Search Results

The FreeStyle Libre 2 Flash Glucose Monitoring System is a continuous glucose monitoring (CGM) device with real time alarms capability indicated for the management of diabetes in persons age 2 and older. It is intended to replace blood glucose testing for diabetes treatment decisions, unless otherwise indicated.

The System also detects trends and tracks patterns and aids in the detection of episodes of hyperglycemia and hypoglycemia, facilitating both acute and long-term therapy adjustments. Interpretation of the System readings should be based on the glucose trends and several sequential readings over time.

The System is also intended to autonomously communicate with digitally connected devices, including automated insulin dosing (AID) systems. The System can be used alone or in conjunction with these digitally connected devices for the purpose of managing diabetes.

The FreeStyle Libre 3 Continuous Glucose Monitoring System is a real time continuous glucose monitoring (CGM) device with alarms capability indicated for the management of diabetes in persons age 2 and older. It is intended to replace blood glucose testing for diabetes treatment decisions, unless otherwise indicated.

The System also detects trends and tracks patterns and aids in the detection of episodes of hyperglycemia and hypoglycemia, facilitating both acute and long-term therapy adjustments. Interpretation of the System readings should be based on the glucose trends and several sequential readings over time.

The System is also intended to autonomously communicate with digitally connected devices, including automated insulin dosing (AID) systems. The System can be used alone or in conjunction with these digitally connected devices for the purpose of managing diabetes.

The FreeStyle Libre 2 Flash Glucose Monitoring System (with modified Sensor tail), hereinafter also referred to as 'modified FSL2 System' and FreeStyle Libre 3 Continuous Glucose Monitoring System (with modified Sensor tail), hereinafter also referred to as 'modified FSL3 System' are integrated continuous glucose monitoring (iCGM) systems that provide continuous glucose measurements every minute to provide glucose levels, trends, and real-time alarms capability to aid in the management of diabetes. The Systems also provide configurable alarms designed to warn the user of Low Glucose or Signal Loss. The user may make treatment decisions based in part on the Sensor glucose results provided by the Systems require a prescription and are intended for home use.

The FreeStyle Libre 2 System consists of the FreeStyle Libre 2 Sensor and either the FreeStyle Libre 2 Reader or the FreeStyle Libre 2 App downloaded to a compatible smartphone as a primary display device. The FreeStyle Libre 3 System consists of the FreeStyle Libre 3 Sensor and the FreeStyle Libre 3 App downloaded to a compatible smartphone as a primary display device.

Both the FreeStyle Libre 2 and FreeStyle Libre 3 Systems are compatible with the Libre Data Sharing API cleared in K223537. The display device of the connected FreeStyle Libre 2 or FreeStyle Libre 3 Systems, which directly receives the data from the Sensor, continues to serve as a primary display device for the glucose data and alarms.

The Abbott FreeStyle Libre 2 Flash Glucose Monitoring System and FreeStyle Libre 3 Continuous Glucose Monitoring System sought clearance for modified sensor tails to reduce vitamin C interference, extend sensor wear to 15 days, and enable compatibility with automated insulin dosing (AID) systems.

Here's an analysis of the acceptance criteria and supporting studies:

1. Table of Acceptance Criteria & Reported Device Performance:

The document doesn't explicitly list a table of "acceptance criteria" with specific numeric targets for device performance. Instead, it states that the devices meet the iCGM special controls for clinical performance set forth in 21 CFR 862.1355. This regulation broadly outlines requirements for accuracy, reliability, and safety of integrated Continuous Glucose Monitoring systems.

However, based on the provided text, we can infer some key performance objectives and the reported outcome:

2. Sample Size Used for the Test Set and Data Provenance:

• Sample Size: The clinical study evaluated performance in adult (18 years and older) and pediatric (2 to 17 years) participants with diabetes. Specific numbers of participants within these age groups are not provided in this summary.
• Data Provenance: The document does not specify the country of origin. It indicates the studies were clinical studies, implying a prospective nature.

