New user, low alarm level and speed of tracking

I have had mine 3 weeks and support from dexcom has been terrible. They seem to be treating me like
Vlad the Impaler and have not responded to my calls and concerns in any way and nor in 3 weeks.

Accuracy is not issue, speed of response when at low end is like slug in artic and misses key events.
My body is large and can really swing glucose quickly on gut emptying - running out.

Low alarm blocked at 100 is useless coupled with slow tracking response really hampers device.
I am back in with handheld checking myself when I get down to 150 and many times actual BG is 130 to 140. If I do not get glucose tabs in early, my system can be at 50 no sweat before DEXCOM has twigged anything is wrong and my liver does nasty dump to 311 or higher.

This is unfortunate as this unit has some nice features and done some things right. AT higher numbers 180 and up delay of response is tolerable. Their field rep referred to me as a dousie I believe.

Response seems they just wish I would go away and leave them in peace.

Jim Snell 805-383-7759

hmmm. THEY can’t say this (FDA rules), so it’s good that you asked here.

You’re offering hints like “my body is large”, and you’re a man. If your frontal abs/belly area has “enjoyed” many years of shots and infusion sets, or your waste size is larger than 34 inches, then you might want to try an alternate site. I’ve found that “love handles”, directly down from the armpit on either side, work best for me. Lots of women in really good shape, with small waists, get better results with an upper butt cheek – but it’s inconvenient. Some people use upper arms (an official location of Abbott Navigator, RIP, but totally “Off-Label” for Dexcom). Try the love handle spots first, about two inches above your belt (don’t get too close to your ribs).

In any case, it sounds like you’ve spent nearly all of your “30-day money-back trial” days already. If you plan to give up and return it for credit, DO NOT keep it past the time limit – even though you’ve been waiting on THEM to call you back, they’re very firm about the maximum length of that “money-back” period.

I’d be inclined to try a bunch of other sites, but only after telling them that you’re going to return it (immediately) unless they send a letter (via fax) which officially extends your trial date for that testing.

I hope it works out for you. But If it doesn’t, it’s almost certainly NOT YOUR FAULT! The two CGMS machines, MM and Dexcom, have a higher likelihood of “YMMV, and it’s behaving as a random number generator” than any other treatment or diagnostic device.

I think we need time for these CGM to get more accurate before relying on them much. The MM was horrible in terms of accuraqcy and discomfort but the tech support was good. Dex seems a little better for me as some days its on, but often its horriblby off. Tech support is useless. Hope you can be positive and do the best you can until better products arrive.

Good luck.

Thank you most kindly for response. Comment about my size has to do with liver size and response. When I trip liver sub 77 - sually - 50, and if metformin does not have liver by neck ( up to strength in blood stream) I can watch on hand held taking many samples and watch meter ramp up to 511 - HI and then ramp down back to 278 as heart mixes the crap pumping around my system. Male/female is not at least as I know not issue…

like i said - response time of cgm is slow and accuracy good when body moves slowly not issue. I do not see dexcom tracking crap sugars in my blood like the pqq accuchek.

I do not know the Dexcom you are describing. Every time I call Dex, I get an informed, caring, and responsive individual. I even called on a holiday weekend when my sensor went crazy. The on-call person was at home with house guests. I heard the person excuse themself from their guests, go to the privacy of a home office, and provide detailed information; saving my sensor.

Keep in mind, all of the electronics discussed here are TOOLS ! We have to learn our tools and use them well. Trying to ask a hammer to hit a nail every time is not the venue of the hammer, it is the venue of the carpenter swinging the hammer. Many people in this group and others expect a CGM and BG meter to be “dead on” 100% of the time. If you believe this, you need to RTM again.

The area of concern I see from your comment “My body is large and can really swing glucose quickly on gut emptying - running out.” is there is a reason you are dropping like a rock from 30,000 feet. Instead of blaming the CGM, look into the reason you drop like a rock. For example, is there a reason all of your insulin hits at once? Do all of your carbs get into the blood stream at once? How long had it been since you ate before this crash? – IMPORTANT POINT – How closely are you working with the physician who prescribed your Dexcom and manages DM for you?

