I hate to even admit this but this is the first I’m ever hearing of standard deviation! Where do I find this number??? My endo has not once even remotely touched on this topic … like someone else said she only looks at my a1c results and I never thought that was fair. Especially the first year or so after going on the pump, my a1c didn’t really go down past 7.5 even though I was seeing such an improvement in my bg readings. She would always explain to me that you can get a “B” average by either getting all B’s or getting some A’s, B’s and C’s. Is this the same theory as the SD? Could that be why I finally saw a big improvement in my a1c? Maybe my SD went down a lot?
Stacey, you use a Minimed, right? Take a look at Grace’s attachment here. If you use Carelink on the Minimed website, it’s figured in your data Quick View Summary.
Thanks Melissa!
How often do you all download to carelink (for minimed users)? I never do. Unless I used the CGMS and want to share the results with my CDE which I will be doing in 2 weeks. I feel it would be missing data since I don’t enter every reading in my pump. Does anyone else feel the same?
I uploaded every Sunday, but I was using a meter that communicated via RF with my pump and I used my bolus wizard for every meal, so I had a lot of data. My CDE and endo were big on making sure that I used the bolus wizard every time. My suggestion would be to upload data every couple of weeks.
Since I got the new Carelink USB thing, I do it every other day at work (I have a Mac at home and a PC at work). My meter wirelessly sends readings to the pump, so it knows my food, insulin, and BG’s for the day.
A standard deviation is showing you or telling you what your average swing in BG is. If you only check you BG 3 times a day and you numbers were as follows (100 in am 200 at lunch and 100 at dinner) your standard deviation would be 100. It is that up and down that is harmful to your body because it shows that you don’t have good control of your over all numbers even though you may have a great A1c.
I upload my data every couple of days to look at how I am doing on the graph that shows my CGM data. I tweak a lot of my pump settings and I almost have it where my BG is in the green zone all the time. Barring a few post meal spikes!!!
Thank you for your response I think I understand a little bit better now!
David,
Your control over your blood sugar is amazing to me. I can hardly believe it. Please post Carelink proof 
Just kidding, but really, congratulations to be in the “green” most of the time. That’s awesome, and if I thought it was actually possible I may have kept trying harder.
Grace
I’ve been downloading my pump for awhile and now I have the CGMS. I am always frustrated by my variability in BGs. This discussion made me curious so I went back to look as the standard deviation and mine is usually 20-38. I test my bgs 10-12 times daily even with the continuous sensor. I think this helps alot. So I am continually adjusting and correcting so that I can maintain tight control. I tell my patients that managing my diabetes is my other full time job and the pay is lowsy but the benefits (no complications) are worth all the hard work.
I try hard when it comes to managing my BG, but please lets not get carried away!!! My goal is simple, I want my body to hold up for as long as possible. That is why I read a lot and frequent sites like tudiabetes, because I want to learn as much as I can from everyone who might have a thought on how to manage diabetes better.
Thank you for the compliment and I decided to take you up on uploading my data from Carelink. This data is for the past week–a continuous work in progress!!!
8506-1217385781175.pdf (35.9 KB)
Those numbers look great David! And I like the comment about a continuous work in progress! It really is 
One remark-- you have VERY few hypos!! Only one this week!! When my control is good, I do have lows that are considered part of tight control (or so I thought)… they are usually between 55-70, but still they happen.
Any tips on decreasing lows?
Honestly I have to admit my control has really increased since I started using CGM. A month ago you would look at my graph and see that after dinner my BG would just climb until 3 in the morning. I started tweaking my basal rates and now it is flattening out because you can really see the trends after each week goes by.
Controlling lows I think (at least in myself) has to do with the basal rate of course that is assuming that the sensitivity rate is set correctly. If your sen rate is set properly and you are accurately counting your carbs then I would turn down the basal rate. I used to have 4 basal rates plugged into my pump, now I have six or seven to control the weird areas that might need a little more or less. Does this help at all? I am sure that you tweak your basal rates, you may just need to find a loner CGM for a couple of weeks!!!