3. Number of Experts Used to Establish the Ground Truth for the Test Set and Their Qualifications:

The document mentions that the clinical study evaluated performance with respect to reference venous plasma sample YSI measurements.

• Ground Truth Providers: YSI measurements are obtained from laboratory instruments, which are considered the gold standard for glucose measurement. The "experts" involved would be trained lab technicians and medical professionals responsible for collecting and processing blood samples and operating the YSI instrumentation.
• Qualifications: Not explicitly stated, but they would be certified laboratory personnel and healthcare providers responsible for patient care and sample collection within a clinical study setting.

4. Adjudication Method for the Test Set:

Not applicable. The ground truth was established by YSI laboratory measurements, not subjective expert review or adjudication.

5. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study:

A Human-in-the-loop or MRMC study is not applicable for this device type (Continuous Glucose Monitoring System). The device provides continuous glucose values, and its performance is assessed against a laboratory reference, not human interpretation of images or other data that would involve multiple readers.

6. Standalone Performance:

Yes, the studies described are primarily focused on the standalone performance of the algorithm (as embedded in the device) in measuring glucose levels against a reference method (YSI venous plasma samples). The "clinical study" directly assesses how well the device's sensor and algorithm perform without direct human intervention in the glucose reading process itself, other than the placement and use of the device by participants.

7. Type of Ground Truth Used:

The ground truth used was venous plasma sample YSI measurements. This is a highly accurate and accepted laboratory reference method for glucose concentration.

8. Sample Size for the Training Set:

The document does not specify the sample size for the training set used for the device's glucose algorithm. It mentions that the "ADC Glucose Algorithm established for the predicate device" is used, implying that the core algorithm was developed and trained previously, and this submission focuses on modifications and re-validation.

9. How the Ground Truth for the Training Set Was Established:

The document does not specify how the ground truth for the training set of the ADC Glucose Algorithm was established. However, given that the clinical performance evaluation uses venous plasma YSI measurements as ground truth, it is highly probable that similar high-accuracy reference methods would have been employed during the development and training of the original algorithm.

Ask a specific question about this device

The FreeStyle Libre 2 Flash Glucose Monitoring System is a continuous glucose monitoring (CGM) device with real time alarms capability indicated for the management of diabetes in persons age 4 and older. It is intended to replace blood glucose testing for diabetes treatment decisions, unless otherwise indicated.

The System also detects trends and tracks patterns and aids in the detection of episodes of hyperglycemia and hypoglycemia, facilitating both acute and long-term therapy adjustments. Interpretation of the System readings should be based on the glucose trends and several sequential readings over time.

The System is also intended to autonomously communicate with digitally connected devices. The System can be used alone or in conjunction with these digitally connected devices where the user manually controls actions for therapy decisions.

The System can be used with the FreeStyle Libre 2 Sensor (14 day) or the FreeStyle Libre 2 MediRx Sensor (10 day).

The FreeStyle Libre 2 Flash Glucose Monitoring System with the FreeStyle Libre 2 App -Android (herein referred to as the 'FreeStyle Libre 2 System' or 'System') is an integrated continuous glucose monitoring system (iCGM) that provides continuous glucose measurements every minute to provide glucose levels, trends, and real time alarms capability to aid in the management of diabetes. The System requires a prescription and is intended for home use. The System consists of the following components: a Sensor which transmits yia Bluetooth Low Energy (BLE), a BLE enabled display device (Reader), and an Android or iOS mobile app (FreeStyle Libre 2 App) downloaded to a compatible smartphone. Scanning of the Sensor via Reader or App provides the user with real-time glucose measurements (glucose values) accompanied by trend information (glucose arrows) and historical glucose information (glucose graph). The user may make treatment decisions based in part on the Sensor glucose results provided by the System. The System also provides configurable alarms designed to warn the user of Low Glucose, High Glucose or Signal Loss.

