News for hackers. Stuff that matters.
Re:A DIY diabetes kit by zaroth on Saturday July 16, 2016 @04:59PM (#12107858)
Sorry to hear about your son! Hopefully your son wasn't in ketoacidosis by the time he was diagnosed, and the initial days were just training and not ICU.

My own experience with using the Dexcom G4 CGM with my 1-year old son was extremely negative. First, it was terrible to insert the sensor -- it is a clunky mechanism which would sometimes fail and leave you with a worthless sensor, and at best would leave your child in screaming pain for minutes afterwards. The sensors would often fail, or start falling off long before their 7-day promised lifespan. The adhesive caused allergic reactions which left nasty red welts, and eventually you run out of fresh places to insert them.

When they were working at their best, you would have a readout which would tell you approximate glucose level, and if you are heading up or down. However, there are no error bars, and errors can be very significant. More calibration (measuring with a strip and entering the measurement into the CGM) is not always better. Sometimes we would find it was exactly correct, sometimes it would be 50 or more points off. You still always have to measure before dosing because the accuracy just isn't there.

The biggest issue we would have is, leaving the CGM under the crib so that it was in-range, and then checking it during the night, many times we would end up with the dreaded '???' in place of a blood sugar reading, and a large gap in the chart. Dexcom blamed the CGM and replaced it twice; the problem kept recurring.

Then, the times when the CGM was not losing its mind or its signal, it would frequently start false alarming -- it does not have a hysteresis algorithm to prevent the alarm from triggering over and over again if you are lingering and constantly crossing over a threshold. As anyone with a young child knows, sleep is a priceless commodity, and unnecessary alarms going off multiple times throughout the night are a total deal breaker.

For a system which was supposed to increase glucose control and peace of mind, the G4, for our 1-year old and us, did neither. Add in the extreme cost ($350/mo for sensors, assuming they all even work) the utility just wasn't there for us. Two G4 Platinums and a few sensors now sit in the back of the closet collecting dust.

We used syringes for almost a year before switching to the OmniPod pump. It is similar physically to the G4 sensor, except it auto-inserts the cannula which helps a lot, although our youngest still cries for a couple minutes after each insertion (every 3 days). A controller acts as the glucose monitor, and also lets you enter in the carbs and it calculates the appropriate bolus. It's a nice system, except the pods also fail early about 20% of the time, they can fall off and leave rashes and puncture wounds, and they waste a huge amount of insulin for children since the minimum capacity for a 3-day pod is nearly 100 units. But it's hard to beat the convenience of dosing for a meal when you're eating out and you just have to push a few buttons.

Re:A DIY diabetes kit by jawngee on Saturday July 16, 2016 @05:17PM (#12107909)
Thanks!

Yes he was DKA on diagnosis. It took them 4 days to figure that out too.

For us the CGM would be about getting rid of and/or minimizing the overnight blood checks. There's nothing more horrifying then sleeping through an alarm now. In fact, that just happened to me. Fell asleep on the couch, woke up two hours past his check, saying "oh fuck" a lot as I run around the house like a headless chicken.

The only CGM we can get out here is the medtronic enlite. Our pediatric endo in Singapore didn't have great things to say about it, which is why we started with the pump. The only pump available to us too is the medtronic series, which, to be honest, has been perfectly fine and man what a game changer. His numbers have really fallen into line since we switched from MDI.

Re:A DIY diabetes kit by zaroth on Saturday July 16, 2016 @06:26PM (#12108113)
I remember well that phase. As you get more comfortable, with or without a CGM you'll start reasonably skipping some of those midnight and 2am checks. You can try dialing back the night-time basal to the point where blood sugar should be naturally somewhat rising overnight. That way your last check of the night is 3 hours after the last bolus, and in theory there should be no more surprises. Unfortunately it doesn't always work out that way, but it's a strategy at least.

