Continuous Glucose Monitoring with Medtronic/MiniMed

UPDATE: THIS THREAD IS CLOSED AND NOT CURRENT. For an updated view of my experience with this product, please go to my more recent post, from 12/5/2007, on the same topic, just click this link.

First I should state that I have not yet been using the continuous glucose metering from Medtronic/MiniMed for long, less than one month. The learning curve appears to be quite large, even for someone with a strong technology background and competence with other electronic devices, even for someone already in tight glucose control.

I would also like to state that even at these early stages and with the lengthy laundry list of issues I’ve had, I have also gotten some very good information from the continuous monitoring, even at this early stage. I have already adjusted my basal rate for a much more even reading throughout the night. I have already learned that I was under-bolusing, especially for breakfast and lunch. I have also had times when the alerts have told me my blood sugar was either high or low that I would otherwise have remained unaware of.

These hint at great things to come as I learn more about the proper use of this device. I would hope that this page will serve to help others set realistic expectations for what they can get from this device and shorten their learning curves to avoid some of the pitfalls I have made. As I get better with this device, I expect to post more threads with pointers for improving results. I will be more likely to do so if others respond on this thread to let me know that it is being read.

These are the Issues I have had so far with Continuous Glucose Monitoring:

  • The transmitter does not work well in a modern office. I had mine less than an index finger’s length from the pump with my cell phone on the opposite side, and even tried turning it off, and still got lost sensor. Weak signal and even lost sensor are common in my office, with wifi, cell phones, cordless headsets, blue tooth, microwave ovens, and other radio transmissions all around. I really believe that the solution is to have a wire connection option. Wireless is simply not yet reliable enough for medical devices, especially at the chosen frequency of 900 MHz.
  • The tape on the sensor needs to be more like the tape on a silhouette infusion set. The front of the sensor, where the tape does not even continue all the way around, is the most critical area since the transmitter pulls away from that end.
  • The relationship between interstitial glucose and blood glucose requires better research and documentation. My experience so far hints at a couple of interesting possible relationships.
    • As is well documented, the rise in interstitial glucose lags the rise in blood glucose, though the time seems to vary.
    • During exercise, I am starting to notice that the interstitial glucose may drop before the blood glucose, which may or may not drop as well. This seems to be the case when I do not have an unabsorbed bolus in me. If I have insulin not yet absorbed, that absorbs rapidly with exercise and drops blood sugar first.
    • If blood glucose is rising and action (i.e. a bolus) is taken, the interstitial glucose will likely never reach the level of the blood glucose.
    • At times, especially during extremes of inactivity such as car rides and watching movies, rises in blood glucose may never trigger rises in interstitial glucose.
  • Behavior of the pump software on lost sensor produces actively worse readings. One problem with lost sensor is that the behavior of “find lost sensor” does not simply reconnect the transmitter and pump using the old calibration. As noted above, getting a time when both blood glucose and interstitial glucose are steady is difficult. When a sensor that was well calibrated is lost, the new reading, required in two hours, is likely to produce radically worse calibration than the nice early morning flat calibration it replaces. Often, this means that the device is useless for the rest of the day.
  • Sensor is often lost after far fewer than 40 minutes. Sensor is treated as lost if transmitter is removed to check the battery. Both of these cause bad calibrations afterward, as noted above.
  • I need to verify this one a bit better. However, my suspicion at present is that alerts generate either vibrate or beep alerts depending on the setting of the alert type at the time that the alert was set up. So, in the case of a meter BG required, it may beep when the pump was set to vibrate, because, at the time that the sensor was attached, the pump was set to beep. This is, of course, a very minor issue.


  • In my experience, calibration is more accurate whenever both your interstitial glucose and blood glucose have been fairly constant for at least a half hour. Whenever this is the case, enter your meter blood glucose to avoid having to enter a less good calibration reading at a later time.
  • The best calibration reading of the day is typically just before breakfast for me, always always always take the opportunity to use this calibration.
  • I have my alerts set at 60 for low and 140 for high. Unless you want to know about every fluctuation, you probably want to stick with the defaults. However, if you are not getting enough alerts, consider moving the numbers closer together for tighter control.