This 510(k) summary describes the FreeStyle Libre 2 Flash Glucose Monitoring System (with FreeStyle Libre 2 App), focusing on the addition of an Android application to the existing system. The document states that the Android app utilizes the identical algorithm and implements the same wireless interfaces with the sensor as the previously cleared Reader and iOS app. Therefore, no additional clinical data was used in this 510(k) to support a determination of substantial equivalence for clinical performance. The original clinical performance data was established in predicate devices K193371 and K211102.

Since this 510(k) relies on the clinical performance data established in prior submissions (K193371 and K211102) and explicitly states that no new clinical performance data was generated for this specific submission (K210943) regarding the Android App, the detailed acceptance criteria and study information for clinical accuracy are not directly present within this provided text for K210943. The document focuses on non-clinical performance testing for the Android app.

However, based on the information provided regarding the non-clinical performance of the Android App within this specific 510(k) (K210943):

Table of Acceptance Criteria and Reported Device Performance (Non-Clinical)

Due to the nature of this 510(k) being primarily for the addition of the Android app, the following information regarding clinical performance criteria, sample sizes, expert involvement, and ground truth would have been established in the predicate device studies (K193371 and K211102) and are not detailed in this document (K210943):

• Sample size used for the test set and the data provenance: Not specified in K210943, as clinical data was referenced from predicate devices. For iCGM, clinical studies typically involve a significant number of participants (e.g., hundreds) across varied demographics to assess accuracy across the glycemic range. Data provenance would likely be from multi-center prospective clinical trials conducted in various regions.
• Number of experts used to establish the ground truth for the test set and the qualifications of those experts: Not specified in K210943. In iCGM clinical trials, ground truth is usually established by highly precise laboratory reference methods (e.g., YSI glucose analyzer) rather than expert consensus on device readings.
• Adjudication method (e.g. 2+1, 3+1, none) for the test set: Not applicable for iCGM clinical accuracy studies where the "ground truth" is typically a laboratory reference method.
• If a multi-reader multi-case (MRMC) comparative effectiveness study was done, If so, what was the effect size of how much human readers improve with AI vs without AI assistance: Not applicable for iCGM clinical accuracy studies as described. The device provides direct glucose values, not interpretations requiring human reader assistance in the same way as imaging diagnostics. It replaces blood glucose testing for treatment decisions.
• If a standalone (i.e. algorithm only without human-in-the-loop performance) was done: Yes, the core accuracy of the algorithm is evaluated in a standalone manner against laboratory reference values in the predicate device studies. The FreeStyle Libre 2 System itself, including the sensor and its algorithm, performs autonomously to produce glucose readings.
• The type of ground truth used (expert consensus, pathology, outcomes data, etc): For iCGM devices, the gold standard ground truth for accuracy studies is typically laboratory reference methods, such as those performed by a YSI glucose analyzer, which measures glucose concentration in blood plasma.
• The sample size for the training set: Not specified in K210943. Training data for the glucose algorithm would have been part of the original development and validation of the FreeStyle Libre 2 system (K193371 and K211102), and would involve a large, diverse dataset of sensor readings and corresponding reference glucose values.
• How the ground truth for the training set was established: Not specified in K210943. Similar to the test set, the ground truth for training would have been established using highly accurate laboratory reference methods (e.g., YSI analyzer) from blood samples collected concurrently with sensor readings.

Ask a specific question about this device

The FreeStyle Libre 2 Flash Glucose Monitoring System is a continuous glucose monitoring (CGM) device with real time alarms capability indicated for the management of diabetes in persons age 4 and older. It is intended to replace blood glucose testing for diabetes treatment decisions, unless otherwise indicated.

The System also detects trends and tracks patterns and aids in the detection of episodes of hyperglycemia and hypoglycemia, facilitating both acute and long-term therapy adjustments. Interpretation of the System readings should be based on the glucose trends and several sequential readings over time.

The System is also intended to autonomously communicate with digitally connected devices. The System can be used alone or in conjunction with these digitally connected devices where the user manually controls actions for therapy decisions.

The System can be used with the FreeStyle Libre 2 Sensor (14 day) or the FreeStyle Libre 2 MediRx Sensor (10 day).