The biggest factor with persistent overnight checks is how does your son respond to lows? They will happen, and when they do, does he wake up crying? It's not something you rely on, so much as... I guess, factor into the risk analysis. I know how physically and mentally demanding it is to be up checking every 2 hours through the night, every night. Ultimately you'll find ways/strategies that will work for you to moderate that a bit, and the day will come where you can get a good night's sleep without feeling like you are taking a risk.

You know what I want.... an integrated lancet/meter/strip with just the right amount of light to make checking a sleeping baby in the middle of the night less of a balancing act... Balance iPhone against pillow to illuminate finger, position lancet and fire, squeeze the little finger to see if you aimed right, try to catch the blood drop on the strip, and try to get it all done before the meter times out and turns off. Man, the number of times I've gotten the blood to the strip just as the meter turned off (which sends me back out of the room to fetch another strip and start the whole process over again...)!!

Wish you best of luck, and it does get (somewhat) easier over time!

Re:A DIY diabetes kit by superdog683 on Saturday July 16, 2016 @06:13PM (#12108082)
thru routine should be able to get sugar stable over night.

i imagine you are trying to keep your child textbook perfect as far as glucose levels. this is close to impossible for a child this young. in general you can keep a diabetic around 100-150 as long as they are not above 150 for extended periods of time it avoids the negative issues associated with diabetes. always safer to run a little high verse low. you could turn down the insulin pump at night until child around 125 and chance of low sugar would decrease. when child gets older can strive for "perfect" control. if diabetic 150 or below and 180 or below at meals most of the long term negative health issues are avoided. 75-95 and 120 meals is the text book numbers but difficult to maintain. good luck when he gets older dealing with the candy and soda issue. one of the very difficult times with a type 1 diabetic explaining and convincing the child they cant have candy like the other kids. also when he gets older avoid all the diet foods and drinks, esp. nutrasweet. stevia might be ok we dont know yet for sure. nutrasweet is bad bad bad for everyone especially diabetics.

Re:A DIY diabetes kit by superdog683 on Saturday July 16, 2016 @05:51PM (#12108020)
you mention some good issues and concerns.

technology is very helpful but for the most part it can not perform any functions 100%.

step one for type one diabetic is routine and diet. yes this is boring but very important. a good solid start in diabetic training would take 100 hours. that is to start and with a good trainer and a good patient. as a pharmacist i like to stay with the old technology for now. insulin pump for basil rate. or use insulin pens long acting agent for basil rate and a short acting for meals. test ye old finger prick and glucose meter. i focus on routine to get consistent glucose levels. develope 10 standard meals to start. 10 brk. 10 lunch. 10 dinner. understanding the glycemic index is key to consistent glucose levels. have to understand carbs and how they are absorbed and avoid high glycemic items b/c those will spike and then drop glucose levels. now this child is very young but older diabetics, alcohol can wreak havoc on glucose levels. next want about 1/3 1/3 1/3 carbs/protein/fat and fiber. example 8oz buckwheat/80z salmon/ 8oz broccoli seems simple and we all talk about a balanced diet but the reason its important is the carbs/protein/fat all get absorbed and metabolised at different rates so this gives the body a sustained level of calories that are digested over a period of hours. there are a 100 important factors i could talk about but the glycemic index and 1/3 ratio and dont exceed 600-800 cal per meal might help some people get started.

just as an aside i have had patients that were motivated, educated diabetics and after a year or two of working with them they could go a week without testing and stay in the 100-150 range. diet, routine, and understanding how and what your body metabolizes are very important.

Re:A DIY diabetes kit by superdog683 on Saturday July 16, 2016 @06:47PM (#12108169)
i agree, you are doing the impossible.

thank you for the feed back, its how i learn.

yes, 2/3 of that post is for the "average" diabetic that maybe be reading this thread, you are are already an advanced diabetic. just keeping the child alive means you are very vigilant. i have not treated a diabetic under 5 years old. i could learn more from you than you can learn from me.

