Anyone having big reliability issues with Freestyle Libre 2 Plus sensors?

[SimonP78]

I have been using the Libre 3 and 3 Plus sensors for a year or so, now. It's rare that I don't experience a failure on a sensor, which either completely fails, or which goes low repeatedly, causing alarms. Subsequent finger sticks with a ReliOn Platinum glucometer show the blood glucose in an acceptable range, while the sensor reading (on my phone) shows red and low. Calls to Abbott have yielded replacements, but only after lengthy calls to operators abroad; frustrating calls with operators who use broken English, and who blame me, the installation, my phone, or anything else they can try, rather than accept that the sensor has gone bad. The readings are often inaccurate, and I take regular finger-sticks and record them, to verify. It's not merely a matter of a time delay; the difference is significant, and I can't imagine trying to make insulin decisions based on the CGM readings. They're wildly unreliable. I'm wearing one now in my right arm; my blood glucose is 95, while the sensor reading flags at 63, with a red low warning. It does not look like the sensor will last it's full period, again.

Apologies...new to the forum, visiting. I'm posting in the Type 1 forum, but have type II, and the thread is Libre 2; I'm using Libre 3. I thought the comments might be useful, however. I found considerable negative reports on the Libre 2 originally, and went for the 3 (which was supposed to be an improvement). My apologies if the commentary is misplaced.

I feel your pain re sensors: https://forum.diabetes.org.uk/boards/threads/cgm-log-how-long-do-they-last-why-do-they-fail.114108

 

[Hobo1960]

Last two Libre 2 Plus sensors have failed at 10 days and 8 days. I previously used Libre 2 and can only remember having one failure. I’ve had type 1 diabetes for 49 years, have had tight control with no complications. I find the whole freestyle Libre system has caused me more anxiety than carrying out multiple finger prick tests. The discrepancies with sensors and “is it or isn’t it correct” has added to the anxiety! I don’t want to control my diabetes on that basis and am now giving careful consideration to reverting back to finger prick tests. My diabetes nurse did warn me that changing over to Freestyle Libre may cause anxiety as my diabetes is well controlled. She said that it offered improved control to diabetics with poorer control, but was unlikely to offer me the same benefits. Has anyone else come to the same conclusions?

 

[Rosie9876]

So I have been using Libre 2 for a while now and haven't had 1 sensor last 15 days. They have lasted between 8 hours and 11 days at best. I have found them very unreliable. Has anyone else had this issue? Does anyone know why these sensors are so unreliable?

I've been using Libre 2+ for several months. Only one fell off and Abbott replaced it promptly. However, I don't trust its accuracy. I don't do finger prick testing, but was tested in hospital, and of about 8 results, only 2 matched the Libre closely. In all other cases, Libre was lower by between 1 and 2.5 mmol/l. I still find it useful for checking fluctuations, but if I were Type1, I wouldn't want to decide my insulin injections on it.

 

[Hobo1960]

A difference of 1 to 2.5% is good. I find the Libre’s slow response (lag) to blood sugar changes is worrying. I regularly go out for long walks in the countryside.When I feel that I’m going low I eat some carbs. The Libre eventually catches up and tells me I’m low and keeps on saying I’m low. I respond by eating more carbs. I finish my walk and test my blood sugar and it’s 11mmol. The Libre says I’m 5! After an hour it catches up and gives a fairly accurate reading. The time lag on the Libre is ridiculous if you’re doing physical exercise. I’ve switched off the alarms and go on how I feel. If I feel low, I eat. If I want to know my blood sugar I use a meter. The Libre is in my opinion only useful when you have periods of stable blood sugar readings, this is unlikely to occur if you’re exercising or doing anything strenuous! These are the times when you’re most at risk of a hypo, but the Libre is too slow to be much use and will continue to warn you’re low when you aren’t? Unfortunately, the Libre is a case of information overload. Do I really need to know how many times I’ve been hypo when the information isn’t accurate. Do I want to make adjustments to insulin when the information is in accurate? Do I want to be woken up several times a night because Libre says I’m Hypo when I’m not? I’d prefer to have less, but accurate information. The only thing that matters to me is having accurate blood sugar readings because that’s the information I need to manage my diabetes, but each to their own.

 

[Jasmin2000]

I only use the Libre2+ to tease my DSN - last week it was at 2.4-2.9 for about 12 hours and my glucometer was 4.5-6.8 throughout this period, but he doesn't trust my readings. Tough.