Kristin,
I tried this at breakfast today, and it worked pretty well. I feel a lot better this morning. Before bkfst my BG was 202, and my meal had 51g carbs plus eggs and cheese. So I used the bolus wizard to calculate the correction and meal dose, and took it all at once as a normal bolus. Then I waited 20 minutes to eat. Two hours later I was ecstatic to see by BG at 80! I think I would have been better off doing the dual-square you mentioned though because I did end up eating 8g glucose. Its nice to get a positive result from trying something new. I was really banging my head against the wall!
Grace
Great news! Please let us know what you find works best for you! There are many of us experimenting with the timing of the bolus… so we can really learn from each other
Let me share with you something from the third issue of
diaTribe which you should keep in mind (it comes from an interview with Dr. Zach Bloomgarden who has a huge clinical practice in endocrinology of more than 1,000 patients in New York City. He has also written many articles on diabetes – nearly 250. Since 1994, he has also written over 150 columns, “Perspectives on the News” for Diabetes Care. Dr. Bloomgarden is currently a Clinical Professor of Medicine at Mount Sinai School of Medicine.):
Kelly: The next area we wanted to jump to was your opinion on continuous glucose monitoring and your views on the outlook for the technology.
Dr. Bloomgarden: Type 1 diabetes is a hugely different burden from type 2 diabetes, and the easiest way to quantitatively understand that is to have a thousand patients or so with diabetes that you treat for years and download their meters day after day after day – when you do that, you’ll find that the standard deviation of the blood sugars in a person with type 2 diabetes is 10 mg/dl or 20 or at most 30 or so. The standard deviation of the blood sugar of someone with type 1 diabetes is at least 50 or 60 and often 90 to 100. So that translates into just huge, huge variability in blood sugars. Some people with type 1 diabetes don’t do the frequent testing that I think that they should do, which we’ve talked about already, so I can’t get a sense of what their overall pattern is. But all the different meters allow you to generate a frequency histogram of blood sugars. The frequency histogram of a person with type 2 diabetes, if you bother to do it, would be a very very narrow bell curve, and the frequency histogram of type 1 diabetes is sometimes like a square wave going from 40 to 400. In that setting, having CGM is just nirvana. Some of my patients who are pump patients who started CGM have said that this is much more important to them than the insulin pump. And it really is. The insulin pump is not all that big a deal in the current era of rapid-acting insulin. It is certainly better, but it’s just a little better. CGM is, however, in version 0.6, and not quite ready for prime time. Patients need to be really technologically savvy and financially well-off to handle it.
The bottom line: while that is a big standard deviation, it is actually pretty normal for a person with type 1 diabetes!
I always enjoy reading your perspective, Scott. Thanks for the informative post.
For all my understanding of standard deviation and how to figure it, I admit that the best control I’ve achieved is in the 50-60 range. And I test 10-12 times per day (though my insurance will cover only enough strips for 5-6 times per day). It’s comforting to know that, although it’s certainly not optimal, that it’s not abnormal either.
Thanks, Scott! How nice to see a reality check coming from a doctor! The pump really does make one quite reactive in dealing with diabetes, and the opportunity to correct down to normal from a high within a few hours must be contributing to a higher standard deviation in T1’s. I’m playing around with timing of insulin before meals to chop off those peaks now – which is just conventional wisdom to someone who learned to manage their diabetes on slower-acting insulin delivered by syringe.
Hello Grace, this is a very useful and important topic. My standard deviations were in the high 20’s for several years before pumping. After pumping for 14 months my standard deviation is 24 for the month of July. My pumping has reduced my number of highs and lows and that has resulted in the lower standard deviation. I have read several times that a standard deviation of 30 or less is considered good. I disagree. I want mine tyo be less than 20 but I have not been able to get it that low. A standard deviation of 30 is much more of a roller coaster type of control than with 20. The higher the standard deviation the more likely complications are to occur, even with a good average and A1c. If the average blood sugar level is low ( ideally 90-100) and the standard deviation is also low (30 or less) then good control is present.
I’m taking AP Statistics this year in school, so I know what S.D. is… but I don’t REALLY know what it is… at least in this context.
Can anyone help a girl out?