The FreeStyle Libre 2 Flash Glucose Monitoring System is an integrated continuous glucose monitoring (iCGM) system that provides continuous glucose measurements every minute to provide glucose levels, trends, and real-time alarms capability to aid in the management of diabetes. The FreeStyle Libre 2 System consists of two primary components: a Sensor that transmits via Bluetooth Low Energy (BLE), and a BLE enabled display device (Reader). User initiated RFID scanning of the Sensor via Reader provides the user with real-time glucose measurements (glucose values) accompanied by trend information (glucose arrows) and historical glucose information (glucose graph). Users may use the Sensor glucose results and information provided by the System in making treatment decisions. The System also provides configurable alarms designed to warn the user of Low Glucose, High Glucose or Signal Loss. The system is intended for single-patient use at home and requires a prescription.

The Sensor is single use, disposable, and powered by a silver oxide battery. The Sensor is provided as two secondary components, Sensor Applicator and Sensor Pack (electron beam sterilized device) which are used to assemble and apply the Sensor to the back of the user's arm. During Sensor application, the sensor tail is inserted about 5.5 millimeters below the surface of the skin through the guidance of a needle. The needle is retracted back into the applicator after insertion, and the Sensor remains attached to the skin with a medical grade adhesive. The Sensor continuously measures glucose concentration in interstitial fluid and has an 8-hour memory capacity. The Sensor is factory calibrated, does not require fingerstick calibration, and can be worn for up to 14 days.

The Reader is a small handheld device that is powered by a lithium-ion rechargeable battery and uses RFID communication to start new Sensors and to scan Sensors to display and record data and uses BLE communication to issue alarms that notify the user to scan his/her sensor when glucose values pass a high or low glucose threshold. The Reader also has a built-in strip port with blood glucose functionality (that is intended to work with the FreeStyle Precision Neo Blood Glucose test strips, cleared under K171941), and a user interface that includes event logging features.

The proposed subject device is a modified FreeStyle Libre 2 Flash Glucose Monitoring System that adds compatibility with the FreeStyle Libre 2 MediRx Sensor, which can be worn for up to 10 days. The FreeStyle Libre 2 MediRx Sensor design is unchanged from that of the predicate FreeStyle Libre 2 Sensor, which remains compatible with the modified System. In addition, the Sensor glucose algorithm and Reader design of the modified System remain unchanged from those of the predicate.

The alternate 10-day wear duration of the FreeStyle Libre 2 MediRx Sensor is achieved by changing a Sensor configuration parameter at manufacturing, which is detected by the predicate Reader to automatically determine the wear duration and accordingly adjust the user interface display of remaining Sensor wear time and ensure the Sensor cannot report data beyond the configured wear duration. In addition, each Sensor type has an end of life parameter, which determines when the Sensor will automatically shut down. This functionality is already built into the Sensor and Reader and was validated as part of previously conducted software validation under K193371.

Other than the differences related to wear duration, the FreeStyle Libre 2 MediRx Sensor is identical to the predicate Sensor, and the predicate Reader functions as intended with either the predicate FreeStyle Libre 2 Sensor (14 day) or FreeStyle Libre 2 MediRx Sensor (10 day).

The provided text is related to a 510(k) premarket notification for the FreeStyle Libre 2 Flash Glucose Monitoring System, specifically for a modification to include compatibility with a 10-day wear sensor (FreeStyle Libre 2 MediRx Sensor).

The document does not describe an AI/ML-based device where the performance is presented in terms of AI metrics (e.g., accuracy, precision, recall, AUC, etc.) or a multi-reader multi-case (MRMC) study. Instead, it concerns a medical device for continuous glucose monitoring (CGM). The "algorithm" mentioned (Sensor Glucose Algorithm) refers to the internal processing of sensor signals to derive glucose values, not an AI/ML model in the context of clinical decision support or image analysis.

Therefore, many of the requested criteria (e.g., sample size for test/training set in AI context, number of experts for ground truth, adjudication method, MRMC study, standalone performance for AI, type of ground truth for AI) are not applicable to this device's submission.