19 Responses to Continuous Glucose Monitoring with Medtronic/MiniMed

  1. Misanthropic Scott says:

    Update: Despite all of the hassles, this really is improving my control significantly. If it helps to keep me from going blind, this is a huge win. Even by the time I wrote the main portion of this topic, I had already adjusted my nighttime basal rate significantly and achieved a much flatter graph through the midnight hours. This is huge.

    I also have a new tip. An hour before bolusing for each meal, check your graph. If it has been flat, take advantage of getting a good calibration reading in there. Do a reading, even if it is comparable to the sensor, get the calibration in, your graph may not be as flat next time you need to calibrate. This pushes it back another 12 hours.

    If the two readings were different, check again in half an hour. If the sensor reading has not moved toward your blood glucose reading, do another meter reading to ensure that you know which is right. If it’s still off and if your new meter reading is close to the old, calibrate again. Repeat just before you bolus for your meal.

    I have found that it often takes 3 or sometimes even 4 calibration readings to get the damn thing to acknowledge that its reading is the one that’s wrong. Of course, I may be overly anal in trying to get the two devices to agree on a number. To me though, the values are as important as the trend, so I like to get them right.

    You may have guessed it by now. I am doing far more readings than I used to. I am getting good information through it all though. It’s just frustrating … but worth it.

    I know this page has been getting read. Please feel free to ask me any questions about any of this. If you want to remain anonymous, just get a yahoo or gmail ID that does not have your name on it and use an alias for posting.

  2. Misanthropic Scott says:

    For more a more general discussion of tight control of type I diabetes, please see my more recent post on the subject.

  3. jane says:

    My daughter has been using the real time for about a onth. We are having issues with calibration errors. A LOT. It is driving me crazy. I can not seem to get more than 3 days use out of a sensor. When we try to “start new sensor” while leaving the old one in, we get cal errors. Also, our “i sig” is usually around 10..Minimed says this is the issue, but how do I change that??

  4. Misanthropic Scott says:

    You can view the isig on the paradigm 522 pump by pressing ESC four times. An isig ofd 10 sounds OK to me, assuming you mean 10.00 rather than 0.10. One thing your daughter can try is massaging the area around the sensor to get new fluid there. That has helped me considerably since I started doing so last week.

    I’ve now had this for about 2 months. The learning curve is huge. I’m getting really good information from it when I can get it to work. However, I will say that even though this is really helping me to adjust basal rates and time my insulin, I am NOT finding it reliable as a mechanism for providing much immediate help.

    Once in a while, an alarm will tell me I’m either low or high when I didn’t know that before. Mostly though, I find that I can find out how to treat myself better overall rather than right this minute. Certainly, IMHO, tell your daughter that she should NEVER take insulin based on the sensor reading. At best, it is useful for telling you that you need to test.

    As for CAL ERRORs, I have complained about that as well. I usually get them when my blood sugar is moving rapidly. If your daughter follows my instructions from my own reply on 9/27 for always calibrating whenever her blood sugar is stable, she may be able to avoid having to calibrate at a time when her blood sugar is moving. This is the best I can recommend for CAL ERROR. I guess, checking the isig to make sure it is in a good range first might also help.

    All in all, this has been among the most frustrating devices I’ve ever used. I’m continuing to use it because it really is helping. It’s just hard. Her health is worth it. I hope she can stick with the stupid thing and get some good results out of it. It’s well worth it. However, I can’t say yet how long it will take before I’m really comfortable using it on a regular basis. And, I’m a geek (computer programmer). So, I’m generally pretty comfortable around technology. Perhaps just knowing that it’s not just her will make this easier.

    Good luck with this. Feel free to continue to ask my advice. I’m still learning too. I’ll keep updating as I learn. Remember, tight control is key to diabetes management and avoiding complications. If this can help, even if it’s hard being an early adopter of a new technology, it is worth it. (OK, now, am I trying to convince you and your daughter or me?)