so only advice i have that may be helpful for you is trust and expand on your expertise. learn the glycemic index. others md and me will be wrong in dosing "your" child. i accept that the patient or parent knows their body and listen accordingly. many medical people will not! when you know your right trust your expertise. and finally, yes the 2 issues you brought up are difficult and unexplained. fever or illness can cause glucose to be erratic. and the low bedtime could be from residual insulin in the tissue. this still wouldnt explain all of it but some of the drop could be from insulin in the tissue, can take 1-3 hours to clear. from the numbers you gave your child has some what erratic glucose numbers. the meal issue is very difficult, maybe you can use the pump just for basil and dose the the child after they eat with humalog. dosing a child before meals is impossible and humalog only takes about 15min.

my post is mainly about how the new technology has limits and using individual dose at or after meals for some diabetics will help.

in a more extreme environment ie if glucose and eating are extremely variable you could wait 30-60 min after meal and dose according to glucose reading this is usually easier than dosing ahead counting carbs etc. this is not a desirable long term dosing method. i would only use this when "forced" to by very erratic eating and glucose readings.

Re:A DIY diabetes kit by zaroth on Saturday July 16, 2016 @06:13PM (#12108080)
This is all completely valid and useful advice, but just not really for infants or young children. You can weigh out to the gram every carb on the plate, and dose 15 minutes before eating, but you never know how much they'll actually eat. So then there's things like planning the "backup food" you are going to bribe them with to cover the carbs they've already been dosed for. Similarly, you can pack a lunch and instruct the nurse on how to dose it, but unless they're bringing their empty plate back to the school nurse for inspection, it's all just guessing and hoping.

Clearly there are different phases to the disease. There is no amount of testing, diet, and routine that will keep my 4 year old within 100-150 range for 24 hours straight, let alone a week. Yesterday I saw him fall from 423 to 72 in 45 minutes with no bolus and only .25 units IOB. Activity level alone doesn't explain it, and the doctors even at Joslin Diabetes Center are not particular great.

For example, a couple months back my son was having persistent lows even after what we were certain were accurate tests and dosing. We called to discuss adjusting carb ratios, and they tried to tell us to go from 1 unit for 30grams to 1 unit for 25 grams. It took 30 minutes on the phone to explain they were telling us to go the wrong way. We ended up going all the way up to 1 unit for 60 grams to stop the lows that week, and then gradually brought it back down to 1 unit for 30 over the next month. Was it a virus? Growth spurt? No idea...

Another common occurrence, testing before dinner and being in range, dosing 18 carbs for a meal of chicken, broccoli, and peas, testing before bed and seeing slightly above range but with IOB (insulin on board, meaning he's still burning through insulin that was previously dosed) and the calculations saying he should end up near a 150 target. Then test again 2 hours later (e.g. 10pm) when he's at 0 IOB, and some days he's at 420, other days 60. Same meal, same portion sizes, etc.

We've also had "fun" watching blood sugar plummet in the middle of the night even with 0 IOB and the basal fully suspended. How can blood sugar fall below 60 while he's sleeping with zero basal insulin AND zero bolus insulin in his system whatsoever? Haven't heard a good answer to that one yet...

The point is, T1D for kids means constant vigilance, make no assumptions, don't trust numbers from a doctor you don't understand and agree with, and be ready to question and adjust expectations month-to-month.

Re:A DIY diabetes kit by tnc-71 on Saturday July 16, 2016 @05:38PM (#12107974)
Not sure if you know this or not, but Dexcom will replace sensors that do not last the full 7-days. Whenever I have one that falls off, I call customer service and they sent out a replacement.
Re:A DIY diabetes kit by zaroth on Saturday July 16, 2016 @05:52PM (#12108021)
Yes, I should have mentioned it. We had a sensor pile. For a while they had us send them back (or was that just the OmniPods?) and eventually they started just shipping replacements whenever we bothered calling.

I had a particularly bad experience with the G4 CGM. I know that some people do actually really like the device. I don't know if it's that my kids are stick-thin and consequently don't have many good insertion points, so the sensor would more likely than not be in muscle tissue, or any number of other possible reasons... But I spent much more time cursing Dexcom's name and holding back the desire to throw the damn thing across the room than actually reading useful data.