When attending the Diabetes Professional Care conference, I noticed a session on the agenda titled “Triangle of Diabetes Care and the Ambulatory Glucose Profile”, by Pratik Choudhary, which is something about which I had never heard before. So I started to do some digging. Unfortunately I was unable to attend the session. If anyone went, I’d be interested to hear your thoughts!
In summary, based on the training that I have watched, it is, in my opinion, a reasonable approach for healthcare professionals to take with a very clear link to Abbott Diabetes Care and an attempt to sell the Abbott Freestyle Libre. Many of the principles being talked about make a lot of sense and it is refreshing to see them being put forward. I think it promotes a lot of good behaviours that perhaps have been understated in Diabetic care in the past.
But what is it? From the MIMS Learning site:
“The Triangle of Diabetes Care is a treatment strategy that is centred on three goals for diabetes care, is an approach that is applicable to all healthcare professionals tasked with treating patients with diabetes and can be delivered in all settings, from hospital clinics to primary care within the community.”
Given this as the summary of it, I decided that it was worth digging further.
On a Google search, the top result was a blog entry from Abbott. Now that’s not to say that there is anything wrong with it, and it might be a very valid approach, but if a pharma company is pushing it, it often sets alarm bells ringing. So let’s take a look shall we?
In visual terms, MIMS have this picture:
Again, this doesn’t seem like a bad idea, but looks very much like, well, what we proactive T1Ds do as a part of everyday life.
So taking a step back, where did this “Triangle” come from? Well it appears that Dr Ramzi Ajjan, an associate professor and consultant in Diabetes and Endocrinology at Leeds teaching hospital came up with it. According to his Bio on the Abbott website, it is something that is relatively recent.
Okay then Dr Ajjan, what does this mean in real life terms for Diabetics? Based on the first of five presentations available on the MIMS website, it can be boiled down to the following three points:
- Extended hyperglycaemia increases risk of complications
- High oscillation of glucose levels increases oxidative stress on cells (amongst other reactions), leading to greater risk of complications (think retinopathy and some aspects of CVD especially)
- Avoiding too low an Hba1C reduces the risk of CVD related death (but studied in Type 2 diabetics specifically); given that increased hypoglycaemia is correlated with increased death by CVD in the patient set studied, avoidance of hypoglycaemia is recommended*
It is good to see that monitoring is being recommended for both Type 1 and Type 2 diabetics. How are type 2s ever supposed to deal with their condition if they aren’t encouraged and supported in observing the behaviour of carbs on their bodies? As part of this, recommending more testing is also a good thing, and something that many T1s already do precisely to observe patterns and understand behavioural and therapy impact. We are the motivated few. There are probably many more who don’t though, and getting this message across is the important part.
What is also clear though, with the inclusion of the line about FGM and the sponsorship by Abbott, is that this is also a sales pitch, all be it a well intentioned one, for the Abbott technology. I’m not against this per se, as a lot of the data is something that I as the expert in my personal condition have already read and I think that most of it is relevant. It’s also a good way to educate HCPs about the technology and why it’s beneficial. I can back that up with a blog post I recently wrote. I’d like to see it stating very clearly that with these types of technology, you can embark on Dynamic Diabetes management or Sugar Surfing.
Tying in to this is the recommendation for use of the Ambulatory Glucose Profile, which is a chart that looks like this:
This is my Ambulatory Glucose Profile, as provided by Diasend, from data provided by my Libre. It shows Median, 25%-75% band, 10%-90% band and 0%-100% bands. This is something relatively recent to Diasend, as it used to only provide a box chart, as below:
Unfortunately, the data in these things is only as good as the data collection device, and as I’ve documented elsewhere, the issues I’ve had with sensors do mean that some of this data is rubbish.
What’s perhaps more useful about the AGP graph is that it allows you to spot patterns in time when you know that there are things that are not quite as they should be. For me it’s the mornings and evenings where I get the greatest variability, and the benefit of such a trace is that it gives you the opportunity to review when and where you are struggling, but perhaps more importantly to review what you are doing at these points in time and see what adjustments you might make. I know, for example, that my mornings are variable based on DP timing (as per this blog post) and as much as changing basal profiles on my pump might assist, there is a large amount of variability day by day and it requires a lot more pattern determination.
Anyone who has used CGM or the Freestyle Libre has had access to this level of data for a while, and many of us find it incredibly useful, but as I have mentioned before, it’s useful as part of an overall strategy, but not necessarily in and of its own right. You need to be involved in your own management to see benefits and those people using these systems generally are as we are having to pay for it ourselves.
The one down side of the AGP tools that are currently available is that they don’t yet allow you to select specific days on which you undertook similar behaviours and review the patterns of those days. For example, I’d like to compare the three days I lift weights every week and see what the AGP looks like, versus the days I don’t. Based on basic review of the data, I already have a feel for this, but seeing it in this fashion would be really helpful. Diasend, are you listening?
Likewise, looking at the days I know I ate a CHO rich diet and those where I ate a low CHO diet would be great to demonstrate the difference to HCPs. So basically, in my view, having HCPs learn how to read these graphs and understand the benefits of them can only be considered to be a good thing.
The proactive amongst us already share this data with our peers on Facebook and various internet media. It enables group review and advice.
That’s fine for those of us who are pro-active in our care, but what of those who aren’t? What would be even better would be that for those patients that struggle, HCPs receiving this data and making a pro-active call to provide feedback/recommendations once a month (and not every six or twelve months) to nudge their behaviour.
In summary, I think that the triangle of Diabetes care is a bit like teaching expert patients to suck eggs. But there aren’t very many expert patients. Every day on the diabetes.co.uk forum we see both Type 1s and Type 2s who don’t understand that they need to understand the patterns in their blood glucose relative to food and to therapy, and everyday we talk people through how to do this. Having it introduced at an early stage in diagnosis with information on how to obtain and read the data would be tremendously beneficial. But the key factor is the education on what it is and how to use it with a DAFNE like approach is critical. It looks to me like a step in the right direction, and that direction is dynamic diabetes management, or Sugar Surfing as we like to call it.
It astounds me that most of what is being described here is new to many HCPs. This series of webcasts was published in July 2015. CGM users have been doing it for 10 years. Libre users for a year. Is there really such a large gap between the experience of switched on patients and UK healthcare? As I’ve opined before, we are here. We want to participate. We know what we are talking about. Use us!
Appendix: Triangle of Diabetes Care – Learning Outcomes
*As anyone with a good grasp of statistics is well aware, correlation does not imply causation. I therefore am making the assumption that there is more than just a correlation driving this recommendation.