A spread between CGM and finger stick of 150CGM and 130 BG is not really bad. The CGM can be up to 20% off - it is the curve, not the real number. Next the BG can be 10% off. So, your real lab measured glucose could have been 140 and both tools were correct.

The information you are sharing leaves many questions unanswered. Instead of damming the darkness, you need to turn on your light.

Unfortunately, this response is off base and not to issues raised. After spending $ 1500 on a new product I too would have hoped for the relationship you mention. Speedy Muffler has better manners. The silence has been complete in last 3 weeks. I do not see what your discussion attempting to co relate speed of monitoring versus accuracy. As I indicated, I felt the accuracy was good but at key times was missing data in its slow response. Trying to explain that under accuracy numbers is while interesting not the key point. Blaming my body for not playing straight or not well behaved against the speed of capability of the data logging machine is a curious form or arguing. Do you work for Dexcom?

For medical reasons - my Doctor, I indicated I needed a low level alarm settable above 100 so I could get an earlier alarm at 140. I was led to believe others had asked for just such capability. I had also indicated this would help compensating out the slow response, Alarming at the point, my doctor does not want me going under is useless - too late at that point.

What is being really said here? I offended the shiboleths of the Dexcom Gods so therefore I can freeze and sit in the dark without contact, support or response and not feel glad I selected and purchased a Dexcom unit.

I do not work for Dexcom.

What I am attempting to share relates to what I read in your first message. My perception gleaned from your first message was you and maybe even your Dr. had expectations inconsistent with the limitations of the tools (speed of response from CGM, Insulin, and correction carbs) at hand today.

My apology.


Thank you for response and apology not needed.

As you clearly and quite accurately pointed out there are many underlying issues and frustrations with diabetes and the human body such that one does not want to streak off like a errant heak seeking missle after an issue irregardless if one thinks oneself assessment correct on surface issues.

My expectations and I would say “needs” have been formed watching and logging this monster with an average according to my data logging program of 32 to 38 finger skin pricks in a 24 hour period and watching my body and catching the race to the basement while adjusting diet and snacks and volume.

Does my diet and my goals accerbate this process. Quite possibly. I am pushing hard on a 1200 to 1500 calorie diet, I am on a mixed hybrid bag of insulin plus pills - metformin, starlix. Insulin at morning and night to compensate out slow startup of my body and to close out lower on BG at late night.

I am not on insulin pump with far better tracking of insulin requirements versus carbs and instead attemptring to ensure sufficient food /carbs against a fixed (sort of) insulin generation by a starlix pill.

On the otherhand, looking at FDA website, I clearly see that CGM technology as you and other posts have pointed out is in its infancy. It reminds me of the early days of 10/100 and 10/100/1gig ethernet meetings I sat in when the industry, interested players, mfgs and regulatory parties all sat in a large room and looked at each other and said “what should we do?” and came up with an initial spec.

I apologize for my fast - qick comments that were not intended to be negative but being pressured by my need to change how I was monitoring and controlling my situation and beating the crap out of my personnal situation.

Thank you.

Jims- Just a suggestion but maybe you and your Doc could figure a way to simplify your regimin. When your glucose takes a dive can you go back and determine the reason for the dive in your glucose? I think it would be terribly difficult with the brief mention of what your regimin is. It sounds like you are taking Lantus/Levemir twice daily with Starlix at meals? Perhaps it may be easiest to drop the Starlix and do a little basal testing. Take your Lantus/Levemir like normal and do not eat for as long as possible. Test like you normally do (32 times a day GEEZ! Your fingers must look like a pin cushion). This way you can see whether the basal may be contributing to your lows (if you do not eat that Lantus/Levemir should keep your glucose stable). I would also suggest you consider taking rapid acting insulin at meals instead of Starlix. This may be a bit easier to “fine tune” than Starlix (I know more shots but with what you are going through the fix, if it works, would be preferable to the problem). Not sure if you are doing it or not but I cannot downplay the importance of really counting your carbs. It may even be helpful to read up on the Bernstein diet and give that a try (by cutting out the carbs you can decrease the possible variables). Keep working at it and I hope you find the answer soon.

