• Guest - w'd love to know what you think about the forum! Take the 2025 Survey »

How do you record your diabetes?

That's great because you make sense of blood sugar patterns, not everyone can statistically analyse. I've been doing stat analysis for over 6 years now.

What i've found useful is dividing the day into 3 hour segments then breakout the readings from 0-3.9, 4-7, 7.1-10, 10.1-15, 15.1-20 & 20.1-33 mmols. Which is then split into 7, 14, 28, 90, 180 & 365 days, compare year on year. Additional analysis is the interquartile range and outliers.

The really useful analytics I put in place was Decile Analysis which each 10% is split into daily segments of early hours, morning, mid morning, afternoon, late afternoon, early evening and late evening. This way i've isolated where any hypos / hypers are concentrated in the day and the excel workbook prompts me to act on where ever the highest concentration of lows or highs have been ocurring.

It shows me rolling averages, frequency distributions, SD, short run to long run 7 days to 365 days. which clearly highlights changes in paterns or behavioural habit changes.

As i mentioned in my response earlier i started using Apps to record my insulin and carb intake again. All in all the more information i have to build a clear picture the better control ive got, keep glucose variability to a minimum and my Hba1c typically straddles between 5.8-6.1%.
 
Wow - That's some in depth analysis!

It certainly is very useful to have the data in an highly manipulable format such as excel (and indeed have the ability to derive stats)

One of my motivations for my post was to perhaps share this with the community and see whether it could be of use - I know there's plenty of naysayers, but human memory is fickle to say the least(e.g remembering BS), and as they say knowledge/information is power.

My analysis is very simular, however slightly different methodologies. I have pretty much automated a lot of the excel spreadsheet; just a case of pasting raw data into a assigned tab in the worksheet.

Id be keen to see your spreadsheet design if your up for exchanging; sure we could reap some mutual benefits.
 
All i do is input my blood test reading, day and time. Everything else is entirely formula driven so as long as the processor can cope. I got more into the statistics area in the last year or so rather than just rolling averages. Now it does near enough what i want and tells me where exactly i go off course which means i don't have to think it through anymore as it does it all for me.

I think that Apps need to be proactive and actually tell users where they need to correct things.
 
Unfortunately diabetes is all about statistics if you read into medical reports, our Hba1c predicts mortality and complications. Therefore i've come to realise over the last few years this is a numbers game, and a serious one at that too. What a ******* disease this is, we need to educate ourselves as much as possible in diabetes management and beat this **** thing.
 
I was using Diabetes Connect for about 9 months after I got my pump and it helped me quite a lot, but I gave up on it finally because it does nothing with the information you give it about basal rates. It doesn't plot them against BS recordings for example, so it's unhelpful for basal rate testing, and the interface for inputting basal rates is hopeless, as is the resulting graph. I am looking for an app that understand pumping.
 

PM me if you'd like to exchange more details on spreadsheet analysis.
 
A great thread to read. I keep mine in a log book as mentioned above and all sport activities I just add onto my Strava account with what I consumed on the run/ride and what my levels are after 1 hour, 2 hours and post-run/ride.
 
If anyone is interested in correlating their data, from logging my insulin and carbs on MySugr i've done a 7 day comparison as i havent been recording my carbs and insulin for very long.

I exported my carb and insulin data from MySugr App then put it in my excel workbook with my blood sugar analysis.

Ok so this is what i analysed. My bolus to carb ratio has a high inverse correlation of -0.85 or -85% to my blood sugar standard deviation.

The standard deviation has a high positive correlation of 0.83 or 83% to my average daily blood sugar reading.

Or directly the Bolus to Carb ratio has a high inverse correlation of -0.69 or -69% to my daily average blood sugar. This will have an some sort of error rate depending on how accurately we count our carbs.

So if you can grasp the maths, as my bolus to carb ratio falls my blood sugar increases which is directly linked to a higher standard deviation.

The higher your blood sugars the lower your bolus to carb ratio is, which in turn means an error in calculating your insulin doses for your carb intake.

This will obviously change depending on illness, exercise, variability in insulin resistance and absorption rates etc. But the basics go hand in hand. I need to look at my results long run and see what meaningful info i can produce.

Complicated, but if i can create a multidimensional analysis and correlate the variables, these will be great prompts to design an indepth app where it tells us where we are screwing up and point us in the right direction.

I can't wait for the doctors to advise me if things change from day to day, then ask them what to do every 12 months when they see me.
 
Last edited by a moderator:
Err, I don't mean to rain on your parade @pseudobob, but if I see that my blood glucose isn't dropping as I'd expect after bolusing and eating, then it's fairly obvious that I've not had enough insulin to cope. If I see that happening on multiple occasions then I may check that my basal is correct. Why do I need an app to show me that?

That's been the standard approach for dealing with diabetes as long as I've been on MDI (20years plus). Review your own records regularly to spot patterns and make changes, don't rely on a doctor to tell you what to do. It's surprising how easy the old log books made that!
 

Yeah sorry that's me stating the obvious, going into a number frenzy doh! I know it's obvious, think i just got carried away and starring blindly into numbers earlier.

You don't need an App for that but even so Accu Chek released the Expert Bolus calculator meter. Same as correction doses for the 100 rule etc.
 
The other thing is, is that things change. A couple of weeks ago my blood sugar was high at 16 mmol, i had to take twice as much novorapid just to even get it back to 8 mmol by the following lunchtime without taking on extra doses to eat. It's not that i hadn't taken enough insulin compared to my regular day, it's that stress and viruses can severally affect things at times.

Contrary to that on wednesday evening i took half of my regular novorapid dose for dinner which was about 40 carbs and my blood sugar continued to drop below 4 for the entire evening even after consuming an extra 60+ carbs and i'd hardly taken any insulin. Things can be quite variable at times. Not been exercising in the last for weeks so my blood sugar drop wasn't exercise related so my blood sugars can be quite variable and take on some slightly random changes. But you know peoples physiology is different.
 
I have the Optium Neo, it probably has a diff name somewhere other than Australia.
It came with a cord I connect to my laptop and dl all my results.
I email them to the Diabetes Educator each week and they adjust my doses.
I really like doing this, I used a log book for many many years, and it was tiresome making it up in the car before my appointments, using different colored pens for that authentic touch
 


That's not unusual, many of us find we are more insulin resistant when bg levels are high.
 
Previously I used Vario bG machine very useful records results and shows on visual chart. Have just changed to Omnipod Pump last week info also recorded on this haven't written down for years. As for computers I'm phobic! often I use one at work something goes wrong.
 
Your sensitivity to highs and lows will change depending on whether you're running high or low. HbA1c is a good guide to your average.

High blood glucose leads to complications.

Its much safer to test, and not difficult. Sometimes the results will surprise you...
 
I just continually check and adapt. Occasionally a pattern emerges, but only every couple of years. However, every 4-5 days, something alters and a temporary basal is required. I can predict the day before I get symptoms that I've caught something. It's a bit like comparing jumbo jets with a fighter. Jumbos glide for ages on their own, but have to travel a very well planned route. A jet has the aerodynamics of a brick, but can adapt in an instant, provided the pilot watches everything.
 
You must log in or register to reply here.
Menu
Log in
Register
Cookies are required to use this site. You must accept them to continue using the site. Learn More.…