CGM Device Guide

Abbott FreeStyle Libre 2 and 3

The same electrochemistry. The same algorithm. Different form factors and connectivity. The FreeStyle Libre family has the highest market share in the UK — and the strongest ±20/20 agreement rate of any CGM in this guide for the Libre 3.

CGM Abbott

Framework status – FreeStyle Libre family

5 / 5

All five framework criteria met
iCGM designation: Yes – FSL2 (FDA, June 2020), FSL3 (FDA), FSL3 Plus

✓ T1D included – studied in people with type 1 diabetes

✓ Meal challenge included – rapid glucose rises tested, not just stable periods

✓ Hypoglycaemia challenge included – accuracy tested in the low range

✓ Venous blood reference – the gold standard reference method

✓ Population and indication match – adults and children. Libre 3: 4 years and older

Data sufficiency: Met – multiple large-scale accuracy studies across the family. Vaughan et al. (2025) meta-analysis covers a decade of evidence.

What is the same across the family

The FreeStyle Libre 2 and 3 use the same electrochemical sensing technology and the same algorithm. Abbott has improved the sensor electronics, miniaturisation, and connectivity across generations – but the underlying measurement principle is consistent.

This means the accuracy improvements between generations are real but measured. The Libre 3 achieves a lower MARD (7.9% vs 9.3% for the Libre 2) largely through improved calibration and electronics, not through a fundamentally different approach to glucose measurement.

Clinically, both devices are non-adjunctive – insulin dosing without a confirmatory finger prick is approved. Both have adequate data for framework inclusion. The differences are form factor, connectivity, and wear duration.

Device comparison

Feature	FreeStyle Libre 2	FreeStyle Libre 2 Plus	FreeStyle Libre 3	FreeStyle Libre 3 Plus
Wear duration	14 days	15 days	14 days	15 days
±20/20 agreement	~93%	~93%	93.4%	93.4%
Outside ±40/40	<0.1%	<0.1%	<0.1%	<0.1%
Form factor	Standard size	Standard size	~70% smaller – world’s smallest CGM at time of launch	~70% smaller
Reading method	Scan (NFC) + optional Bluetooth alarms	Scan + Bluetooth alarms	Continuous Bluetooth, every 1 minute	Continuous Bluetooth, every 1 minute
Scanning required	Yes (to view current reading)	Yes (to view)	No – pushes automatically	No – pushes automatically
Separate reader	Optional	Optional	No reader available	No reader available
Calibration	Not required, not offered	Not required, not offered	Not required, not offered	Not required, not offered
AID compatible	No	Omnipod 5	CamAPS FX / YpsoPump	CamAPS FX / YpsoPump
Age indication	All ages	All ages	4 years and older	4 years and older
iCGM (FDA)	Yes (June 2020)	Yes	Yes	Yes

[Image: FreeStyle Libre 2 and FreeStyle Libre 3 size comparison – awaiting editorial images from Abbott]

Accuracy

The accuracy figure that matters is the ±20/20 agreement rate — the proportion of readings close enough to the true blood glucose to support an insulin dosing decision without a confirmatory finger prick. For the Libre 3, this is 93.4% overall in the pivotal study. The DSNFUK CGM comparison chart, on which this guide is built, uses ±20/20 and ±40/40 agreement rates as the primary accuracy measures. MARD — a statistical average — tells you little about what matters clinically and is not used here as a primary measure.

Vaughan et al. (2025) provides the most comprehensive overview – a meta-analysis of a decade of FreeStyle Libre accuracy data across the full sensor family.

The black swan – when to trust and when to check

Between 0.5% and 1% of readings from framework-qualified CGMs fall outside the ±40/40 accuracy band. This applies to the Libre family as it does to all other approved CGMs. Despite the lowest published MARD of any CGM, the FreeStyle Libre 3 is not immune to this. See the accuracy page for the full precautionary framework and when a finger prick is still needed.

AID system compatibility

AID compatibility splits across the Libre family. The FreeStyle Libre 2 Plus integrates with Omnipod 5 – Abbott’s current AID partnership with Insulet. The FreeStyle Libre 3 (and 3 Plus) integrates with CamAPS FX on the YpsoPump. Neither generation integrates with t:slim X2 / Control-IQ or MiniMed 780G.

AID system	Pump	Libre 2 Plus	Libre 3 / 3 Plus
Omnipod 5	Insulet tubeless pod	Yes	No
t:slim X2 / Control-IQ	Tandem tubed pump	No	No
CamAPS FX	YpsoPump (tubed)	No	Yes
MiniMed 780G	Medtronic tubed pump	No	No

For patients on or considering Omnipod 5, the FreeStyle Libre 2 Plus is the compatible Abbott option. For patients on CamAPS FX / YpsoPump, the Libre 3 is approved. Patients on t:slim X2 or MiniMed 780G who want an Abbott CGM are not currently served by an AID integration – the Dexcom G7 is the CGM choice for those systems. For patients on MDI or pump without AID, either Libre family device is appropriate based on clinical factors and patient preference.