Jim- I had assumed T1 diabetes, as most of the users here are people with that illness. With T2 and possibly “metabolic syndrome”, the illness which you live with is a lot different than mine – and YMMV, definitely.

T2 is a complex disease- more complex than mine (although I sometimes have T2-like symptoms too; it’s not that uncommon among older T1 PWDs.) So, I’m going to take a wild guess that the support people are also frustrated and unfamiliar with possible causes and solutions of your Dexcom failure-to-perform … and in frustration with not having a clear, concise answer for you, they possibly DO have snarky comments about “wasting their time” after they hang up the phone. And some of those feelings are leaking into their speech with you, and their failure to call back. It’s a bad thing- NOT professional, and harmful to the company but it does happen in a lot of support organizations. :frowning:

You might ask your MD whether s/he thinks a low-carb diet (and I mean RADICAL low-carb, less than 10% of calories) might smooth things out? Or, you might research yourself: Google sets of words like “hba1c low-carb study results” and you might find things which you want to try. Is it too late to send the Dex back?


All excellent feedback and astute feedback. Some more details.

  1. The fast drops on glucose are due to insulin pressure (extra) during day. At early morning
    and late at night prior to any lantus - BG will sit very stable at whatever number I leave it at and as long as liver nailed down by metformin, BG will slowly drop on tiny amounts my pancreas put out in absence of humolog/startlix/lantos.
  2. Reason for nasty dawn effect (238) and liver emergency at 278-311 is due to natural basil insulin is small and only ramps up on starlix type - glyburide pills.
  3. I take 21 units of 75/25 humolog in am and ride till lunch and in fact it is the smoothest part
    of my BG charts. I just have to watch some hot spots of the slower parts of Humolog at 10:00 am and near lunch. I take at 5:10 am and usually out of body by noon. No starlix in am at all. That did cause riot then.
  4. I try to count carbs as best I can then back in balance of diet.
  5. first starlix I take at lunchtime synced to eating lunch. I try for snack at 1:15 to 1:30 am. My digest time is 2 to 2.5 hours - slower and faster based on carbs type, fat and glycemic type. I usually ensure of Have some high glycemic content to ensure starlix digestion of 20 min initial up in blood and 1.5 max to full strength does not outpace gut sugar risetime out.
    6 second starlix at dinner usually 5:00 pm synced to dinner eating.
    7 i ride this till digestion complete and starlix exhaust.
    8 at 10:00 pm on , i look at bg and hit with lantus shots - usually 2 of about 7 to 10 units.
    I try to organize last lantus shot at 12:00 am or as close as possible and I usually find that BG will flatline to 120- 128 till I wake up at 5:00am at which point I watch my stinking liver banging up the BG with glucose in am - no food or sgars or carbs at all. As I take metformin
    at 5:00am , with 21 nits of 75/25 humolog by 8:30 am liver gets caged and stops that. I can stop that early by having good alcohol drink and roll it over early at 148 to 150.
    Brteakfast at 7:30.

I never wanted to be checking on my BG as much as I found myself but it really is liver adding all the fun. Metfomin at 10:00pm and 12:00am -\500 mg locks down dawn effect.
Metering allows me to watch zones where gut is coming up/down and whether I need to jump in and add glucose tabs to block sub 100. Doctor said no sub 100 please.Ie my liver buffering does not work with a crap and sugers me up.

WHat this means folks that prior to catching my liver’s nonsense on dawn effect and buffer control, my H1ac was 13.3 %, eyes had hemorriging, on actos, body leaking water, overweight etc. After metformin change at midnight and playing liver buffer with glucose tablets, H1ac was 6.9, eyes healed, actos booted, lwgs and water gone and my weight started dropping after metformin change.

