A concerned parent with expertise in data analysis, creates an electronic solution for type one diabetes management
I’m Simon Carter, and I’m here to tell you about my journey with diabetes, which led to the development of the PredictBGL app.
My journey started when I was 16 years old, when I was diagnosed with type one diabetes.
I didn’t do anything wrong to get diabetes. I didn’t eat the wrong foods. It’s not something I had any control over.
When my daughter was born, at age two she was diagnosed with diabetes. This was the last thing on my wife and I had expected. We had been living healthily, and doing all the right things, growing our own vegetables. It didn’t make any difference.
My wife cried for three months because she was looking ahead to her to daughter having her own children and living her own life, she thought how difficult it was going to be for her.
Our daughter was otherwise the most perfect, adorable child that you can possibly imagine.
So we were very concerned about how we could make her life better, and reduce the impact diabetes would have on her life.
Kindergarten staff were not trained to handle type one diabetes
When my daughter started kindergarten, we would sometimes drive there to give her an extra dose of insulin, because the people at kindergarten were not really trained for handling diabetes.
A year later or two years later, when my daughter started at school, we had to write a plan for her teachers, on how to deal with her diabetes. That was again, really, difficult. Because a diabetes plan can’t possibly take into account all the things that you know as a parent.
Yet you’re expected to summarize it into an a four page booklet, and then give it to the school, as though that’s the Bible of how to treat your child.
Of course it leaves out about ten pages of information, about what you learn as a parent of a child with diabetes, for example the timing between insulin and diet and exercise.
“Type 1 diabetes represents around 10 percent of all cases of diabetes and is one of the most common chronic childhood conditions.” Diabetes Australia
We even had a meeting with the school principle, where we were told that if our daughter arrived at school with a high blood sugar, that she would not be allowed to continue at school. And apparently the high blood sugar was our fault.
It just reflected that a complete absence of knowledge, about how diabetes is managed.
They had no clue that a high blood sugar, is not something that we can necessarily control.
Simon started keeping a close record of his daughter’s blood sugars and insulin doses
Following out of this, we had a log book to record her blood sugar results during the day that would come home with her on Friday. And we would try to decipher this, to see what had happened during the week. We tried to work out, like which teacher did what, and why did they do that?
So for example, if there was a school assembly that delayed her morning tea, and she had a low blood sugar level, the teacher would give her jelly beans for that, which was fine. But then they would also give her the school morning tea.
The effect of having double the amount of carbohydrate would obviously gave her a high blood sugar.
So we had to look at all these things and try to train the teachers, to understand, what was required.
I started thinking about how we could make things easier, so all the teachers could have the same view of her data.
Simon creates his first website to log his daughter’s blood glucose levels and insulin doses
So what I did is, I created a website, where the teachers could log all of her data. What her blood sugars levels were, what her carbohydrate intake was, and anything else that might apply.
It was really important what my daughters blood sugars were, as she moved from the class room to sport. It was very important to know when her blood sugars were before she got in the pool, as well as after she got out of the pool.
For example if her blood sugars started at 10 before she got into the pool, because when she got out the pool her blood sugars might drop to 4.
Because if her blood sugars had dropped to less than 4, that when you start getting problems, from dizziness leading into unconsciousness and eventually death if untreated.
So from this web-based logging system, we were really able improve her time at school.
School teachers refused to even perform blood sugar tests
At her school we still had teachers who were very archaic about their management. The school actually employed an extra helper to perform the blood sugar tests because her sports teacher refuse to perform them.
Even though she had a duty of care to look after my child, she refused to do what was required.
“A UK study showed that people who kept their blood glucose levels as close to normal as safely possible with intensive diabetes treatment, reduced diabetic eye disease by 76 percent, diabetic kidney disease by 50 percent and diabetic nerve disease by 60 percent.”
Blood Glucose Control Studies for Type 1 Diabetes: DCCT and EDIC
One of the school nurses was rude to his daughter when she needed insulin injections
And we also had issues with one of the nurses at the school.
My daughter would go to the nursing station for her injections, and this nurse would treat my daughter as an inconvenience. She would be quite rude to her about her diabetes and make out that it was somehow her fault for being a diabetic.
So we asked my daughter, how would you feel if you injected yourself with insulin? Would that be better?
And she straight away said –
“Yes I’d rather do my injections myself, than go to see the nurse at the nurse station.”
Simon adds an insulin calculation system to his website
So in addition to the logging system, I added a dose calculation mechanism.
So if her blood sugar was high at lunchtime the nurse or my daughter could correct down for that, and make sure her blood sugar was totally normal by the end of the school day.
As time went on this system got better and better, and I thought, if this system is working for us, then we should be offering it to other people.
Simon’s website becomes open to other parents of children with type one diabetes to use
So we created a website that was open for other people, with logins and and other facilities.
I was looking at the data; and we had blood sugars and carbohydrates and we had insulin. I thought surely we must be able to make some future predictions. Maybe not accurately, but a hell of a lot better than having no data at all. What we were trying to do understand, was where the blood sugars were going in the future.