However, I can extract information related to the device's clinical performance evaluation based on the provided text, which supports its substantial equivalence.

Here's a summary of the relevant information:

1. Acceptance Criteria and Device Performance:

The document states: "Clinical data from the adult and pediatric iCGM clinical studies that supported clearance of the predicate device were re-analyzed to show that use of the subject device with the FreeStyle Libre 2 MediRx Sensor for a 10-day wear duration meets the iCGM special controls for clinical performance set forth in 21 CFR 862.1355."

This indicates that the acceptance criteria are based on the iCGM special controls outlined in 21 CFR 862.1355. The document does not provide a specific table of numerical acceptance criteria or reported device performance metrics (e.g., MARD values, clinical accuracy zones) for the FreeStyle Libre 2 MediRx Sensor in this specific 510(k) submission. It relies on the re-analysis of data from the predicate device's clearance.

2. Sample Size and Data Provenance for the Test Set:

• Sample Size Used for Test Set: Not explicitly stated for this particular 510(k). The clinical performance is based on "re-analyzed" data from "clinical studies that supported clearance of the predicate device (K193371)." To find specific sample sizes, one would need to refer to the original K193371 submission.
• Data Provenance: Not explicitly stated regarding country of origin. The data is based on "clinical studies that supported clearance of the predicate device." It's retrospective in the sense that existing data was re-analyzed for the new sensor configuration.

3. Number of Experts and Qualifications for Ground Truth:

• This question is not applicable in the context of this device. The "ground truth" for a glucose monitoring system is typically a high-accuracy reference method for blood glucose (e.g., YSI analyzer in a controlled lab setting), not expert consensus from radiologists or similar.

4. Adjudication Method for the Test Set:

• Not applicable for this type of device. Adjudication methods like '2+1' or '3+1' are common in image analysis studies where human readers provide interpretations and discrepancies are resolved. This is a continuous glucose monitoring device where performance is measured against reference glucose values.

5. MRMC Comparative Effectiveness Study:

• No, an MRMC comparative effectiveness study was not done as described for AI assistance. This device is a direct-to-patient glucose monitoring system, not an AI-assisted diagnostic tool that human readers would use to improve their performance.

6. Standalone Performance:

• Yes, in essence, standalone performance was done. The device itself provides glucose readings without human interpretation or intervention in the measurement process. The "re-analysis" of clinical data to meet iCGM special controls essentially evaluates this standalone performance of the FreeStyle Libre 2 MediRx Sensor. The output of the device (glucose value) is directly compared against a reference method.

7. Type of Ground Truth Used:

• The ground truth used for glucose monitoring devices is typically central laboratory reference glucose measurements (e.g., from a YSI glucose analyzer) taken from blood samples during clinical studies. The document states a re-analysis of "clinical studies," implying the use of such a reference method from the predicate device's trials.

8. Sample Size for the Training Set:

• Not explicitly stated/applicable in the context of AI/ML training. The "Sensor Glucose Algorithm" mentioned is likely a deterministic or model-based algorithm, not a trainable deep learning model in the sense of a "training set" for AI. If the algorithm involves parameters that were "trained" or optimized, the document does not specify the sample size used for this internal process. The primary evaluation here is of the modified sensor with an unchanged algorithm.

9. How the Ground Truth for the Training Set Was Established:

• Not explicitly stated/applicable in the context of AI/ML training. As above, the ground truth for any underlying algorithm development would refer to the reference glucose measurements used to build or validate that algorithm, but this is not typically referred to as a "training set" in the AI sense for this type of device. The document explicitly states: "the Sensor glucose algorithm... of the modified System remain unchanged from those of the predicate."

Ask a specific question about this device

The FreeStyle Libre 2 Flash Glucose Monitoring System is a continuous glucose monitoring (CGM) device with real time alarms capability indicated for the management of diabetes in persons age 4 and older. It is intended to replace blood glucose testing for diabetes treatment decisions, unless otherwise indicated.

