A DIY diabetes kit
bbc.com
bbc.com
Unfortunately, we live in Vietnam where access to even basics like pumps, let alone CGM, is non-existant, or incredibly hard. We had to fly to Singapore to get a pump, and then fly back every three months to get new supplies. CGM is a pipe dream.
One thing this article doesn't touch on, which is unfortunate, is how much of a nightmare modern medical systems make diabetes. In the US, insurers decide what pumps you get, what CGM device you might be able to get, even what insulin you are using. I see it everyday on the support groups I'm on. And the costs are astronomical. In Vietnam, we get our son's insulin for $12 a vial. The same vial in the US is about $400 or more (insured price). Test strips in the US can cost $1-2 per strip; we spend about 20-30 cents per.
I am very hopeful about the bionic pancreas though, a device that does what these DIY systems do; but also delivers glucagon as well as insulin. There are also some stem cell trials that are very interesting too. I'm hoping that by the time this kid hits college, his care is automated as possible, if not outright cured.
http://www.npr.org/sections/health-shots/2015/12/14/45904732...
Pricing info isn't super apparent, I guess it probably isn't a whole lot more than the $25 a vial suggested in various articles and comments about it.
Please elaborate...
Modern diabetes management includes either pumping or multiple daily injections (MDI). With MDI, you use a long acting insulin as a basal, or background, insulin. And then you'll use a rapid insulin to bolus for meals and to make corrections for unexpected spikes.
The pump is similar to MDI but you only use rapid insulin and the pump delivers a steady basal stream throughout the day, and you still bolus at meals and for unexpected spikes.
These two things allow diabetics to eat whatever they want.
We did the old school way for the first few weeks because that's the only way the doctors in Vietnam knew how to do it. It really sucked numbers wise.
Going further, the calculations are rule of thumb and generic and each person has to make adjustments based on intuitions about their metabolism, what they ate, what they plan to eat, what activity they anticipate, and when they last dosed. It's every day, for the rest of their life.
[Source]
Scott Hanselman interviews OpenAPS hacker Dana Lewis:
http://www.hanselminutes.com/514/the-open-artificial-pancrea...
It is exceedingly easy to overdose on tylenol, and yet it's available without any controls.
Sex-negative has nothing to do with it. Plan B is non-prescription.
The question is, does something about visiting a doctor to get a prescription for birth control mitigate the risk of a blood clot? Something that there would be no way to replicate with say better women's health education or a nice info-graphic next to the birth control section?
Besides, insulin is available OTC anyway (in the US). It's just that prescriptions make it cheaper/free.
We (EMS) give glucagon to hypoglycemic diabetic patients when we are unable to obtain IV access. I am not a doctor, but we almost always prefer to give them D50 via IV for diabetic emergencies in the field. Most of the time, they don't need to go into the ER after such intervention. As a last resort we give glucagon IM and wait, and wait... more often than not, using glucagon requires a trip to the ER. Bottom line, glucagon and the resulting liver release of glycogen doesn't seem like a healthy way to regulate blood/glucose levels long-term.
Again, I am not a doctor... But as I understand the process, your liver needs time to build up glycogen stores. If you start hammering on the liver to deliver, you may trigger other unintended consequences. It just doesn't seem like a good idea from my admittedly naive, and emergency medical point of view.
But using glycogen stores isn't necessarily bad for you metabolically. Glucagon is used by your body everyday during periods of fasting between meals to maintain a minimum blood sugar. Your glycogen stores can supply enough glucose for several hours of fasting before you turn to other routes of metabolism, which would next preferentially be fatty acids. Glucagon IM injectors are given to lots of diabetics to prevent hypoglycemic emergency if drinking fruit juice or something isn't enough.
I happen to live where a very, very large cluster of diabetics are studied. We do a lot of diabetic emergencies... I can count on my hands how many times we've used glucagon. Like I said, we can treat and release with D50. Not so with Glucagon.
