Reliability of glucose monitors

[janeliz]

I was recently given a new glucose meter by my practice nurse as she didn't like the one I had from the hospital. As an experiment I tested my blood with both monitors at the same time. I was expecting the same or very similar results, but to my amazement my first monitor read 11.3 and the second monitor read 15.6. All strips were in date and both monitors had control test in limits. How do I know which monitor is correct? I could be having a hypo on one monitor but not the other. I have only been on insulin 2 months so am still very much a novice. Any advice ?

 

[noblehead]

The meter are can be 15%+/- out and are by no means a precision instrument, I believe the higher the bg the more they can be out, with lower bg they are more accurate apparently.

Just stick with the one meter.

 

[Spiker]

It does not matter that much if the reading matches a lab test result, let alone another meter. What is more important is that results are consistent from one test to the next. So as @noblehead says, best use just one meter, or just one type of meter.

 

[SamJB]

Did you wash your hands? That makes a large improvement in precision, in my experience. If you're interested in getting an accurate and precise meter, the MyLife Pura and Abbott Freestyle Lite are the best out there.

 

[janeliz]

The low reading meter was an Abbott Freestyle Lite, but this would mean I was often reading under 4. The other meter a Gluco RX Nexus would read 8. Think I'll stick with the higher reading as I felt too well for an under 4 reading. After all they are only a guide. Although we rely on them for driving and DVLA regulations. And yes I do wash my hands first.

 

[forge]

Both meters are right they are telling you your BS is too high.

They should have just made them with no numbers and a green zone and a red zone with a blue zone in-between.

Because that is all we need to know.

Mostly they read high because of contamination of the sample site.

 

[Heathenlass]

They should have just made them with no numbers and a green zone and a red zone with a blue zone in-between.

Not of any use at all if you are Type 1 and need figures to calculate a bolus dose from .

Signy

 

[forge]

In my opinion the meters are too iffy for calculations. But others are probably happy with them.

I expect it is important to know where you are and what you are going to eat but relying on accurate meter readings is going to have problems because in my experience it is rare to get the the same reading with three consecutive samples.

My question is if you are T1 and you get a very high reading that you think is genuine do you
a) Try to correct it with one jab and not worry if you end up with a low.
or
b) Jab first with 80% of what you think and make another adjustment later.

I do not know what is best I am just curious. I am off jabs now after about 12 years.

 

[janeliz]

The meters are as you say forge iffy. As it is all we have to determine how we can try to control our blood sugar levels, then it does become important. For driving , tweaking insulin doses, what foods are good and bad and how you feel. I have 2 jabs a day of mixed short and long acting insulin. Only been on it 2 months since acute severe pncreatitis which wiped out my pancreas. (not due to alcohol as I don't drink alcohol)

 

[Spiker]

This is known I think as "the engineer's problem". What do you do if your data is unreliable? The engineer's answer, the real world answer, is you use the data anyway, and you adapt through experience. In our case you learn your meter, learn its variability, its likely range, its uncertainty. The uncertainty and variability in meters is no greater than the uncertainty and variability in the injection process, portion sizes, nutrition analysis, day to day changes in physiology. If you are a scientist you are paralysed by all these compounding errors. If you are an engineer you rely on the fact that over the long term they all average out. We need to be engineers.

Even bad data is worse than no data.

Shout out to @Engineer88

 

[Spiker]

My question is if you are T1 and you get a very high reading that you think is genuine do you
a) Try to correct it with one jab and not worry if you end up with a low.
or
b) Jab first with 80% of what you think and make another adjustment later.

If you get a high reading you

-Wash hands and retest to confirm
-Correct on the assumption the reading is accurate
-But, correct toward a safe target BG rather than a tight one. So aim for 5.0 or 5.5 or 6.0, not for 4.0. That's asking for trouble.

Engineers always have safety margins.

 

[Heathenlass]

Thank you , @Spiker , you answered that beautifully and far better than I could !

@forge , are you Type 2 ? It sounds as if you are, and have been on mixed insulin ?

Signy

 

[Heathenlass]

Apologies, I also meant to say that personally I know my meter is not 100% accurate but is accurate within certain parameters. I'm just used to that meter , and have a "feel" for it. It's rare that I make mistakes based on readings . I do believe in " The Law of Small Numbers " ( Dr. Bernstein's term I believe ) .

Signy

 

[forge]

I am T2 and over the years I went from low GI diet and exercise through to 3 x 30 units of Nova rapid per day + 1 x 40 units of Lantus per day = Total 130 units I determined my own dosage

In between I was on mixed slow and fast.

and now with no major carbs diet I am just on 4 metaformin tabs once per day.