Seriously, though, as @Hobo1960 and others have said Libre's are only accurate when BG is stable, so depending on your lifestyle it may be better to rely on a glucometer - I do intense exercise at odd times, eat low carb high protein diet and am up at all hours. The poor Libre can't handle it and I'm regularly in the red all night, while my protein drives BG says all green over the same period. I'm a candidate for a Lancet paper!

 

[Hobo1960]

I only use the Libre2+ to tease my DSN - last week it was at 2.4-2.9 for about 12 hours and my glucometer was 4.5-6.8 throughout this period, but he doesn't trust my readings. Tough.

Seriously, though, as @Hobo1960 and others have said Libre's are only accurate when BG is stable, so depending on your lifestyle it may be better to rely on a glucometer - I do intense exercise at odd times, eat low carb high protein diet and am up at all hours. The poor Libre can't handle it and I'm regularly in the red all night, while my protein drives BG says all green over the same period. I'm a candidate for a Lancet paper!

Good to know that I’m not alone in how I feel about the Libre. I was also worried that my Diabetes nurse would believe all the erroneous Libre data. Fortunately, she’d tried a Libre 2 sensor and was getting hypo alarms in the middle of the night. She doesn’t have diabetes and her husband wasn’t vey happy with having his sleep disturbed! I’d done the same as you and had kept a record of meter readings to show the levels of inaccuracy.

 

[Jasmin2000]

Good to know that I’m not alone in how I feel about the Libre. I was also worried that my Diabetes nurse would believe all the erroneous Libre data. Fortunately, she’d tried a Libre 2 sensor and was getting hypo alarms in the middle of the night. She doesn’t have diabetes and her husband wasn’t vey happy with having his sleep disturbed! I’d done the same as you and had kept a record of meter readings to show the levels of inaccuracy.

Wait - a diabetes nurse that tried a sensor? That's a good one - keep hold of her!

 

[Hobo1960]

Wait - a diabetes nurse that tried a sensor? That's a good one - keep hold of her!

The rep from Abbott visited the Practice prior to the full rollout of Libre 2 to all type 1 diabetics. He gave the nurses the opportunity to try the sensors out for themselves and said that as they would be instructing patients it would probably be helpful for them to do so.

 

[Pinkfimble]

So I have been using Libre 2 for a while now and haven't had 1 sensor last 15 days. They have lasted between 8 hours and 11 days at best. I have found them very unreliable. Has anyone else had this issue? Does anyone know why these sensors are so unreliable?

My daughter used these for a couple of years and had constant issues. New sensors gave unreliable readings for the first 24 hours, some sensors consistently gave readings that were too high/low, some died in the bath. One that constantly said she was having a hypo, another that left the metal application needle in her arm! Dexcom are much better.

 

[Hobo1960]

My daughter used these for a couple of years and had constant issues. New sensors gave unreliable readings for the first 24 hours, some sensors consistently gave readings that were too high/low, some died in the bath. One that constantly said she was having a hypo, another that left the metal application needle in her arm! Dexcom are much better.

Is your daughter using a pump? I’m on multiple injections and I don’t Know of anyone on injections that has been offered Dexcom sensors. I’ve been told that I can have as many blood testing strips as I need. I feel really sorry for diabetics that have had their testing strips restricted because they’re using Freestyle Libre. I’ve had accurate readings, but the performance of the device has been inconsistent. The failure rate of the Libre 2 Plus has been dreadful. The Libre 2 worked better for me, but was still very inaccurate at times when my blood sugar was changing rapidly. This is a problem and can only be addressed by carrying out backup finger prick tests. I know of diabetics that trust the Libre 2 entirely and hardly ever carry out finger prick tests. I wonder what they’re missing?

 

[becca59]

I don’t feel we’re missing anything. My TIR is excellent I’ve only had two hypos in the last three months. If I feel the sensor is miss behaving I do increase blood tests. But the truth is pre Libre we were not on board with what our levels were constantly and micro managing them. We did blood tests, but not constantly. And we accepted what they told us.
As a point of interest all type ones are eligible for Dexcom one. If you ask to change they cannot refuse. The choice is in the guidelines.

 

[Hobo1960]

I don’t feel we’re missing anything. My TIR is excellent I’ve only had two hypos in the last three months. If I feel the sensor is miss behaving I do increase blood tests. But the truth is pre Libre we were not on board with what our levels were constantly and micro managing them. We did blood tests, but not constantly. And we accepted what they told us.
As a point of interest all type ones are eligible for Dexcom one. If you ask to change they cannot refuse. The choice is in the guidelines.

Thanks for the info about Dexcom One. I’ve just had a look at the NHS Formulary Guidance for my area and the Dexcom is listed, so thanks for that.

Out of interest, the annual cost to the NHS of the Dexcom One+ is £911.41 and the Freestyle Libre 2 Plus is £912.50 per year.