Metformin in blood up to strenth as best as I can tell shuts off the crtc switch in liver from make sugar to fast mode. Salk institue published this fact. I routinely can tell when liver caged and not pumping in extra sugar. I aslo take 500mg charge at 7:00 pm of standard metformin ( 2.5 hours in, 2 hours up in body and exits in 1/2 hour. Monster large charges are waste of medicine and the 500mg charges work fine.

Yes, I am type 2 but with a need to play type one watching my BG does not trip off liver to do emergency sugar add.

Previously to this nonsense, I was walking 2 hours a day routinely to burn morning sugar \off that was so bad cells could not take any more glucose. Actos will ram more in.
Some would call that insulin resistance. I called that cells saturated with glucose from liver. Once I walked off the excees glucose, I would see humolog work and finally get BG to 100 by lunch. Rest of day was normal.


Regarding customer service at DEXCOM and some other fine folks, this is the trouble with the complexity of this disease and well meaning non diabetics trying to work with crib sheets looking for the magic silver bullit one answer to try and help - not.

Prior to design, I was involved in DEC field service when they where still in business and a pharmacist whom we put on line with computers told me - I realize you and the company cannot fix all problems but in the Spirit of the Japanese Iron chef , I will always be pleased as long as you do your best to help me.

Just a couple last suggestions. Since Lantus is mostly peakless (or it does not peak enough at least in your circumstance) perhaps by switching the Lantus shot to NPH alone you might be able to time the dose just right so that it peaks at about the same time your liver kicks into overdrive. Obviously you risk nightime lows with this strategy.

If you have the insurance and the money you may also consider an insulin pump that could be programmed to give you more or less insulin at your problem times. Not sure where you are from but the US Medicare guidelines (which a lot of insurance companies mirror their policies on) is that you have to have significantly diminished pancreatic function to get one- not sure if you would qualify if Starlix is still effective for you (but you never know).

I hope you have more luck in the future with the Dexcom customer service and tech assistance people. Keep on pushing as you and/or your insurance company paid good money for the system and they should exhaust all possible solutions to get you results acceptable to you or admit to you perhaps their system just does not work for you (and hopefully offer compensation).


Thank you very kindly for your time and patience. All comments have been germaine, helpful and suggestive as well. It remains a work in progress and your help has cheered me up.

Doctor and others has suggested insulin pump. It is both frustrating and annoying about pancreas sleeping off with tiny basil insulin secretion yet on glyburide - starlix kicks it into gear with sufficient insulin or at least to process meal.

It is curious the tortuous path walked to find myself at this place of fun.
Humolog was picked in AM to boot back liver’s early morning hoot - even on metformin being on liver all night, as soon as it wears off my brain and liver go back to work in am till clamped off either early by booze or finally by new metformin coming on at 7:30 am at earliest. I have watched in am that BG will be flat line and minutes after wake up, there goes BG in action from liver. Brain and liver are co-sonspirators in this nifty riot.

Humolog was added due to metformin 2.5 hour ingest time in am that was originally screwing me up. Lantos was added at night to pull back BG at night due to early wear off of starlix. I generally run evening BG down to 120 so as to give head room against evening brownout and that has worked.

It seems I cope better with the humolog peak issues and really have to work the gut/starlix insulin glucose load matchup mess. Once up - no problem, once down and over at night I watch things slowly decay at 9:00pm on without a scramble.

SUffice to say this has been worse and more challenging than organizing state machines in a complex interactive digital processing system.

Once again, many special thanks for your time, comments and suggestions.


Special thanks for reply and your assessments and observations. I wanted to resolve comments to prior responder first.

I am type 2 - no doubt.

I am doing three things here:
a) counting and ensuring sufficient carbs.
b) following Mediteranean diet - low glycemic for all else. Carb count and amount essential as I do not
want to trip liver to do emergency glucose add at 50 and swamp my system.
c) I am frustrated with much of the diets and literature that does not deal with body and organ size. This is not joke. In order to get low a1c numbers and keep from popping the glucose safety’s I have to run minimum fire just to blow whistle and run turbine and small amounts of movement. The minute I have to do serious exercise, I have to ramp fire up to 180 or more so I do not flatten BG in 10 minutes.
d) I am on min diet of 1200 to 1500 calories and have been from Dec 2007. I could not lose weight till May 2010 when the metformin and testing approach was implemented and mentioned in earlier post.