    Oh, I just remembered. This morning, I did something right!! I hope I can repeat it whenever my sensor is 3 days old and I want to continue using it. It had not yet gotten to SENSOR END, being 2 days and 19 hours old. I calibrated it. Then I remembered it was old and that 2 hours before breakfast was a good time to do new sensor. So, I did. I did Sensor->Start->New Sensor. And it started picking up new readings within 10 minutes or so. It did not make me wait 2 hours!! I hope this is repeatable as it was one of the few times when I had it accurately calibrated and was glad not to lose that.

  5. Misanthropic Scott says:

    One more thing. For anyone thinking that this will help you test less. No. At least it isn’t doing so for me. I had been testing 4-12 times a day usually about 6 on routine days and 8-12 on weekends when my schedule is erratic. Now I test probably 10-15 times or so every day.

  6. jane says:

    thanks for the encouragement. My daughter is actually testing less. Her normal has been 8-12 daily and more on the weekend. We are averaging about 4-6. When the sensor is working it is right on and we are very happy with it. It has just been cal errors that have sent me over the edge. Tonight will be the end of day three; we will ahve to see, I am going to not remove the transmitter, just keep it attached and not recharge it, and “start” a new sensor. The first 6 days that we were on the sensor, she was in range (70-140), 87% of the time. Hormones have kicked in the last few weeks and it is not as good, but I do have faith in the ability of this device. I know, as a nurse, that this can and will help her in the long run and help bring her A1C lower than it’s current 7.4. She actually has had pretty tight control in the almost 7 years of being diagnosed. She has never had an A1C above 7.6. (I am knocking on wood, as puberty is upon us)

  7. Misanthropic Scott says:

    Well, I’ll be curious whether calibrating whenever the graph shows a nice flat period can help with the calibration errors. This should give better calibrations and always put them off for the next time that the graph is flat instead of being forced to calibrate when blood and interstitial fluid readings are in flux.

    It may help if she checks her isig before calibration, especially if the numbers are wildly different. I sometimes see differences like the sensor saying 52 and my meter saying 114, even though the graph was flat. I haven’t seen that since I learned about massaging the site and its possible effect on isigs. I’ll let you know if there is any correlation between wildly differing numbers and a low isig value. I expect this to be the case, but can’t say yet.

    I also tried to convey to the engineers that if the graph is flat for at least an hour and blood sugar reading is entered, the blood sugar should be taken as golden and just used immediately. This should never cause a calibration error, no matter how far apart the readings are.

  8. Larry says:

    Hi – I use a paradigm 512 and it’s good to hear that continuous glucose monitoring is somewhat effective. I chose not to use it, as my insurance didn’t cover it and I feel like a lab rat sometimes with the infusion sets sticking out of me. I guess the best part of the pump is people asking me why I have ‘pager’. Always brings a chuckle. I then usually say I am a doctor.

    Anyhow, thanks for your input on your pump usage!-Larry

  9. Misanthropic Scott says:

    I think I finally have this technology down now. The final pieces for me were to combine massaging the sensor site and entering calibration values only when my graph is flat for at least a half hour, preferably longer. I am now beginning to test less, sometimes as often as I used to, sometimes even less than that.

    I have not had a CAL ERROR in over a week. I did notice yesterday that my isig values were getting lower than they had been for the prior 5 days. I think I will likely not keep a sensor in place the full 6 days, but change it after only 5 instead.

    Please feel free to keep asking questions. I’ll keep doing my best to answer, whatever that means.

  10. jane says:

    well, I think I solved my daughters cal error issue when starting a new sensor. I need to place the sensor, cover it with tegaderm, and leave it for hours, usually overnight, without connecting the transmitter. I just left the transmitter on the current sensor and inserted a new one, so the poor kid had 2 sensors and an infusion set in. She tends to be “dry” and it taked the fluid a while to get up the cannula. By doing this, her new sensor started right up, no cal error.

  11. jane says:

    oh yeah, went to the endo last week, her A1c is 6.9, down from 7.2 yeah!!!!!!

  12. Misanthropic Scott says:


    That’s wonderful news! I’m sorry I wasn’t able to help with this issue. I overlap mine too, but not by nearly as many hours. Either way, anything that helps the A1C is a really great thing, even if it is a pain in the abs/butt/thighs/whatever.

