These are chat archives for nightscout/intend-to-bolus

21st
Mar 2015
kenstack
@kenstack
Mar 21 2015 00:44
The commands are there somewhere becusse you used to be able to do it with the old paradigm software from your pc
Toby Canning
@TC2013
Mar 21 2015 03:07
Do you think that program would still work with a 722 if I can find it?
Jason Calabrese
@jasoncalabrese
Mar 21 2015 04:11
I've always heard the 522 and 722 are the same
just the different cartridge size
Ben West
@bewest
Mar 21 2015 05:53
5xx and 7xx are the same, just different reservoir
some of the constants are different, but the mechanics are all the same
@kenstack indeed, the way I learned about them is decompiling a copy of paradigmPAL software that someone sent me
Scott Leibrand
@scottleibrand
Mar 21 2015 06:13
@tc2013 Ya, we query for pump history every 3 minutes (before taking action), and again if we took any action.
diabeticgonewild
@diabeticgonewild
Mar 21 2015 11:09
So @TC2013 , I have a copy of the ParadigmPAL software if you want it....

By the way, I am done with my school work for the weekend. Got 100s on everything...

I also got the getting distracted mode out of my system, so I will get a lot of programming done, at least today, if not all weekend.

All I have left is
  1. Finishing up intrinsic Gaussian Conditional Autoregression, which I have almost finished
  2. Getting the differential equations I set up to work
  3. Adding a line of code to implement the Markov Chain Monte Carlo process
  4. Testing my work
diabeticgonewild
@diabeticgonewild
Mar 21 2015 11:14
That's with respect to the mathematical modeling
diabeticgonewild
@diabeticgonewild
Mar 21 2015 11:29
@TC2013 , I have the ParadigmPAL software and the dinky USB test strip connector for the BD Paradigm Link meter. All you would need to do is buy this off of EBay, and run Windows XP on your computer.... http://www.ebay.com/itm/paradigm-link-blood-glucose-kit-/331508311377?pt=LH_DefaultDomain_0&hash=item4d2f6f6551
$1
I mean, I'm not going to use it, so you might as well take it from me.
kenstack
@kenstack
Mar 21 2015 11:34
@tc2013 @bewest im even wondering if it would work on the newer pumps .... @diabeticgonewild is right Re needing xp - doesnt windows 8 have an xp emulator ? I'm now a Mac person so my windows knowledge is not great :)
diabeticgonewild
@diabeticgonewild
Mar 21 2015 11:35
yeah, but drivers are an issue
I can try loading it on my computer right now...I just need to find my USB disk drive
I have Windows 8.1, and according to forums, ParadigmPAL does work on the 522/722 series
*up to
In lieu of using the dinky BD Paradigm Link meter, ComLink can be used, but it's not cheap (also it's a serial connection, so you might want to get an adapter if you only have USB ports)
Like I dropped that meter on a tile floor and it cracked the screen...it sucks
@kenstack , I used to be a computer repair technician
diabeticgonewild
@diabeticgonewild
Mar 21 2015 11:42
the cheaper solution is the BD Paradigm Link meter
diabeticgonewild
@diabeticgonewild
Mar 21 2015 13:05
I actually just had a physical reaction to a low...for once...I broke out in a sweat! That probably hasn't happened to me in 5 years, courtesy of autoimmune autonomic neuropathy, so obviously I am getting better...
Toby Canning
@TC2013
Mar 21 2015 15:02
@diabeticgonewild That's good about feeling the low... and bad too that you are having one. I just bought the meter off the link you sent me, so forgive me if you were planning to--I will give it to you if so!
Toby Canning
@TC2013
Mar 21 2015 16:04
@diabeticgonewild I just re-read your post and see you weren't going to use the stuff... :) my brain isn't working this morning...
Toby Canning
@TC2013
Mar 21 2015 21:47
Spent more time with the 522 today understanding how basal delivery works. On the 522 that I have, it delivers the basal in equally spaced .05U increments, starting at the top of the hour.
.1u/hr=a click every 30 minutes(presumably .05U are delivered with every click)
.5u/hr = a click every 6 minutes
1u/hr = a click every 3 minutes
2u/hr = a click every 90 seconds
3u/hr = a click every 45 seconds
6u/hr =a click every 22.5 seconds
35u/hr =a click every 5.2 seconds
Toby Canning
@TC2013
Mar 21 2015 21:54
The challenge is to know how much insulin was added or removed when you increase or decrease the temp basal over a 5 minutes period.
diabeticgonewild
@diabeticgonewild