For 28 years I was an engineer specializing in Instrumentation and Process Control of mega factories. Yes lab instruments and process instruments can be different, and yes all processes can have uncontrolled variables.

I don't want to give you a lecture on how it all fits together but these things are important
In industry we have.

1. a set point (desired value) that you aim for.
2. alarms called annunciators (flashing lights and sirens) that tell you when you have exceeded an acceptable error
3. You have trips that start shutting down the plant to ensure safety.

That is why I suggested the numbers are not important and green and blue and red zones are important

When it comes to auto control there are a few options
a) is called proportional i,e how far off set point determines how much correction
b) is as above but more is added or taken away depending how long it has been off set point
c) is the two above with a speed of change adjustment added.

These three are adjusted until the system cycles (hunts) and goes too hi then too lo then too high etc.
When it hunts it is backed of a bit so it wanders within acceptable limits.

For us lows are bad news and highs + time is bad news so we need to err on the high side.

Computer control on cars is different again and they use factory settings plus what is called INTEGRAL GAP Control but I will not bore you with the detail but what is worth knowing is

If your car will not start you can disconnect the battery negative for 15 minutes the computer will go back to factory settings and will probably start for you. It is like turning your PC off and on at home when it gets all confused.

So back to the meters the numbers only do half a job.

We have to decide what the zones are call them what you like
a) Good 4 - 7 maybe or 5 to 8
b) Iffy 7 -9
c) Dangerous above 10 and below 4

It is up to you where you zone it but the other factors are if you are driving or using machinery you might like to go a little higher because a low could cause death or injury if you pass out.

So back to square 1, 11,3 and 15.6 are both in the red zone.

 

[Spiker]

That is why I suggested the numbers are not important and green and blue and red zones are important

When it comes to auto control there are a few options
a) is called proportional i,e how far off set point determines how much correction
b) is as above but more is added or taken away depending how long it has been off set point
c) is the two above with a speed of change adjustment added.

You fail to meet your own requirement a) in your red / blue / green scheme because a) needs a numeric input. The clue is in your words "how far off set point".

Apart from that yes red / blue / green is a simplification of what we all do. For some it might be enough - except every T1 needs a numeric value to be able to calculate correction values, as @Heathenlass already pointed out.

Can you explain how any system with only 3 possible values, red / blue / green, could ever solve the correction problem, your "control option a)" ? And in a reasonable time, since for T1s, low errors and high errors are harmful and time-critical to resolve?

 

[JTL]

Both meters are right they are telling you your BS is too high.

They should have just made them with no numbers and a green zone and a red zone with a blue zone in-between.

Because that is all we need to know.

Mostly they read high because of contamination of the sample site.

I liked this then unliked it when I read further which proves ... I haven't a clue!

 

[forge]

Hello @Spiker I tried not to get too techo

The fact is with set point controls the measurement has to be off the set point to make an adjustment.

Example
A) when the two pointers are exactly aligned the control valve is exactly 1/2 open
B) If the control requires anything other than a half open valve the two pointers cannot be exactly aligned.

So every set point needs an acceptable tolerance.

In industry tank levels can overflow and they can get too low and the pump runs dry and burns out but somewhere in-between is fine. Someone has to decide what the set point should be and provide a safety margin and if necessary automatically shut the pump down.

We are the same we need to take defined actions when the readings are within defined levels.

The set points tho critical are not important, the action is taken on the difference between the desired value (set point) and the real value. The difference between the two defines the amount of action we take but that too has a tolerance level.

You might decide that a reading of 5 is ideal and between +4 = 9 and +5 = 10 requires X amount of jab and that is fine but if you decide that 9.3 requires different to 9.5 then you are just fiddling about, because the meters and sampling are not good enough.

 

[Spiker]

Ok @forge so you accept in your example that you need a quantified numeric input for corrections. Your red / blue / green model does not allow for this. We need meters with numbers.

 

[Spiker]

In industry tank levels can overflow and they can get too low and the pump runs dry and burns out but somewhere [/SIZE]in-between is fine. Someone has to decide what the set point should be and provide a safety margin and if necessary automatically shut the pump down.

We are not as simple as storage tanks. We don't just need a binary response to an out-of-range condition. We need a response with a measured variable magnitude. Therefore we need a numeric measure of deviation from the acceptable range. As you accept in your reply. So time to admit that the red / blue / green approach is not sufficient, please.

 

[robert72]

Before we hade meters we had colours (with numbers) but it was not always so easy to get the exact shade (you had to put a drop of blood on the test strip, wait a while, wipe it off, wait a while more, then match the strip to the colour chart. A bit like ketostix but so much slower )

Anyway - meters are a vast improvement in accuracy over what we had before.

I still have these in my diabetic kit drawer for some reason