My frustration with Dexcom is what is your response time - no answer.
OK I ask for higher low level alarm so that I can get earlier alarm and get in with handheld to check reliably.

Currently, I routinely see my body step down 20 to 30 points faster than dexcom can see. This is not accuracy issue, it is speed of response. To compensate I watch unit’s numbers and as it is heading to 160, I check on handheld and if at 140 and 130 down; I start adding glucose tablets to cut off fall. As I said, I am not 4 foot 80 pound race horse jockey with small glucose system that much of the industry uses as its model and response times on glucose swings.

a 7000 horsepower challenger/1221 norfolk western articulated locomotive does not work like a little yard switcher engine and requires different subtles to be coped with.

Thank you for writing and sharing your comments - all helpful.

Jim, about the delay in response:

THAT part happens to everyone. What most of us do is watch it like a hawk, and when we see slope of the curve change in a way which we don’t like, and didn’t expect-- treat early.

Don’t just treat the number: treat the trend.

Well said Rickst29 about treat the trend. Really, it should be treat the patient not the number.

Another point hit me a few minutes ago – I want a “carved in stone” formula for everything - breaking distances, insulin carb ratios, correction factors, you name it. The CDE in my Endo’s office told me treating diabetes and WILD swings of BG is NOT A SCIENCE! It is a series of educated guesses or SWAGs. So, I guess the gang here, including me, just go SWAGing around. <<GRIN>>

I agree with the other posters regarding the trends and the delay between fingerstick and interstitial happens to everybody. That’s just physiology and not Dex messing up or being inaccurate. Dexcom claims a 5 minute lag compared to medtronics 15 to 20 minute lag between fingerstick and CGMS readings but I’ve found personally that the lag varies somewhat. Also it’s not so much that the CGMS values will match the bg in 5 minutes but it will reflect the trend. Hopefully that answers your question about the expected timing. For me the CGMS is a lot slower to come up after a low than to change with the highs also. I’ve also found there’s a lot of variability in fingersticks too so I don’t think of any of this as an absolute. Also Dexcom tends to get more accurate for me the longer I use it. You might review a lot of the calibration tips people have posted too.

I too am type 2 and tend to have a strong dawn phenomenon and will drift up if I skip meals especially in the morning. Not as bad as what you seem to experience though. We’ve (endo and I) have tried various things and right now best solution for past 9 years has been insulin pump. I was on metformin but then I forgot to refill and low and behold it wasn’t doing anything so we stopped it. I also recently am trying Victoza.

Have you been using the event markers on the Dexcom? You may find the downloads to be helpful in terms of figuring out patterns between rises/falls and timing and amount of carbs eaten. Make sure you download and get all of your data if you send the Dexcom back.

One last comment. I wish they’d give people longer than 30 days before deciding whether to keep it or not. When I first started with Minimed in 2007 it took me several months to figure out CGMS and how to use it/troubleshoot.


Thank you for responding and offering your experiences. You have talked about many issues as well as CGM so I will try to respond:

CGM Dexcom comments:

  1. Yesterday , Dexcom and I got together on phone finally and had good discssion.
  2. I am not throwing rocks at them while I had some issues - we all do.
  3. Interstitial delays are a given and on FDA website on early spec issues, numbers were bandied about being 3 minutes to 10 minutes. No argument - as indicated. This was not the argment either.
  4. My sense on accracy were stated but not possibly as clear as I needed to be. Given settling time, I consistently find my accucheck and freestyle track well with Dexcom. I have no issue with accuracy.
  5. Delay is what it is, but when some body limits low alarm trip to arbitrary 100 instead of letting me set it higher to trip earlier so I can get in with handheld without having to have my eyes on the dexcom constantly to see visual number. My liver crashes occur very fast after gut emptying and I need all the spare time I can have to get glucose tablets on board to stop. I had crash yesterday where dexcom went down to 148 min and my two handhelds and tracking got down to 119 and doddled arond 124, 128, 133 and finally lifted to 162. Dexcom only went down to 148 and then backup all in 15 minutes.

