. [8]
    LoopDocs CGM frequently asked questions · Open-source community — Original project documentation: G7 15 Day support added in Loop 3.10.0 in January 2026. Project compatibility is not regulatory clearance.
  9. [9]
    Trio documentation: compatible CGMs · Open-source community — Lists standard G7/ONE+ support. This review did not establish explicit G7 15 Day support from this page.
  10. [10]
    AndroidAPS documentation: Dexcom G7 and ONE+ · Open-source community — Current project warnings say BYODA is obsolete and DiaKEM fails new installations after authentication changes; alternative G7 data routes are documented. No explicit 15 Day support established here. Display smoothing does not establish a universal physiological delay.
  11. [11]
    Dexcom US FAQ: use during pregnancy · Manufacturer — Manufacturer guidance permits G7/G7 15 Day during pregnancy for type 1, type 2 or gestational diabetes, subject to product age limits. This is permitted use, not proof of pregnancy-specific clinical benefit.
  12. [12]
    Dexcom US FAQ: how much does G7 cost? · Manufacturer — US access, pharmacy and coverage information; prices depend on coverage and individual eligibility. Medicare eligibility is conditional, not universal reimbursement.
  13. [13]
    Dexcom UK FAQ: upgrading from G6 to G7 · Manufacturer — Describes self-funded UK purchase and clinical-team support for funded users. Does not establish universal NHS eligibility.
  14. [14]
    Dexcom Germany G7 shop · Manufacturer — Commercial access in Germany; this does not establish access or reimbursement in every EU country.
  15. [15]
    Dexcom Canada G7 product and access information · Manufacturer — Prescription pharmacy purchasing and named private/provincial coverage pathways; not universal coverage.
  16. [16]
    Dexcom G7 to be subsidised through the NDSS · Regulatory decision · 2025-02-21 — NDSS announcement: subsidy from 1 March 2025 for eligible registered people. Eligibility, application and ordering conditions apply.
  17. [17]
    Dexcom Japan G7 product information · Manufacturer — Japanese product information: ages 2+, 10 days plus grace period and local insertion-site instructions. Do not substitute these instructions for the US label.
  18. [18]
    FDA Class II correction Z-2163-2026 — G7 iOS app · Regulatory decision · 2026-05-11 — Versions 2.9.0–2.11.2; initiated 14 April 2026. Stale glucose/trend replay and delayed alerts. Includes G7 and G7 15 Day. Follow the update action; this is an app correction.
  19. [19]
    FDA Class II correction Z-2164-2026 — G7 watchOS app · Regulatory decision · 2026-05-11 — Companion watchOS record for versions 2.9.0–2.11.2. The date is FDA posting, not the initiation date.
  20. [20]
    FDA notice: certain Dexcom receivers may miss audible alerts because of speaker failure · Regulatory decision — Affected model/serial scope and June 2025 action update. Check the serial number, request replacement if affected, and test the speaker when charging. A successful test does not rule out later failure.
  21. [21]
    Dexcom Uncovers Theft of Scrapped Product, Notifies Potentially Impacted Users — FDA-hosted announcement · Regulatory decision · 2026-05-29 — Company announcement dated 26 May; FDA publication dated 29 May. Identifies scrapped standard G7 lots 1725204004 and 1725069002, the risks and lot-check/replacement actions.
  22. [22]
    FDA Class II recall Z-2574-2026 — stolen G7 sensor lots · Regulatory decision · 2026-06-24 — Firm-initiated 27 May; FDA posted 24 June. Lot-specific scope, not a finding that all other G7 products are fault-free.
  23. [23]
    Dexcom Japan information notice: G7 Android app v2.11 · Manufacturer — Japanese manufacturer notice DOC-1000688 Rev001, dated September 2026. Versions 2.9–2.11: under specified Focus Mode/settings conditions, vibration could be suppressed while sound and visual alerts still occurred. Recommends the latest Google Play update. No exact publication day supplied; not described here as an FDA recall.
  24. [24]
    Dexcom FAQ: how much longer will G6 be available? · Manufacturer — Manufacturing deadline of 1 July 2026; remaining stock may still be available. Does not establish immediate worldwide withdrawal.
  25. [25]
    Dexcom provider FAQ: how long will G6 app support be available? · Manufacturer — G6 app support continues after the manufacturing deadline.
  26. [26]
    Dexcom UK FAQ: when will G6 be discontinued in my market? · Manufacturer — Describes a phased global transition; availability differs by market.
  27. [27]
    Dexcom G7 15 Day US user guide · Manufacturer — Current US instructions: ages 18+, 60-minute warm-up, 15 days plus 12-hour grace period; meter backup, receiver update, Direct to Watch and sensor water limits. Printed pages 1, 5–6, 38–39, 70–73 and 117 onward.
  28. [28]
    Dexcom Investor Day presentation — May 2026 · Manufacturer — Slides 23 and 28–30: future 15-day baseline, G8 target of 50% smaller size and expected late-2027/early-2028 timing. Slide 90 separately describes glucose/potassium work; these slides do not establish G8 ketone functionality. Future company targets are not cleared specifications.