Well done on your hypo count. It would be impossible with my lifestyle not to have low blood sugar from time to time. I can go from sedentary to slogging my guts out in a matter of minutes. My hypo awareness is excellent and my low is quickly self treated. My diabetes is tightly controlled and with my regime I would be concerned if I didn’t have an occasional low as I’d start worrying about running my blood sugar too high. After 49 years things have changed a lot. When I was diagnosed in 1976, only hospitals had glucose meters and they were cumbersome and took several minutes to give a result. It wasn’t something that you could put in your back pocket. For the first few years I monitored my diabetes by using Clinitest tablets. Every morning I had to collect a urine sample, put 4 drops into a test tube and then add a tablet to it that caused the urine to boil and change colour. Blue indicated no sugar in the urine. Yellow 1% and Orange 2%. How times have changed! I must have been doing something right as I’m free of complications. When I was first diagnosed my consultant said “If you don’t control your diabetes, the diabetes will control you”. Those words have stayed with me me!

 

[Rosie9876]

Last two Libre 2 Plus sensors have failed at 10 days and 8 days. I previously used Libre 2 and can only remember having one failure. I’ve had type 1 diabetes for 49 years, have had tight control with no complications. I find the whole freestyle Libre system has caused me more anxiety than carrying out multiple finger prick tests. The discrepancies with sensors and “is it or isn’t it correct” has added to the anxiety! I don’t want to control my diabetes on that basis and am now giving3 careful consideration to reverting back to finger prick tests. My diabetes lnurse did warn me that changing over to Freestyle Libre may cause anxiety as my diabetes is well controlled. She said that it offered improved control to diabetics with poorer control, but was unlikely to offer me the same benefits. Has anyone else come to the same conclusions?

Oops. Note I corrected the post you replied to @Hobo1960. I meant to write mmol/l, not percent on glucose level. To me, a peak reading 9mmol/l, when it is really 11.5 is misleading. False lows even more worrying.

A difference of 1 to 2.5% is good. I find the Libre’s slow response (lag) to blood sugar changes is worrying. I regularly go out for long walks in the countryside.When I feel that I’m going low I eat some carbs. The Libre eventually catches up and tells me I’m low and keeps on saying I’m low. I respond by eating more carbs. I finish my walk and test my blood sugar and it’s 11mmol. The Libre says I’m 5! After an hour it catches up and gives a fairly accurate reading. The time lag on the Libre is ridiculous if you’re doing physical exercise. I’ve switched off the alarms and go on how I feel. If I feel low, I eat. If I want to know my blood sugar I use a meter. The Libre is in my opinion only useful when you have periods of stable blood sugar readings, this is unlikely to occur if you’re exercising or doing anything strenuous! These are the times when you’re most at risk of a hypo, but the Libre is too slow to be much use and will continue to warn you’re low when you aren’t? Unfortunately, the Libre is a case of information overload. Do I really need to know how many times I’ve been hypo when the information isn’t accurate. Do I want to make adjustments to insulin when the information is in accurate? Do I want to be woken up several times a night because Libre says I’m Hypo when I’m not? I’d prefer to have less, but accurate information. The only thing that matters to me is having accurate blood sugar readings because that’s the information I need to manage my diabetes, but each to their own.

 

[Hobo1960]

I wondered whether you’d mistakenly typed percent. If I’m below 9 mmol within 2 hours of injecting, I won’t generally give an additional insulin injection. If I’m 9 or higher I will, unless I’m planning to do some exercise. I live in Devon and a brisk walk up a hilly lane is guaranteed to lower my blood sugar levels by several points. If I’m out on the coast path doing a 14 mile walk, I can half my insulin, have a large pub lunch and keep in balance. If I’m going to make insulin adjustments I always try to check my blood glucose first. I recall that this was how we were told to do it, but I believe there has been more recent guidelines from some doctors saying that you can do it on the basis of the sensor readings. In my experience the Freestyle Libre tends to under read on the lower numbers. I’m guessing that this is for safety reasons. An untreated hypo could do you harm, but over eating to correct a non existent low is unlikely to cause you harm in the short term. Just imagine the claims against manufacturers where someone has suffered a bad hypo because a sensor gave an erroneous reading. The manufacturers are reasonable in advising us to check using a glucose meter if the way we feel/symptoms don’t match the sensor readings.