He does research into predicting future blood glucose levels
So I did a lot of research into insulin models and also carbohydrate models. I found there were no carbohydrate models.
There were some insulin models for one of or two of the newer insulins. But there was nothing for the older insulins.
When it first produced the results, and they were lined up on the screen, it predicted I’d have a low blood sugar level about 2 hours after this current time. So this was 9 o’clock and it predicted I’d have a low blood sugar level at 11 o’clock.
And as it turned out, I went out shopping, and I started feeling a blood sugar, and I looked at my watch and went okay. So it’s five past 11.00pm.
“It is somewhat extraordinary, and it proves the point, that you can pretty much predict future blood sugars ahead of time.”
Simon’s daughter’s life was saved by this future prediction algorithm
So then we tried this system on my daughter.
About a week later I was looking at my daughter’s blood sugars before she went into the school pool, and saw that she would have a low blood sugar level while in the pool, based on the software’s predictions.
So I rang up the school and I spoke to the secretary; she actually has a husband with type one diabetes. I explained the situation and she raced up to the pool and they dragged my daughter out and did a blood sugar test. It was 1.7 which is exceptionally low.
This means that there would have been no warning, she would have just slipped under the water and the teacher might not have even noticed.
So having this system has given us an amazing peace of mind, to know what blood sugars are going to do.
I can’t explain to you the level of confidence that gives to us, knowing that our daughter will be safe in her sleep. It stops us getting up at 2:00 AM and 3:00 AM and 4:00 AM and 5:00 AM like some parents do, to measure her blood sugars. And that’s what we used to do.
Whereas now we look at her blood sugar around 10:00 at night, and look at the charts and we go, okay, that looks like that is fine. Then we can go to bed without the stress of wondering whether our child will be alive in the morning.
I can’t convey to you what that means.
The syndrome of sudden unexpected death or “dead in bed” syndrome is considered to account for around 5% of deaths in type 1 diabetes – Diapedia.
So that’s where the system has come from. That how it was built and why it was built.
We tried to translate this same model to other people with diabetes, and other people found that this model worked for them.
And I’m telling you, this is amazing, because you asked an average doctor and they’ll tell you you can’t predict blood sugars.
And I’m telling you that you can.
We get a lot of resistance from people.
Some Doctors and Australian bureaucrates refuse to accept Simon’s electronic system
So for example, my daughter went to a diabetes camp recently and all the dietitians, educators and the doctors have been worded up that we were the people using the app, we were the pariahs, that were not using an insulin pump and two injections a day like everyone else.
We were, you know, heaven forbid, using a smartphone app to manage diabetes and not just any app, but one that predicts future blood sugar levels.
And we were basically told that we couldn’t use this app on the camp. And also, the doctors changed my daughters insulin doses quite dramatically. The result was that she had massively high blood sugars for the entire four days of the camp. The Doctors ended up quite sheepish, at the end of the camp, and admitted, they were wrong.
And then our reply was like, why did you do that? Why did you not follow the information provided by this app?
We get a lot of resistance from the medical profession, who can’t seem to comprehend that the people doing things differently can result in an improvement on what they’ve been doing for 30 years.
PredictBGL is now a global app
At the moment, the app is being used by about 7,000 people around the world. But the next thing we want to do is take live data from their blood glucose meters and live data from their sensors to go straight into our app. So there is no need for double entry and they are not typing in data all the time.
We are also trying to make it a lot easier for people to get better food information from a food database. So we’re looking at a 2 million entry food database which is customized for people in Australia. So people can understand exactly, not just the carbohydrates in the food but also the fat and protein. Because they also play a role with with the prediction of future blood sugars.
So by having this information, and making it easier to use, will help make the predictions more accurate in the future.
We’re looking to go to markets like France and Germany and even Russia.
The amount of diabetes care is very limited in Russia compared to Western countries.
And we really provide them with a system which is, better than world class. For example, in Russia, patients only have funding for the two blood test strips a day. Whereas I’ll use about seven to eight. So they have to be very strategic about how they manage their blood sugars.
They’ll do that perhaps one in the morning and one in the evening. I have a customer in the Ukraine who has been using the app and his predictions are astounding, unbelievably good. He puts all the data carefully and his blood sugars predictions are incredibly accurate.
Simon has forced the medical industry to move away from managing blood glucose levels on an ad hoc basis, and towards a predictive model
This is beyond my expectations of how good a system like this could be. I want to make this the system the centre of how diabetes is treated. Because it’s doing a far better job than any current system of how diabetes is treated.
And really, it’s changing the way doctors look at diabetic data. It’s forcing them to realize that just looking at blood sugars on a chart when the patient comes to see them once every six months is not enough, and has never been enough.
In reality, the patient needs to see a visualization of blood sugars, not just where blood sugars have been, but where they’re going, hours ahead of time. So they can actually plan their lives and prevent low blood sugars from happening.
So yeah. Thank you for listening and I hope it’s been of interest to you.”
© Wikihospitals July 2016