This report is being submitted retrospectively as part of internal review.The review of the glucose plot showed that on (b)(6) 2023, there were 4 consecutive suspicious calibrations which led to the sensor suspend phase, where the system blinds the sensor readings for 6 hours which allows the algorithm sometime to recalibrate the glucose calculation and start in initialization phase.After the system was out of the sensor suspend phase, it restarted in initialization phase and displayed better agreement between the sensor readings and fingerstick measurements.On (b)(6) 2023, there were another 4 consecutive suspicious calibrations, resulting in another sensor suspend phase.The system is designed to trigger a sensor replacement alert when two drop-out phases occur 14 days apart within 21 days after the sensor insertion date.Further analysis of the in vivo glucose data shows that the rate of change in the glucose values frequently exceeded 5mg/dl/min.This noise is the most probable cause of the failure.The investigation shows the system correctly disabled the sensor due to performance failure and the system's self-test functions are working normally.An rma was authorized to offer the user a sensor replacement.
|