If "they" figure out the delicate balance between all the systems in order to use glucagon the same way it is regulated in a non-diabetic; it stands to reason that "they" should have a way to repair/replace the islets of Langerhans as the first priority, don't you think?
The constant fear of trying to catch if you're low is exhausting. Especially when a diabetic like me goes through sad phases, can't get enough sleep etc. The main symptom of being low is dizziness, and if you're blood sugar levels are not in control you always feel that diziness.
I've had panic attacks, thinking I'm low when I'm not, got really depressed, my blood sugar levels were like a roller coaster.
This year I managed to get a grip of my life, right now I'm on vacation and relatively happy, I noticed that I get by with much less insulin. Stress definitely has an effect on diabetes management. During exam weeks for example, it's nearly impossible for me to keep my levels under control.
A CGM could end all this suffering, but the cost of having one is very high. The insurance, at least here in Turkey, covers only 3 strips a day of the cheapest brand available.
I've seen 22 mg/dl at a time, and didn't lose consciousness. Most diabetics can feel when they're hypoglycemic.
PS: These are my own experiences, sorry if it's too verbose.
I've found the best ways to stabilize my glucose are, well, eating lots of protein and fats, eating smaller carb portions in more meals, and smoking cannabis.
22 mg/dl is really scary. The lowest I've found myself at was 30. Talk about scary! Is glycogen cheaply available in Turkey? It's really damn expensive here.
The biotic pancreas uses glucagon to bump you up from lows:
http://diatribe.org/introducing-beta-bionics-bringing-ilet-b...
If I'm looking at the right advice [1], insulin is £1.50-£3 ($2-4) in the UK, so that may be one possible source.
(The UK suppliers I found don't have information on overseas delivery on their sites; it's probably worth a phone call.)
[1] http://www.diabetes.co.uk/insulin-pumps/buying-an-insulin-pu...
The other things (like infusion sets) should be easy to ship. The UK suppliers generally won't sell them unless you have a UK pump and an account with them, though. It could probably be worked around though, using some sort of international forwarding arrangement.
I have type 2, and don't know much about type 1, but I am assuming that the test strips are the same. If so, you can do better than 20-30 cents in the US.
The key is to avoid the big brand name meters, like Contour, FreeStyle, AccuCheck, and such.
Get a meter from True or ReliOn. These are available retail online and at major brick and mortar pharmacies for around $10-15 for a kit that includes the meter, some lancets, and around 10 test strips. Example: [1].
Strips are available retail for around $16-18 per 100 strips. Example: [2].
The big brand name meters are nicer. My Contour just feels higher quality than my True, and it shows the results a little faster, and the lancet device that came with it works a lot better...but the True gets the job done too, and there is no rule that says you have to use the lancet device that came with the meter.
It's easy to overspend in the US on health care. I'm on Metformin. I have insurance. If I get that prescription filled at Rite Aid a 30 day supply is about $90, with the cost to me being a $20 copay.
If I get that prescription filled at WalMart it is $4, and there is no point in bothering with insurance.
[1] https://www.amazon.com/TRUE2go-Blood-Glucose-Starter-Kit/dp/...
[2] https://www.amazon.com/TRUEtest-Test-Strips-100-Count/dp/B00...
> The key is to avoid the big brand name meters,
> Get a meter from True or ReliOn. These are available retail online and at major brick and mortar pharmacies for around $10-15 for a kit that includes the meter, some lancets, and around 10 test strips. Example: [1].
Agree. When I shopped for a meter at my local Kroger pharmacy (King Soopers here) my only decision was how much are the strips. Mine is an OnSync, and the strips are branded Kroger; the in-store pharmacist told me that Kroger worked with a manufacturer to make an affordable setup.
Everyone has their needs and such, but all things being equal, shop for the cheapest strips, and then the meter that takes them.
https://medium.com/@chrishannemann/a-craftsman-blames-his-to...
https://medium.com/@chrishannemann/measure-seventy-five-time...
