26 Comments
How important is the standard deviation (SD) measurement to you? (SD is a measure of a person’s variation in glucose readings)
You must be logged in to post a comment.
© 2023 T1D Exchange. All Rights Reserved.
SD is a helpful additional measurement in determining how well the TIR reflects BG control. I monitor both on a routine basis.
Use close guidance of Endo.
It’s not a measurement that I see that often. But, I know that it’s a good measurement to tell me whether I am staying within my range. If I am going above and below my safe range, my SD will be high. Best to keep the SD low.
Important enough that it let’s me see rise and fall of random overall BG patterns. However, quite frankly I generally don’t pay all that much attention to it because I have learned to trust the Tandem CIQ w/Dexcom G6 integrated system. My TIR is pretty spot on at ~ 88 to 92%.
It’s very similar to TIR but far more difficult to explain to most people.
Oh yeah…just think back to statistics classes! Fortunately I upload my TSlim/Dexcom every month and get a nice visual of the range between my high’s and low’s.
Thank you for this question. Have often questioned A1cs because I have so many lows. Multiple endocrinologists and ophthalmologists and primary care have all convinced me after 3+ years to lower basal rate and some bolus calculations. Last A1c preadjustment was 5.6, too many lows. Next one coming up. I’ll ask endocrinologist’s assistance in calculation SD.
It’s ‘important’, but I’m already doing the best I can, and as long as it’s reasonable, and I know why things vary, I’m content. Can’t say it’s ‘very important’ because it’s not – I can’t stress over something further that I know I’m already doing my best to keep in control.
Never heard of it. Imagine its something I can see just by looking at CGM graph.
Time in range is much more important and may be replacing the value of an SD measurement.
TIR “replacing” SD? – insufficient and not fully accurate, I think the common data will show. They are NOT the same, not mutually exclusive. Like saying tire pressure is so much more important than tire tread or roadway condition or steering stabilization … NONE of which replace another (except drunkenness or inebriation, just as gross operator error is 100% more important in prevention or can be held 99% causative in a crash).
I don’t like having to scroll past all these articles to get to question I’m trying answer. Please put the question back at the top.
The more information I have, the better chance I have of best possible control.
It is conceptually important to me but, in truth, after 50 years of this disease and now hormonal changes, I am struggling most days to keep my bG below 200. 🙁
Me too, Julie!
This makes me wonder if there have been any studies done to try and help in this area!
SD woud be important to gauge the response of a true closed loop system. Since current CGM is not true closed loop it’s pretty much irrelevant.
My SD is lilrly very large because my SG rings like a bell after I get up until I go to bed. Right now, TIR is what I watch. Even that has a large variation 92% to 56% from day to day.
Even when you refer to a true closed-loop is NOT like internally created insulin which goes directly from Islets in Pancreas **and** glucagon directly from the liver’s stores into bloodstream into cells’ glucose/insulin receptors. But we are stuck with subcutaneous skin vascular nature, accounting for the nominal 15-minute delay in achieving insulin’s effect. Not to mention disruptive barriers to immediate and predictable insulin effect of scar tissue or infusion site near large muscle when I gets exercised! So, everything we do AFTER CLOSE-LOOP is perfect is still a long shot away from true pancreatic endocrine activity.
SD in the mid 20’s and below indicates I’m doing quite well coupled with TIR in the mid 90’s.
This measurement is probably important for me to make note of because my glucose readings peak and valley A LOT
Vitally important. An SD is a measure of the amount of variation or dispersing of a set of values. A low standard deviation indicates that the values tend to be close to the average, while a high standard deviation indicates that the values all over the map.
The math formula is usually represented by the Greek letter sigma σ . The goal is to have σ < 3. In other words Average Blood Glucose divided by σ should be less than 3.
Many analyses attempt to contain variation or dispersions to less than 3 SD. If you go beyond 3, you and your data might be called 3 sigma deviants. I may resemble that remark in some respects. ☜(ˆ▿ˆc)
I have a strong background in statistics and certainly understand the measure and its importance, in principle, in managing blood glucose. That has not caused me to notice or use it as a metric for assessing my control. My control is limited mainly by my errors. When I don’t make any errors, such as forgetting to bolus until I hear a high alert, or mis-estimating the carb (or fat) content of a meal, my BG is very well controlled. So the source of variation is errors that I’m already working to control, and the actual number seems irrelevant to my decision making.
Ceolmar (SP??), I commend your thinking and explaining. A tool which doesn’t have immediately percievable cause+effect relationship to our behavior is not as effective on changing it as those that do.
Interested to see where such discussion might lead, especially in the hands of diabetes educators!
GOOD WORK.
Up until a few weeks ago, I had no idea what it meant. I read the definition and went, “huh?” Then a lovely CDE explained it to me and I finally “got it.”
And, while I understand it now, I am still trying to come to terms with how I can use this measure to help me manage my diabetes. Having all this data really doesn’t help if you have no idea what to do with it… I am still in learning mode and will get there, I hope!!!
I do think it’s important as swinging from highs to lows makes life difficult – it is easy to see your SD on the Dexcom Clarity reports. I have read less than 50 is a goal for most diabetics.
If you have read that less than 50 is a good target, I think you should be reading better stuff. As others here have said a target of AROUND HALF OF THAT is the ballpark we should all be playing in. I, too, am disappointed at SD of 38 or 42 etc…. despite good A1c and average SGs, etc.
Without knowing SD, an average Sensor Glucose or Blood Glucose reading (A1C) is limited in determining good control. I am a brittle diabetic and my sugars abruptly rise and fall. My A1C is great (<7 my entire lifetime) and TIR is 85-90%, but my SD is often runs 30-40. I prefer 20-25 SD which means I'm in better control.