Mar 21 2015 21:54
That depends on the time stamp of when the signal to change basals was sent to the pump. You need a timer function
Toby Canning
@TC2013
Mar 21 2015 21:55
If your basal rate is 1U/hr and you temp to 0, did you reduce you insulin intake by .05 or .1?
diabeticgonewild
@diabeticgonewild
Mar 21 2015 21:55
With a minimum of 5 minutes per interval
Actually the interval depends on the frequency of the micro bolus delivery
And you know it delivers in 0.05 increments so you are set
And the relationship and delivery times can be described algebraically and tracked via that time stamp
Toby Canning
@TC2013
Mar 21 2015 21:58
@diabeticgonewild All good points, but it isn't quite that simple
The radio communications is slow, so we depend even more on the pump telling us what it did.
As you know, I broke my radio and I haven't replaced it because I was working on getting the cc1111 to work
diabeticgonewild
@diabeticgonewild
Mar 21 2015 21:59
What about it pinging back?
I have 2 CareLink USB devices
Like and finding out the average time to read and ping back
Toby Canning
@TC2013
Mar 21 2015 22:01
I couldn't quite figure out how is was timign the temp basal delivery.
Scott Leibrand
@scottleibrand
Mar 21 2015 22:01
How does it change timing/spacing when you change temp rate?
diabeticgonewild
@diabeticgonewild
Mar 21 2015 22:01
So does it change the basal as soon as the command is issued via RF or after the communication is done
Scott Leibrand
@scottleibrand
Mar 21 2015 22:02
Does it go to whatever timing it would've been on if you'd started the temp at the top of the hour?
diabeticgonewild
@diabeticgonewild
Mar 21 2015 22:02
With the pump
Toby Canning
@TC2013
Mar 21 2015 22:02
So, while you might know that the pump was set to a temp basal of X at t, you don't know when it delivered or what it delivered during the timeframe
diabeticgonewild
@diabeticgonewild
Mar 21 2015 22:02
Yeah, top of the hour
Scott Leibrand
@scottleibrand
Mar 21 2015 22:02
Or does it start the new spacing from the time you set the new temp?
Toby Canning
@TC2013
Mar 21 2015 22:02
NO
I ran a few experiments and didn't understand the results.
I can type up my results, but I think it is best to run your own.
Scott Leibrand
@scottleibrand
Mar 21 2015 22:03
I think the key thing is not to change the basal rate until necessary. Once you set a rate, let it run until you need to cancel it, it completes, or you need a stronger temp.
IOW, don't be constantly adjusting every 5 minutes.
diabeticgonewild
@diabeticgonewild
Mar 21 2015 22:04
Crazy
Toby Canning
@TC2013
Mar 21 2015 22:04
That is a reasonable solution
Scott Leibrand
@scottleibrand
Mar 21 2015 22:05
With DIYPS, if we do a 0.2U/hr low temp, and then new data says we only need to be 0.6U/hr low, we let the 0.2 run.
At that point we either let it run, cancel, or set a 0/hr if needed.
Toby Canning
@TC2013
Mar 21 2015 22:06
But in reverse you make the switch
?
Scott Leibrand
@scottleibrand
Mar 21 2015 22:06
Sorry, gotta drive. Dana might reply. :-)
Toby Canning
@TC2013
Mar 21 2015 22:06
LOL
diabeticgonewild
@diabeticgonewild
Mar 21 2015 22:07
Like Hovorka changes his basals either every 12 or 15 minutes in his clinical trials but usually every 15
Did you use a timer @TC2013 ? I am pretty sure the pattern can be found with a lot of patience...
Like a lot
Toby Canning
@TC2013
Mar 21 2015 22:08
Yes, I'm sure you're right.
But yes, was using a timer.
diabeticgonewild
@diabeticgonewild
Mar 21 2015 22:10
Were you testing it by entering it in manually or remotely? I would start off manually doing it and then figure out an algorithm for the read/write/ping back discrepancy
I mean I can test it out myself.
diabeticgonewild
@diabeticgonewild
Mar 21 2015 22:45
Here's how I would approach it since we know how it does regular basal rates.
  1. Normal Basal Rate Delivery Interval [minutes] = (X [U/h])/(0.05 [microboluses])/(60 [min/hour])
  2. Anticipated Temp Basal Rate Delivery Interval [minutes] = (X [U/h])/(0.05 [microboluses])/(60 [min/hour])
  3. Start timer and normal basal rate simultaneously
  4. Wait for first delivery at anticipated interval and record time at delivery
  5. Change to temp basal rate and record time
  6. Wait for 5 microbolus deliveries and record elapsed times
  7. Calculate elapsed time from start of temp basal for the past 5 microbolus deliveries
  8. Repeat steps 1-7 until a pattern is found