The System also detects trends and tracks patterns and aids in the detection of episodes of hyperglycemia and hypoglycemia, facilitating both acute and long-term therapy adjustments. Interpretation of the System readings should be based on the glucose trends and several sequential readings over time.

The System is also intended to autonomously communicate with digitally connected devices. The System can be used alone or in conjunction with these digitally connected devices where the user manually controls actions for therapy decisions.

The FreeStyle Libre 2 Flash Glucose Monitoring System with the FreeStyle Libre 2 App (herein referred to as the 'FreeStyle Libre 2 System' or 'System') is an integrated continuous glucose monitoring system (iCGM) that provides continuous glucose measurements every minute to provide glucose levels, trends and alerts. The System requires a prescription and is intended for home use. The System consists of the following components: a sensor which transmits via Bluetooth Low Energy (BLE), a BLE enabled display device (Reader), and an iOS mobile app (FreeStyle Libre 2 App) downloaded to a compatible smartphone. Scanning of the sensor via Reader or App provides the user with real-time glucose measurements (glucose values) accompanied by trend information (glucose arrows) and historical glucose information (glucose graph). The user may make treatment decisions based in part on the sensor glucose results provided by the System. The System also provides configurable alarms designed to warn the user of Low Glucose, High Glucose or Signal Loss.

The provided text describes the FreeStyle Libre 2 Flash Glucose Monitoring System (with FreeStyle Libre 2 App) and its substantial equivalence to a predicate device. However, the document primarily focuses on explaining the device, its intended use, comparison with a predicate, and the general types of performance testing conducted to support substantial equivalence for the app component.

Crucially, the acceptance criteria and detailed study results proving the device meets those criteria (specific performance metrics, sample sizes for clinical trials, ground truth establishment, expert adjudication, MRMC studies, etc.) are NOT explicitly present in this excerpt. The document states "no additional clinical data beyond that provided in K193371 was used in this 510(k) to support a determination of substantial equivalence" for clinical performance, indicating reliance on prior submissions for the predicate device. Therefore, a comprehensive answer for the requested points regarding the study that proves the device meets the acceptance criteria cannot be fully extracted from this document alone.

However, based on the information provided, I can infer and answer parts of your request.

Here's an attempt to answer your questions based on the provided text, and where information is missing, I will clearly state that:

Device: FreeStyle Libre 2 Flash Glucose Monitoring System (with FreeStyle Libre 2 App)
Predicate Device: FreeStyle Libre 2 Flash Glucose Monitoring System (K193371)

A Table of Acceptance Criteria and Reported Device Performance:

The document does not provide a specific table of acceptance criteria with corresponding performance metrics. It generally states that "Results of executed protocols met the acceptance criteria" for various tests, but does not quantify these criteria or the results.

1. Sample sized used for the test set and the data provenance:

• Sample Size for Test Set: Not specified in this document. The document states that no additional clinical data was used for this 510(k) beyond what was presented in K193371 (the predicate device's submission). The performance testing described (Software V&V, Human Factors, Wireless Coexistence, Cybersecurity, Electrical Safety/EMC) are generally system-level or component-level tests, not direct patient clinical trials with specific "test sets" in the sense of clinical data.
• Data Provenance: Not specified for the clinical data that supports the predicate device (K193371). For the testing described in this document (Software, Human Factors, etc.), it's implied to be internal testing by Abbott Diabetes Care Inc. and conducted in laboratory environments. The document does not specify if it's retrospective or prospective.

2. Number of experts used to establish the ground truth for the test set and the qualifications of those experts:

• The document does not describe specific "ground truth" establishment for a clinical test set in this submission, as it defers to a previous 510(k) (K193371) for clinical performance data. Therefore, the number and qualifications of experts for establishing clinical ground truth are not provided here.
• For Human Factors testing, "usability requirements" were met, implying expert evaluation or user testing, but the specifics of who established the "ground truth" or expert qualifications are not detailed.