Features

Libre 3 – continuous Bluetooth streaming

The Libre 3 automatically sends a reading every minute to the LibreLink app without the user needing to scan. This is the most significant behavioural difference between generations – passive monitoring rather than active scanning. No Bluetooth scanning range limit on older models.

Libre 2 – scan to read

The Libre 2 requires the user to hold their phone or reader close to the sensor to obtain a reading. Real-time alarms are sent via Bluetooth (the phone does not need to be held to the sensor for alarms), but the glucose reading itself requires a scan. This can limit visibility during sleep, exercise, or when the phone is not nearby.

Alarms and alerts

Both Libre 2 and Libre 3 offer customisable high and low glucose alarms. The Libre 2 alarm architecture means the phone or reader must be within Bluetooth range (~10 metres). The Libre 3 app receives continuous data, making alert delivery more reliable.

LibreLinkUp – sharing

The LibreLinkUp app allows family members, carers, and HCPs to view real-time glucose data from a connected Libre. Useful in paediatric and young adult settings.

No calibration

The FreeStyle Libre family does not require or accept calibration. If you believe the reading is wrong – because of symptoms, a clinical picture that does not match, or sensor behaviour that concerns you – a finger prick blood glucose test is the appropriate response.

LibreView Pro – integration with GP records

LibreView Pro integrates directly into SystmOne and EMIS, allowing CGM metrics – time in range, GMI, average glucose – to appear inside the electronic record system without switching windows. Metrics are SNOMED-coded, which matters for audit and may become relevant to QOF in primary care as CGM use expands. Abbott provides education and mentorship support for primary care teams adopting the technology.

Low glucose events – what the data does and does not capture

Low glucose events on LibreView are recorded only when glucose drops below 3.9 mmol/L and remains there for 15 minutes or more. A patient who treats a hypo quickly and effectively may have no recorded low glucose event – despite experiencing a genuine hypoglycaemic episode. Healthcare professionals should be aware of this when reviewing downloads and should not dismiss a patient’s report of a hypo because the data does not show one. The minute-by-minute sensor trace may show the event even when the 5-minute graph and event log do not.

Connectivity tips

The LibreLink app needs to remain running in the background for uninterrupted Bluetooth connectivity. Swiping the app closed can interrupt the sensor connection. For patients experiencing repeated disconnections, clearing the phone cache and keeping the LibreLink app visible on the home screen are the most effective troubleshooting steps before escalating to Abbott customer care.

UK context

Abbott FreeStyle Libre holds the largest market share of any CGM in the UK, with the Libre 2 and Libre 3 widely prescribed in both NHS primary and secondary care. The Libre 3 is available on FP10 in England. NHS formulary position and prescribing routes vary by ICB – patients who do not have access should contact their ICB. Abbott offers face-to-face and remote starter clinics for healthcare teams and primary care practices, and provides mentorship and education support for HCPs new to interpreting CGM data.

CGM is now also recommended for people with type 2 diabetes who use insulin injections. Access is currently uneven across ICBs – patients not yet receiving it should ask their GP or diabetes team and be prepared to advocate.

Clinical evidence

Vaughan N et al. (2025). Meta-analysis of a decade of studies assessing accuracy of Abbott FreeStyle Libre continuous glucose monitoring. Diabetes Technology and Therapeutics.
Alva S et al. (2020). Performance of the FreeStyle Libre 2 – second generation iCGM. Paediatric and adult accuracy data.
Abbott (2022). FreeStyle Libre 3 pivotal accuracy study. MARD 7.9% overall. ±20 agreement 93.4%. Parkes error grid A+B 99.9%.
Alva S et al. (2023). FreeStyle Libre 3 accuracy in children aged 4-17. MARD 8.7% (6-17y), 10.1% (4-5y).

GNL Podcast Episode 38 – Michael Skarlatos, Abbott Medical Affairs, speaks with John Pemberton about the FreeStyle Libre family, AID integrations, LibreView Pro and its integration into SystmOne and EMIS, what makes for effective sensor use, and a glimpse at what is coming next from Abbott. Watch on YouTube or find Episode 38 on theglucoseneverlies.com.

Forthcoming: Abbott have a continuous ketone monitoring sensor in development. As of ATTD 2025, validation work is ongoing. This page will be updated when published data is available.