target ranges on the roche accucheck combi pump

Status
Not open for further replies.
Dory

Dory

Well-Known Member
Relationship to Diabetes
Type 1
Hi all,

I've noticed a trend on my pump concerning the target ranges that accompnay each time block i have for my bolus/CF ratios. at the moment they are as follows:

00.00-09.00: 1:8 bolus ratio, 1:1.6 CF ratio, target range 5-7mmol
09.00-15.00: 1:7 bolus ratio, 1:1.6 CF ratio, target range 5-7mmol
15.00-00.00: 1:8 bolus ratio, 1:1.6 CF ratio, target range 5-7mmol

I've noticed that during the first two time blocks, any correction it calculates is to take me back to a target of 6mmol. this is normal for the accuchek pump, as it always attempts to correct you back to the mid point of your target range.

....EXCEPT....at some random points throughout the day, I've noticed that any CF it gives me is to take me back to 7mmol. Which is clearly not what it is doing on other occasions! I'm having to manually work out the extra CF i need on top of what it suggests and add it to the suggestion CF dose. It can't be any IoB/active insulin, as I've checked when it underestimates and the active insulin levels are showing as 0.0 on the screen. In fact, on some occasions, it even underestimates WITH IoB. for example, last night at bedtime:

23.50: 10.9mmol
active insulin remaining: 1.0u.
No carbs eaten.
recommended CF dose: 1.4

according to the above (ie IoB+CF) and using the CF ratio of 1:1.6 that I'm on, that would bring me down 3.9mmol, ie down to 7mmol. Buy accuchek apparently brings you to the middle of your target range, so for me the meter should have worked the maths out to bring me down to 6mmol (as it was, I manually worked it out and ended up giving another 0.6u on top of the 1.4 it suggested; hence a perfect reading this morning of 6.0mmol).

I have noticed that also the calculator sometimes doesn't suggest a correction to bring my reading within the target range (ie one night, the active insulin was showing as 0.0 and the correction it suggested would have brought my reading down to 9.0mmol - nowhere near my target range!!)

The question I'm asking is: has anyone else noticed their Roche pump correcting to different points in the target range other than smack bang in the middle? I'm worried that my equipment is reaching the end of its life and needs changing (even thoug the meter is under 3 years old as it's a replacement for the original).

The other alternative (and this is something that was discussed in a separate thread by Mike and Tropywench) is regarding what the pump considers as active insulin. We've already worked out that the figures shown in the 'active insulin' field only reflect CF bolus and not any food bolus. BUT what I have noticed is that the underestimated CF suggestions my meter is suggesting (when there's a '0.0' in the active insulin field) seem to come after i've had a food bolus. SO......it's almost like it's taking the food bolus into account when suggesting a CF, but it's not showing the remaining food bolus in the active insulin field. This is even more worrying!

thoughts guys?
 
Haven't read all through your post as brain tired today. So probably misunderstood what you are saying asking.
Simple solution is change your correction factor for the different times of day if you are not on target, and your basal is correct.
A pump will not be faulty as you discribe. 🙂
 
sue - thanks but think you need to re-read my post later on...🙂

the meter is not correcting to the numbers I'm asking it to 🙂
 
I know what you mean but I don't notice it regularly probably because as I don't do very much regular exercise (I mean it's sporadic so I make adjustments at the specific time rather than have it programmed in, or even have a set procedure for it) so things are usually pretty straightforward and fairly simple for me now.

I did get quite a lot of glitches like this a while ago so what I did is gonna sound ridiculous and it shouldn't do - in fact anyone would say it can't possibly - but it worked. Madness.

Make your ranges eg 5.5 to 6.5, or 4.8 to 7.2 - or any other two numbers different from the original where smack in the middle of them is THE number you want to be. Just changing it for the sake of it, is all it seemed to take.

I wondered if it's summat to do with - like predictive text, where it gets fixated on predicting one word all the while once you type in a C or whatever, no matter how much of the real one you accurately input - or something? Clueless.

Unless anyone knows the real answer to this fascinating question LOL
 
hi TW,

yes somehing random like doing what you've suggested may re-set it! seems odd but as the machine is being odd anyway, fight odd-ness with odd-ness!

thanks!
 
I think that late night adjustments maybe accounting for you going to bed so aims for the hihher end of your target so that you don't hypo in your sleep morning djustments may just take you to midrange.

I may need to read again too lol.
 
That may explain some of mine Paul, that hadn't struck me.

God - I wish I could teach my brain to be more analytical.
 
wel that's what i thought both....but surely the machine can't be that intuitive? surely it simply works off an algorithm that says at XXX times these rations YYY and ZZZ are functioning. After all, that would require the programmers to somehow programme into the machine that between the hours of AAA and BBB a human is asleep - which I don't think they would do as this is too proscriptive and they'd be inundated with people saying 'i work shifts that's not correct' or 'i don't sleep during those hours so that's not correct' etc etc

plus it's not just at bedtime, this happens occasionally through the day too, ie anytime from lunchtime through to 9pm.

so baffling!
 
Last edited:
Dory said:
wel that's what i thought both....but surely the machine can't be that intuitive? surely it simply works off an algorithm that says at XXX times these rations YYY and ZZZ are functioning. After all, that would require the programmers to somehow programme into the machine that between the hours of AAA and BBB a human is asleep - which I don't think they would do as this is too proscriptive and they'd be inundated with people saying 'i work shifts that's not correct' or 'i don't sleep during those hours so that's not correct' etc etc

plus it's not just at bedtime, this happens occasionally through the day too, ie anytime from lunchtime through to 9pm.

so baffling!
Click to expand...