3. Adjudication method (e.g. 2+1, 3+1, none) for the test set:

• Not applicable/Not specified. Since clinical data specifics are not provided in this submission (deferring to K193371), there's no mention of an adjudication method commonly used in clinical trials for ground truth.

4. If a multi reader multi case (MRMC) comparative effectiveness study was done, If so, what was the effect size of how much human readers improve with AI vs without AI assistance:

• Not applicable. This device is a Continuous Glucose Monitoring (CGM) system, not an AI-assisted diagnostic imaging device that involves human "readers" interpreting cases. Therefore, an MRMC study is not relevant to this device type.

5. If a standalone (i.e. algorithm only without human-in-the-loop performance) was done:

• The document implies that standalone performance of the algorithm was assessed for the core glucose measurement. It states, "The App utilizes the identical algorithm and implements the same wireless interfaces with the Sensor as used by the Reader. As the App calculates glucose information identically to the Reader..." This suggests that the algorithm's performance in converting raw sensor measurements to glucose values was established and relied upon from the predicate device's data (K193371). No specific standalone performance metrics are provided in this document, but the fact that the algorithm is "identical" means its standalone performance is assumed to be the same as the predicate.

6. The type of ground truth used (expert consensus, pathology, outcomes data, etc.):

• For the clinical performance of a CGM, the ground truth typically involves lab-based reference glucose measurements (e.g., YSI analyzer or similar highly accurate methods) from blood samples taken concurrently with sensor readings. However, none of these specifics for ground truth are provided in this document, as it refers back to K193371 for clinical data.

7. The sample size for the training set:

• Not specified. The document does not discuss the training of the glucose algorithm. It merely confirms that the app uses the "identical algorithm" as the predicate device.

8. How the ground truth for the training set was established:

• Not specified. As the document does not detail the training set, it also does not elaborate on how its ground truth was established.

In summary, this 510(k) summary focuses on demonstrating substantial equivalence for the addition of an app to an already cleared CGM system. It relies heavily on the clinical performance and established ground truth from the predicate device's prior submission (K193371) for the core glucose measurement accuracy. The testing described in this particular document pertains mostly to software validation, human factors, wireless communication, cybersecurity, and electrical safety/EMC, which are specific to the new app component and its interaction with the existing system, rather than new clinical accuracy studies.

Ask a specific question about this device

The FreeStyle Libre 2 Flash Glucose Monitoring System is a continuous glucose monitoring (CGM) device with real time alarms capability indicated for the management of diabetes in persons age 4 and older. It is intended to replace blood glucose testing for diabetes treatment decisions, unless otherwise indicated.

The System also detects trends and tracks patterns and aids in the detection of episodes of hyperglycemia and hypoglycemia, facilitating both acute and long-term therapy adjustments. Interpretation of the System readings should be based on the glucose trends and several sequential readings over time.

The System is also intended to autonomously communicate with digitally connected devices. The System can be used alone or in conjunction with these digitally connected devices where the user manually controls actions for therapy decisions.

The FreeStyle Libre 2 Flash Glucose Monitoring System (herein referred to as the 'FreeStyle Libre 2 System or 'System') is an integrated continuous glucose monitoring system (iCGM) that provides continuous glucose measurements every minute to provide glucose levels, trends and alerts. The FreeStyle Libre 2 System consists of two primary components: a sensor that transmits via Bluetooth Low Energy (BLE), and a BLE enabled display device (Reader). User-initiated RFID scanning provides the user with real-time glucose measurements (glucose values) accompanied by trend information (glucose arrows) and historical glucose information (glucose graph). The user may use the sensor glucose results and information provided by the System in making treatment decisions. The System also provides configurable alerts and alarms designed to warn the user of Low Glucose, High Glucose or Signal Loss.