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.
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.
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.
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!
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.
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.
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.
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.
There are many areas the article does not cover, but this is expected when writing space is limited. Main areas that I feel need more coverage...
1) The community itself, it is overwhelming to see how much work has been done by people in their spare time. Many of these developers go unnamed and provide some amazing work - I would not be able to complete my app without this community, I'm just putting the puzzle of community work together and filling in the blanks where needed.
2) Access to tech, it is almost insulting to see how may Diabetics cannot access Pumps and CGMs, or can access one and not the other. The cost for such systems and limited access is not justified.
3) Why, when we do own the required tech purchased from our own money do we then have to "hack" such devices to access the data that they have about MY body!?
4) The dangers of the community work, this is no riskier I say than commercial systems but this subject must be openly discussed in the community to be sure we understand and limit possible risks
5) I agree with the comments about glucagon, this should ideally be a worse case treatment. Why do we need a system that can provide glucagon when early alerts from an APS can notify the patient that a carb treatment is needed. Let's stop trying to find a 100% solution and produce something that covers 60% of our needs now
My wife's been on the software since last September, and it's made a huge difference to us.
If you've any questions, ask here or grab me on twitter at @oskarpearson
I'm not sure where you are, but in the USA compatible pumps probably start at around $1,000. That doesn't include the monthly cost of the CGMS sensors, which is also significant.
If you're in a country with universal healthcare, and the healthcare provider is willing to cover the CGMS sensors you might have success. (Some govt healthcare services will cover CGMS sensors for people that can't tell when they are hypoglycemic.)
If the person with diabetes (PWD) is very mobile, reduced size and weight will have to be at the forefront of your thinking. However, that requires the most expensive OpenAPS hardware (the Intel Edison - see the mmeowlink wiki).
If the PWD is ok with a larger device due to a bigger battery, you could use cheaper hardware (Pi zero and a 20,000mah battery). It'll still need daily charging, though.
One idea: you might find running things overnight only might have an advantage, depending on whether they have a stable relationship with a shelter. The system would then cover them for a reasonable amount of time every day without the challenge of waterproofing, batteries, etc.
If this is feasible, the best place to start is with the simplest config, as documented at https://github.com/openaps/docs/blob/master/docs/docs/walkth...
Let me know if I can be of any support. My details are in my profile. Alternatively, you can find us on the intend-to-bolus channel on gitter.
I use xdrip with a CGM and I have to say, it's been the best thing for my care (after insulin, lol). The main challenge of diabetes is keeping your blood sugar in range, not too high, and not too low, and getting it back in range if it deviates too far. Blood sugars dropping too low during exercise is a major concern as it can result in seizures and worse.
Thanks to the combination of my dexcom G4 + Android + Xdrip + smartwatch, I can always get a pretty good estimate of where my glucose is, and thus minimize the risks and harm from my (chronic, uncurable) condition. Definitely a great case for open source development for propreitary hardware. The CGM has helped me even more than my $10,000 insulin pump :X
Where it gets tricky is with in-patient care. There I would agree that it's more like representing yourself in court.
insur. depends on your employer and 100 other factors. some plans very little cost others maybe 25%. if you are independently employed insur will run about $25,000 to 35,000 per year for the family.
if your son is "healthy" and is a "good" diabetic he can get by on about $6,000/year. this would be if he eats well, moderately active and maintains sugar well. fyi if athletic this doubles or triples amount of insulin consumed. if your son takes very poor care of himself very erratic sugar 100-350 and 5-8 extra trips to the ER and or doctor and you want the top end insulin pump, humalog, humalog pens etc then the cost could be as high as your imagination.
Or if you are buying your own plan through the ACA, the cost will depend primarily on your family income. The unsubsidized cost depends on zip code, sex, age, and whether you smoke tobacco, but assume $300 per month per adult, $150 per month per child. The subsidies which reduce the cost of premiums and also out-of-pocket costs, they start at effectively 100% if you earn $0, and phase out as your family income goes up to about $80,000.