Have you reported it to Roche? It may be a known problem. As a programmer in a former life I have come across the most glaring of errors for situations the programmer clearly never envisaged and never tested fully.
 
yep, will have to give them a call tomorrow alan and see what they have to say. will keep you all posted!
 
I don't use the combi pump but I do have an expert meter, which I believe is the same as the one that comes with the combi pump to all intents and purposes.

I've also noticed that it doesn't suggest a big enough correction factor to bring me to my target at bedtime - the correction it suggests is just to the top of my target range. I think that is built into the algorithm and agree that this is far too inflexible. I have set my ranges very small to minimise this effect, although I don't make great use of my expert meter having managed without it for several years and finding it far too much of a chore telling it how much I'm eating etc.

I agree that this is not a fault of your pump, but could be a fault of the meter, if the answer that's been suggested doesn't account for all the anomalies! However I imagine it's a design fault rather than a fault with your particular meter.

I'd be very interested in the response you get from Roche - but I wouldn't hold out a lot of hope that you'll get to talk to someone who really fully understands it!

I took part in the market research testing for this device before it first came out, and the instructions contained some errors which demonstrated to me that whoever wrote them really didn't understand how the meter worked!
 
Dory, I notice on another thread that you are a bit of a grazer... Is it connected to the delay that the Combo/Expert uses to avoid stacking. Can't remember what it's called, not meal rise (though that would also reduce corrections) and not insulin duration, but the other one... acting time? Basically the period before you respect your BG to fall. Along with the meal rise these could bothe be reducing the corrections the meter will suggest if you are eating/injecting close together I think

EDIT: just looked it up. The setting I was thinking of was 'offset time'. When boluses are given close together (within the acting time set for the insulin) AFAIK the initial correction/meal bolus is assumed to be working perfectly and additional correction is only calculated on the BG value over and above meal rise/offset time parameters to avoid stacking (p99 and p100 of the big Expert 'training handbook'). I suspect these factors are at play in the behaviour you are seeing?
 
Last edited:
god Mike - you could have it there - the meal rise limit.....

just checked and that seems to add up - my target is 5-6 (ie 6) and my meal rise is 3.0mmol. all the corrections are looking to correct to 9mmol......

my acting time is 4 hours. So basically the settings are currently allowing my reading to go as high as 9.0 for hours hours after every single thing I eat and bolus for! Not good for grazers as that effectively means if I'm over 9 it will only ever correct to that!

That also means the settings are programmed so that the meal rise can be allowed for as long as your acting time (ie 4 hours in my case). Which isn't really a very good setting - after all meal rise limits should really only be valids for up to about 2 hours after I eat!

Huge programming error! Question is, should I reduce my acting time to 2 hours (to allow a 'proper' correction after 2 hours of eating) or should I reduce the meal rise limit to 6.0mmol (which kind of ruins the purpose of the function!)????? hmmmm....
 
ps as a perfect example, last night:

9pm - 11.4mmol, active insulin 0.0u, suggested CF 1.5u - which would have brought me down to 9.0mmol.

I had eaten 40g of carbs at 7.45pm, 1hr15 mins before hand......
 
And what IS your Offset Time set at? - mine is 1 hour?
 
I await the outcome of your conversation with Roche!
 
Hiya TW,

I was recently told by my DSN that the offset time was the time taken for all the insulin you have injected/pumped to be completely active, so have that set at 2hrs (as per DSN's advice that it takes 2hrs for humalog to be 100% active and being used).

Just about to phone Roche now.....
 
Dory said:
Hiya TW,

I was recently told by my DSN that the offset time was the time taken for all the insulin you have injected/pumped to be completely active, so have that set at 2hrs (as per DSN's advice that it takes 2hrs for humalog to be 100% active and being used).

Just about to phone Roche now.....
Click to expand...

Keen to know their response.....i have the same issues!
 
right, just had a chat with 'Stephanie' from Roche. very helpful lady.

Essentially, Mike has hit the nail on the head. Any CF doses done within the acting time you have programmed into your meter will only correct to the upper limit of your meal rise. So if your target is 6mmol, your meal rise is 3mmol and your acting time is set at 4hours, and you test 3h59mins after eating/bolusing and your reading is 11.0mmol, the mater will calculate enough correctional to bring your reading down to 9mmol, NOT 6mmol.

I explained that isn't helpful for grazers who eat little and often within the hours of the acting time, as your pump is therefore always assuming that it only needs to correct to your meal rise limit. She agreed with me and even said it was a little ironic given the pump was designed to allow that freedom to eat little and often if you wanted to!

She said she wasn't aware of any options to unlink the acting time and meal rise, her only suggestion was that i might want to turn off the meal rise limit function which is what other users do. She did however promise to pass my feedback on (maybe this will affect future productions, who knows).

I also raised the issue of the phrase 'active insulin' on the bolus calculator screen (ie wher eit tells you how much insulin you have left in you). She also confirmed that this was what Mike and I (and others) believed - namely that this only refers to CF doses, and not meal bolus insulin. The reason given was that the machine assumes that any insulin you give for food will be used directly up by the food item you are eating, therefore ther ewill nto be any 'leftover' to count as 'spare' insulin that would cause insulin stacking. That kind of makes sense. Howeve rI have told her that if that is the case, the screen should be clearer marked to that effect (ie to say 'active corectional insulin' or something along those lines).

So in short, no solution other than not allowing for a meal rise (begs the question why have that option then Roche?)......
 
frustratingly, I have just tried to set my meal rise limit to 0.0mmol and the lowest it will go is 2.8mmol!

So i will constantly have to manually work out CF doses from now on!!

HOW UNHELPFUL!!!

EDIT: unless someone knows how you turn that specific bit of bolus advice off on the Roche pumps??
 
Last edited:
Status
Not open for further replies.
Back
Top