The difference in numbers from handheld and dexcom for me actually tell me if I have had fast drop in glucose prestage a possible liver dump. At all other times tracking is excellent that DEXCOM cannot follow or guess and estimate at interstitial level.

My goal was to uncover data/specs not specified in manals, tech service.
6. Dexcom publishes no spec on this response item and that may be ok. My goal was to uncover what they think their response time was and b) allow me some flexibility on low level alarm set and c) commnicate so we can resolve problems which has now been done.


I/my doctor have excellent sucess with metformin at 500mg doses. Dawn effect was 238 to 258 reliably in am. My doctor and I have tested this many times and note:
one large dose of metformin will not stop dawn effect.
one 500mg dose at 10:00pm and 12:00am midnight reliably cut this off till 5:30 am when pill wears off. Based on 2.5 hours ingest, 2 hours up and 1/2 hour exhaust on my body.
Most generixc metformin work well - have had no sucess on ER, Teva and single large doses - waste of medicine.
I can see on the Dexcom the effect of metformin up full strength on the blood on liver and sugar addition stops.
Many think there is lasting effect of metformin and I only see it do its good work when dose up full strength in blood on liver (500 mg - less than does not work)
a1c was 13.3 before adding metformin as indicated and it dropped to 6.9.
Also, I find that when glucose drops sub 70 and forces a liver glucose add, I get 278 to 311 average. It actually peaks up to 511 or hi. When metformin up in blood, liver hard pressed to hit me hard.

My worst dumps always occur when liver is foot free and fancy loose off of metformin.

See Salk institte over the CRCT2 switch in liver and action of metformin. Type 2 diabetics have their liver jammed in make sugar mode all the time not fasting. Metformin in blood will switch it back to fasting.


last comment I had was that your excellent comments clearly reveal the complexity of Type 2 diabetes and that no two cases are the same.

I do not know what you mean by heavy dawn effect - did this disrupt your a1c to 13.3 % or was this a case where you where in the low sevens with the dawn effect responsible for lifting your a1c to seven from 6.x.

Your data does not give one like me any info to better understand your situation.

usally as I am told, very low basil inslun numbers usually cause the liver to overload glucose on dawn effect and other liver emergency glucose add. This is the case for me where Basil Insulin release is tiny till
starlix/glyburide pills cause it to ramp and in my case useable for meal handling.

For me metformn has been successful and saved my eyes and other organs and helped boot out actos.

Thank you.


I realize much of excellent response while interesting and helpful in treating diabetes issues misses a critical point.

One buys a piece of test gear to help with data collection, managing details and hopeflly catching serious exceptions that when caught on timely basis and enable my response; limit a full fledged fire when one is not looking.

Accuracy and data collection are not in question.

So far the Dexcom unit has done a great job on most of my hit list except the part of catching and warning me of a serious fire irregardless of speed of action on my body.

When I read their mantra of “take control - live uninterrupted”. For me that means catching the unusual and giving me sufficient heads up to catch and recover. Whether or not my body is well behaved or I have done correct treatment of snacks, pills, insulin is not the point. I expect it to catch the extra-ordinary and warn me irregardless of mistakes on my behalf and give me opportunity to recover.

Snoozing thru a crisis of mine without warning me and such I have to closely monitor, supervise and watch unit based upon my prior experience manually doing this for past 3 months doing 20 to 38 pin pricks a day does not yet match up with my concept of their mantra as a helpful tool in all regards.

My goal in writing was to identify how best to work with device as it is and what simple changes and stratagies would give a earler warning and complete Dexcom’s unit as a critical tool in my bag of tricks.

I do not see where the excellent and helpful comments about treating diabetes have to do with the performance of what has been an excellent job so far except on warning and catching unsual situations that occur fast.