Once you've paid the premiums, then you start paying for the care you use in terms of deductibles, co-pays, co-insurance, up to some annual out-of-pocket maximum. The details depend on the plan, but the basic rule is you can pay more premiums and your out-of-pocket maximums will decrease. But, doing the math, you can rarely "save" month by paying more in premiums for lower maximums, even if you know you will max it out, since insurers know that it's mostly sick people who will opt for the higher premiums in the first place. But it is convenient to only have to send one check a month, instead of constantly writing checks for every script, appointment, etc.
If your co-pay is, let's say, $25, expect to be paying that amount ~15x every 3 months for T1D. Everything is it's own script (blood strips, ketone strips, needles, glycogon, insulin, lancets, etc.) so you get hit with a lot of individual $25 co-pays, although they are not all being filled every month.
If you end up having to go to the ER after a particularly bad low, expect that could max out your out-of-pocket expenses for the rest of the year. So keep an eye on that annual out-of-pocket maximum, which typically will be $5,000 - $7,500. After you hit that maximum, all other benefits are paid 100% by insurance. For a particularly bad year, it helps to note, the family maximum annual out-of-pocket will be 2x the individual maximum.
Another factor is if you want to use the latest tech (CGM, Pods) your insurance likely will cover it, but not always, and there's probably a 20% "durable equipment" co-pay (which counts toward your maximum). This can increase costs for a year you are buying new expensive equipment to get started.
If you are getting insurance through work, and you lose your job, you are allowed to continue paying the full share of the premiums and keep your insurance, through a program called COBRA. That's worth it especially if you've maxed out your out-of-pocket and there are a few months left in the year where you can get 100% coverage. Otherwise, it's usually cheaper to switch to an ACA plan.
Bottom line, fully un-subsidized insurance costs for a family actually using the benefits (paying most if not all of the out-of-pocket maximums) are between $20k - $30k per year. Most Americans don't realize this because the ~$15k premiums are part of their pay package but don't really show up clearly on the pay stub.
Of course any analysis which only looks at personal spending on premiums and out-of-pocket expenses without also looking at tax rate and the percent of those taxes which also went to healthcare costs is, I think, fundamentally flawed. For example, "Healthcare should be free for everyone!" †To be funded by a new 10% income tax.
Finally, without also adjusting for public-sector healthcare R&D costs as well as cost shifting to countries more willing to subsidize private R&D, then... well, what exactly are we even comparing? ("Healthcare here is so cheap, we pay generic prices for all the latest wonder-drugs! So glad we benefit from all this great R&D everyone else is funding!")
In short, your data point tells me nothing about what people in NL are really "spending" on healthcare, as well as to what extent the "lower" costs of healthcare in NL could be a result of freeloading on other nation's R&D.
Consider applying for medicare or similar?
One thing that's been really awesome to see is how T1D patients and parents with little or no experience with FOSS have taken this stuff: "You mean you guys just made this amazing thing 100% for free?!?!?!". It's like watching the birth of open source all over again :-)
The T1D hacking movement is huge - for example, my GF and I recently replaced the batteries in her CGM transmitter (you're not supposed to be able to do this, so it involved a dremel and solder and playing the tiny fingers game). The older transmitters last about 6 months, and then you have to order them through a horrible medical supply company that knows you're not their real customer (your insurance company and device manufacturers are) so this is a big deal.
An even bigger deal: the newer Dexcom transmitters turn off after 3 months. Dexcom claims that new features mean they need more juice, but T1D hackers reversing the new protocol discovered that the app they come with turns them off. Fortunately, it seems possible to turn them back on without dremels and the tiny finger game.
A closed loop APS is the next step for us :-)
I hope his work can help many people who are in need of this type of tool.