I’ve just signed up to a free 10 day trial of the Dexcom One + sensor and will see how that works out

 

[Rosie9876]

I wondered whether you’d mistakenly typed percent. If I’m below 9 mmol within 2 hours of injecting, I won’t generally give an additional insulin injection. If I’m 9 or higher I will, unless I’m planning to do some exercise. I live in Devon and a brisk walk up a hilly lane is guaranteed to lower my blood sugar levels by several points. If I’m out on the coast path doing a 14 mile walk, I can half my insulin, have a large pub lunch and keep in balance. If I’m going to make insulin adjustments I always try to check my blood glucose first. I recall that this was how we were told to do it, but I believe there has been more recent guidelines from some doctors saying that you can do it on the basis of the sensor readings. In my experience the Freestyle Libre tends to under read on the lower numbers. I’m guessing that this is for safety reasons. An untreated hypo could do you harm, but over eating to correct a non existent low is unlikely to cause you harm in the short term. Just imagine the claims against manufacturers where someone has suffered a bad hypo because a sensor gave an erroneous reading. The manufacturers are reasonable in advising us to check using a glucose meter if the way we feel/symptoms don’t match the sensor readings.

I’ve just signed up to a free 10 day trial of the Dexcom One + sensor and will see how that works out

It would be interesting to have a Libre on at the same time as the Dexcom, and see how they compare, also with a finger prick metre.

 

[Hobo1960]

It would be interesting to have a Libre on at the same time as the Dexcom, and see how they compare, also with a finger prick monitor.

Good idea, I might try that. I’ve had a look at some comparison videos on YouTube and there doesn’t appear to be much difference in overall accuracy. I’m just hoping that reliability is better on the Dexcom, but will have to wait and see.

 

[Jasmin2000]

It would be interesting to have a Libre on at the same time as the Dexcom, and see how they compare, also with a finger prick metre.

Been there, done that.
Dexcom (right arm) was closer but lower than the Glucometer. but within acceptable error (given that there's a margin of error on both). Libre2 (left arm) was consistently lower and way off during exercise and other rapid BG changes.

The technologies are similar and the differences, especially the Libre's tendency to over-report at lower levels is due to the different algorithms. Abbott have said that the Libre will underestimate BG on purpose, as it adds an element of safety. Unfortunately, that causes me to hit 2.9 on sprints while Dexcom is still at 4.5 and glucometer at 5.5.

pre Libre we were not on board with what our levels were constantly and micro managing them. We did blood tests, but not constantly. And we accepted what they told us.

That's a great reminder - we're having the discussion on real time BG measurements and algorithm errors because of the technology. Some years ago I would have sprinted, relied on a few glucometer tests and my hypo awareness. Way less risky now.

 

[SimonP78]

The technologies are similar and the differences, especially the Libre's tendency to over-report at lower levels is due to the different algorithms. Abbott have said that the Libre will underestimate BG on purpose, as it adds an element of safety. Unfortunately, that causes me to hit 2.9 on sprints while Dexcom is still at 4.5 and glucometer at 5.5.

I've also heard this about the libre erring on the low side for safety. Do you know where this was reported? It would be interesting to ascertain just what in their algorithm they are doing to err on the low side as I am over the last 4 or 5 sensors invariably shown as being at least 3 mmol/l lower than finger prick readings across the board.

I assume it's probably just a case of a look-up table which errs on the low side when low, but you never know, perhaps there's something time-domain they do with the sensor data which could be adjusted for those of us who suffer from offsets across the range.

I feel I may be grasping at straws in my frustration though!

 

[Hobo1960]

Perhaps some of us have a lag greater than 15 minutes between blood glucose in capillaries and interstitial fluid. An algorithm based on 15 minutes lag (if there is an algorithm) could be inaccurate for people with a greater or smaller lag. You may have seen this video before.

 

[Jasmin2000]

Perhaps some of us have a lag greater than 15 minutes between blood glucose in capillaries and interstitial fluid. An algorithm based on 15 minutes lag (if there is an algorithm) could be inaccurate for people with a greater or smaller lag. You may have seen this video before.

We all have a real lag that is about 5-20 mins (not just different folk, but for everyone at different times and activities). This lag is "compensated" by the predictive algorithm to give an approximation of BG. It takes the trend in the last couple of minutes to predict the next reading - leaving a lag of about 2-3 mins. If there is an obvious difference between real BG and sensor glucose, the algorithm can partly correct it, and you can watch the line change on your chart as it corrects the value. So the 5-20 minute lag is replaced by a 2-3 min lag with a corrective feature.

What is still missing for me is the correction also limits itself to a small part of the original prediction and if the difference between real BG and sensor glucose was off by more than 1 mmol/L it cannot correct it and your curve will be wrong. This is what I see when doing exercising and recovery - BG and sensor are off by 2-3 mmol/L for a while.