Here's a breakdown of the acceptance criteria and the studies conducted for the FreeStyle Libre 2 Flash Glucose Monitoring System, based on the provided document:

Acceptance Criteria and Device Performance

Study Information

1. Sample Size and Data Provenance:

   • Clinical Accuracy Studies (Study 1 & Study 2):
     • Study 1 (Adults): 146 subjects (91.1% Type 1, 8.9% Type 2 diabetes, all requiring insulin).
     • Study 2 (Pediatric): 139 subjects (98.6% Type 1, 1.4% Type 2 diabetes, all requiring insulin).
     • Age Groups in Study 2 (for YSI comparison): 77 pediatric subjects for <70 mg/dL range, 124 for 70-180 mg/dL and >180 mg/dL range. Specific breakdown for 4-5 and 6-17 years where YSI was used.
     • Age Groups in Study 2 (for SMBG comparison): 8 pediatric subjects (4-5 years old) only had SMBG as comparator.
     • Data Provenance: Prospective, multi-center studies conducted at five (adult) and four (pediatric) centers in the United States.
   • Ascorbic Acid Interference Study: 60 subjects (adults, age 18 and older) with diabetes, across 4 sites.
   • Precision Study (from Clinical Studies):
     • Adults (18+): 146 subjects, 26,791 paired readings.
     • Pediatric 4-5 years: 7 subjects, 248 paired readings.
     • Pediatric 6-17 years: 130 subjects, 10,623 paired readings.

2. Number of Experts and Qualifications for Ground Truth:

   • The document implies that the ground truth for glucose values was established using a "laboratory glucose reference" (YSI) analyzer. While YSI analyzers are highly accurate, the document does not explicitly state the number of experts or their qualifications for establishing the ground truth. It refers to the YSI as the comparator method, which is a laboratory instrument, not human experts. For pediatric subjects aged 4-5 years, Self-Monitoring Blood Glucose (SMBG) tests were used as the comparator method.

3. Adjudication Method for the Test Set:

   • Not applicable in the traditional sense of medical image adjudication. The "ground truth" was established by a laboratory reference method (YSI for most cases, SMBG for some pediatric cases). There's no indication of an adjudication process involving multiple human readers to resolve discrepancies in ground truth measurements. The comparison is between the device and the reference instrument.

4. Multi-Reader Multi-Case (MRMC) Comparative Effectiveness Study:

   • No, an MRMC comparative effectiveness study was not done. This type of study typically involves human readers interpreting cases with and without AI assistance. The FreeStyle Libre 2 is a standalone diagnostic device that measures glucose directly, so its performance is compared against a reference standard (YSI, SMBG), not human interpretation with or without AI.

5. Standalone (Algorithm Only) Performance:

   • Yes, the studies primarily assessed standalone performance. The device automatically provides continuous glucose measurements. The "iCGM glucose values" are the output of the algorithm. The intent is for the device to replace blood glucose testing for treatment decisions, which is a standalone use case. Human interpretation of trends and patterns is part of the overall management, but the device's accuracy itself is assessed in a standalone manner against the reference.

6. Type of Ground Truth Used:

   • Laboratory Reference and Self-Monitoring Blood Glucose (SMBG):
     • For adult subjects and pediatric subjects aged 6-17 years, the primary ground truth was established by YSI glucose measurements. YSI (Yellow Springs Instrument) is a common laboratory method considered a highly accurate reference for glucose.
     • For pediatric subjects aged 4-5 years, the ground truth was established by SMBG (Self-Monitoring Blood Glucose) tests.

7. Sample Size for the Training Set:

   • The document mentions "clinical evaluation" and "pivotal clinical studies" but does not explicitly state the sample size of a training set specifically for algorithm development or machine learning. The provided clinical studies (Study 1 and Study 2) describe the test sets used for performance evaluation, not necessarily for training the algorithm. CGM devices typically use complex calibration algorithms developed with extensive internal data that is not always detailed in 510(k) summaries.

8. How the Ground Truth for the Training Set Was Established:

   • As the document does not provide details on a specific training set or its size, it also does not elaborate on how the ground truth for such a set was established. We can infer that similar high-accuracy laboratory reference methods (like YSI) would have been used during the development and calibration phases of the device, prior to the validation studies described in the submission.

Ask a specific question about this device