  13. Marnie Franze says:

    I have had – A1C all over the place in the past 30 years, most of them good. Became a diabetic in November 19, 1976. Started the 522 pump last year, love it, Just started the CGM last week, boy is it frustrating, got two warnings sensor error and lost sensor, working through them, learning as I go. But even after a week of using the CGM, I would never go back – bring on the frustration. I have been well educated in diabetes for 28 years, fought about trying the pump for years, after 2 weeks I would never be a non pumper. No complications or hints of complications to come, even A1C not exactly perfect all the time. But the pump was easy to learn, the monitor well looked easy, but it has a hidden learning curve….

    I talked with my nurse, she has one patient that refuses to take insulin, older patient running around 200 to 350 all the time. The complications with diabetes might be coming from uneducated individuals who refuse to take any care of themselves. I have a neighbor that refused to pay any attention to his diabetes for 3 years, no blood tests, no diet, no education. Now he has eyesight issues, and was hospitalized. And a sister in laws mom that has had diabetes for 15 years not taking a blood tests ever, always ignoring it and I mean always ignoring it, now has stage 3 kidney failure. My sister in law said she would drink diet coke constantly, which was to get rid of the diabetic dying of thurst you get with high blood sugars. Then my husbands cousin was always so thin, and never took care of it either for about 8 years. He is doing great now, good doctor and great control. Now these are hopefully the individuals who make up the numbers for complications, gives us hope that the care and time we have put into learning about diabetes and learning all the new technology gives us life….

    you made a comment about tight control….so I am sending this…there is no option, you do it, sometimes you get lost, but there is no other options….do not be a diabetic that is included in the type that get put on the complication list

    All insurance companies will eventually care it…I waited 1 year for my insurance company to cover the CGM

  14. Marnie Franze says:

    Me again…what does Sensor Isig stand for….? You mention not to have it at 0. I looked all over medtronic site and my books, can not find it. okay blast it now my sensor has a week signal with no Isig. Must be a day of just error messages. I am going to call medtronic help line….

  15. Misanthropic Scott says:

    Hi Marnie,

    Sensor ISIG stands for Interstititial SIGnal. I strongly recommend reading my later post on this topic and moving any further discussion there. I think that will help a lot. This was my initial reaction to using the sensor. My updated reaction includes several months of additional learning.

    For updated post, click here.

  16. james says:

    Hey Scott – thanks for your posts and discussion. I’m also new to the Guardian with my 522 pump –

    After almost two I’m just now become confident in my readings – but it’s based on my knowledge of when to expect accurate vs inaccurate results. I’ve tried various places on my body, but my readings still tend to lag more than I want. But even based on that I still love using it – and with that knowledge I still know my trends (but I have to be careful not to overcorrect based on that discrepency. )

    And yes you definitely have to check MORE. I’ve been checking more WITH the CGM the past two months to obtain my unscientific results of when it works best. I honestly think everyone is a bit different – thus everyone requires they’re own learning curve.

    Now I’m finally feeling like a pro with it – and feel like I OWN the sensors and readings rather than they owning me. (Like people overwhelmed at first with errors and cal issues. Yesterday (after forgetting to take the tape off the sensor,) I even pulled it out, rethreaded the sensor in the needle, and re-inserted it. No errors 🙂

    I think this technology will only get better and better – but we’re still the first wave of CGM users. But even with the dicreps it IS great to have for activities in which it’s awkward or inconvenient to test. (i.e. in movies, out drinking with friends, etc etc.)

    I detailed some of my thoughts on my blog too if you are interested

  17. james says:

    Sorry I meant “after two months”

  18. Misanthropic Scott says:

    Thanks James, please be sure to check out my more current post though at the following link. I left this post for reference about my initial experience only. I have learned a lot in the 6 or 7 months I have been using it now.

    I don’t know how to make it clearer to readers of this post that this is no longer current.

  19. james says:

    Yeah I did see your other post, of course after I commented. You may want to close the comments to redirect…

    Good info again — All the best

    [ed. Thanks. I had trouble figuring out how. Turns out to be easy once you find it. Thread is now closed.]

%d bloggers like this: