|
DEXCOM, INC. DEXCOM G6 CGM SENSOR ; INTEGRATED CONTINUOUS GLUCOSE MONITORING SYSTEM, FACTORY CALIBRATED
|
Back to Search Results |
|
| Lot Number 5252867 |
| Device Problems
Device Difficult to Program or Calibrate (1496); Incorrect, Inadequate or Imprecise Result or Readings (1535)
|
| Patient Problem
No Known Impact Or Consequence To Patient (2692)
|
| Event Date 10/29/2019 |
|
Event Type
malfunction
|
|
Event Description
|
|
We have been experiencing an extremely high failure rate for dexcom g6 cgm sensors since switching from g5.The g6 system is significantly better than g5 in most ways when all is working, but the sensors are apt to fail at least 25% of the time or higher.This has been so frustrating for our 16 y/o son with t1d that i discussed it with his cde at (b)(6) medical ctr and learned that this is a common experience with their pts.Dexcom when contacted both directly and indirectly via tandem diabetes, readily replaces the equipment but that does not change the difficulty, inconvenience, stress and most importantly, the increased time spent without a functioning cgm.All of that detracts from the benefits that cgms provide and add burden to dealing with a chronic disease.It also adds complexity to planning and mgmt of cgm supplies.We had to be particularly mindful of sending many extra sensors for a 3 week camp session in order to allow for failure which did happen on 07/29/2019.That, in turn, requires exerting pressure on supplier and insurers to ship sensors in advance of their expectations based on 10-day wear time.In practice, we have seen the following issues and have reported them both to dexcom and to tandem since the cgm is integrated into the t:slim x2 insulin pump: one out of every 3-5 sensors frequently will not function correctly.The following is an example of a typical experience and occurred while away at camp: sensor inserted in evening of 07/29 with factory calibration code.Readings begin after 2 hour warm - up while sleeping.Cgm alarms for low bg in 60s and then the 40s.However, finger testing confirms that actual bg is between 158-174 at those times.Manual calibrations are entered but dexcom rejects them the a calibration error.Testing and calibration are done repeatedly with the same result and during this time there are no readings.He gives up and sleeps the night without the cgm.In the morning, a calibration succeeds but the sensor starts reporting 181 when the tested bg value is 101.Since it is useless, we decided to replace the sensor again.When first started, that one reports a calibration error after 15 mins even though it should be in 2 hour warm-up.We decide to cancel and restart the sensor session with the same calibration code.This time it works and has been working ever since.We contact dexcom for replacement sensor.This scenario has happened many times.There appear to be 2 related problems.First, that the calibration codes do not always result in accurate bg readings which is the sole purpose of the cgm.Second, the g6 algorithm often rejects manually entered bg values rather than using them to correct its readings.Repeated attempts to manually calibrate either have no effect or cause a "sensor failed" error at which point you must change the sensor.The bottom line is that the combination of these defects is a significant problem that dexcom needs to address going forward.Fda safety report id #(b)(4).
|
| |
|
Event Description
|
|
Add'l info received from reporter on 09/05/2019 for mw5088768; on (b)(6) 2019 (b)(6) inserted a new dexcom g6 sensor.After the 2 - hr warmup, the sensor started reporting readings in the low 40s.That did not make sense or seem accurate so he checked with a finger stick and the glucometer reported 93.He calibrated the sensor with 93 and it initially appeared to accept it and continued reporting a few values in that range.However, he then went to bed and, while sleeping, the sensor stopped reporting any values at all between 1 am and about 10 am when he woke up.When he got up, he tested and tried to calibrate several times and each time the sensor reported an error and refused the calibration.His bg overnight was fine but changes during sleep are potentially extremely dangerous and is precisely the type of situation that the cgm is intended to monitor.This occurred with sensor lot 5256707 which had a calibration code of 5931.Since we could not get the sensor to work, we removed on (b)(6) 2019 and replaced it with another from the same lot (5256707) with calibration code 9551.That sensor worked correctly, without issue.It should be noted that all working sensors from the same lot had calibration codes 9xxx while the failed one had code 5xxx.We have found a high correlation between sensor failures and calibration codes 5xxx.On (b)(6) 2019, it was time to change sensors again.He used another g6 sensor from lot code 5256707 with calibration code 9551.After the warm-up period the sensor began reporting values in the upper 40s and low 50s which seemed extremely unlikely given what he'd recently eaten.Indeed, a finger check indicated that bg was 88.He calibrated the sensor at 88, expecting it to fail.This time however, it accepted the reading and has been working correctly.Nonetheless, the g6 has been approved for use without fingerstick calibrations and yet it does not seem to work properly a significant percentage of the time.Pts (or closed loop system) relying on readings without verification would make decisions based on inaccurate bg data.In this case, two out of three in a single box had issues, with one failing completely.This is not an isolated situation.The g6 system is excellent when it works, but its consistent reliability is lacking and in need of improvement.
|
| |
|
Event Description
|
|
Add'l info received on 11/05/2019 for report mw5088768.This is an add'l f/u to previous medwatch reports submitted on 08/01/2019 regarding dexcom g6 sensor issues that occurred on (b)(6) 2019, (b)(6) 2019 regarding dexcom g6 sensor issues that occurred on (b)(6) 2019.As mentioned in previous report, the quality and reliability of g6 sensors has been noticeably poorer in later 2019 than it has been in the past.In any 90 day period, we estimate that we are now experiencing a sensor failure rate of between 25% and 33%.On (b)(6) 2019, after a successful 10-day sensor session expired, our son (b)(6), inserted a new one in his upper left arm from lot 5256698 with calibration code 5937 (udi: (b)(4)).Within about ten mins, instead of receiving the two-hour warm-up message, the tandem t:slim x2 pump was displaying errors.We immediately restarted the sensor session but the errors continued for another 45 mins until a sensor failure message was received.We called tandem diabetes for advice and they told us to change the sensor and sent a replacement.We then removed the bad sensor and inserted another one on a different location on the upper left arm.This second sensor was form lot 5258572 with calibration code 9311.Although the session warm-up began normally, when it began producing readings during the night it started causing low bg alarms due to inaccurate readings in the low to mid fifties.A finger stick check on a glucose meter showed actual bg to be 128.This continued for several hours clearly interrupting everyone's sleep.More significantly however, it also triggered the tandem basal iq to stop basal insulin delivery.That problem lies in the sensor providing faulty data to the tandem algorithm and not with basal iq at all.Nonetheless that is potentially quite dangerous, especially at night.We temporarily disabled basal iq and by the morning the sensor seemed to have corrected itself and was in line with glucometer readings.Such behavior is not supposed to be occurring in the g6 which was approved for use without finger sticks and is one of the major advances over g5.I want to reiterate from earlier reports that we have found a high correlation between sensor failures and calibration codes 5xxx.Whatever the significance, we have noticed a smaller chance of failure for sensors shipped with calibration codes 9xxx.Additionally, i want to report a similar chain of events that occurred throughout (b)(6) 2019.In that case, a g6 sensor from lot 5258165 with calibration code 9117 was inserted on (b)(6) 2019.It worked initially and then suddenly stopped reading and then failed on (b)(6) 2019 (before the 10 day wear time had passed).It was replaced on (b)(6)2019 with a new one from lot 5252867 with calibration code 9117.That second one also walked at first and then began to sporadically drop readings including an entire afternoon.We finally gave up and replaced it again with a third sensor from lot 7258395 with calibration code 9311 which functioned normally.This series of failures was reported directly to dexcom for replacement.It is clear from these events and the others previously reported that between (b)(6) and (b)(6) we have encountered a highly unacceptable sensor failure rate.We have discussed this with our endocrinology practice and they are seeing this as a widespread issue.The g6 inconsistency is frustrating, dangerous, costly and will decrease the effectiveness of upcoming control iq as well as future ap systems.
|
| |
|
Search Alerts/Recalls
|
|
|
