Latest guest blog is from Professor Kath Barnard @DrKathBarnard
It has been an exciting year for diabetes related technology, particularly for those with type 1 diabetes. There have been widely reported leaps forward in the development of the artificial pancreas with research teams around the world conducting clinical trials outside of hospital and laboratory settings. Summer camps and ‘at home’ trials are ongoing around the world with some very positive results thus far.
New self-monitoring of blood glucose devices are becoming available which add explanation and context to the numbers presented on standard meters. Such devices present an evolutionary development away from simple ‘snapshot’ finger prick testing, to support systems for users and furthermore to flash monitoring systems. This is where retrospective trend data can provide more detailed information about patterns of blood glucose and help the user fine-tune their basal and bolus insulin delivery to best match their needs.
So, it’s all good one would think on the technologies front and before too long we should be able to report significantly improved outcomes for people living with type 1 diabetes using this advanced technology. Or perhaps it’s a little early for such optimism.
Where is the evidence-base? How do we actually know that new, ‘whizzier’, faster and more detailed are actually beneficial? How do we know whether there are any specific groups of people with type 1 diabetes who would benefit most from these from these devices and, if there are any for whom the potential for additional burden and increased anxiety associated with these devices could result in unintended detrimental health outcomes?
This is the thing with evidence-based medicine, as practiced in the NHS. It very much relies on a critical appraisal of the evidence-base, and for these new devices this currently seems to be lacking somewhat. How then are we able to assess the clinical, psychosocial and cost effectiveness of these new devices? Taking this argument a step further, what outcomes should be we evaluating in such a critical appraisal? By knowing this, we can then design the clinical trials required to provide the evidence-base that we require.
There is no doubt that glycaemic control, as assessed by HbA1c is a key outcome. As much as it may pain me as a psychologist, it is difficult to get around the argument that with even the best psychological outcomes, sub-optimal glycaemic control remains the gold standard marker of risk of long-term diabetes related complications. I would obviously counter that argument with the overwhelming evidence that improved psychosocial and psychological functioning are associated with improved glycaemic control and therefore are of equal importance but this does not always wash with my medical colleagues.
So let’s look at technologies currently available to support diabetes management for people with type 1 diabetes. There is undoubtedly engineering excellence associated with these devices and on paper they should support optimal glycaemic control for users ….. so why is it that so many users still have sub-optimal glycaemic control? That’s where I come in. The simple answer that I always give is ‘because life gets in the way’. On paper, type 1 diabetes seems a ‘not too challenging’ mathematical problem, particularly with the devices available to facilitate optimal self-management. In reality however, it could not be further from that. Everything impacts on diabetes control, everything. Even a maths genius would be hard-pushed to achieve perfect glycaemic control (whatever that is!) without sacrificing other aspects of their psychosocial well-being. Whilst this may be acceptable in the short term, diabetes is not a short-term condition and for even the most persistent personality, it will eventually become onerous and unreasonable.
Insulin pump therapy for example is associated with numerous quality of life benefits such as increased independence, fewer injections, greater food related freedom and improved psychosocial functioning, as well as biomedical benefits such as reduced frequency and severity of hypoglycaemia. Yet there are also a number of downsides including constant attachment to the device, potential increased visibility of disease state, cost and risk of occlusion and pump failure. Continuous glucose monitoring devices are associated with improvements in glucose control, however evidence shows that this is dependent upon regular use. Psychosocial benefits include increased confidence about diabetes control, ability to identify trends and act to remedy patterns of hypo- or hyperglycaemia, and reassurance for parents whilst their child is at school. Alarm fatigue (particularly false alarms), technical failure and accuracy problems are limiting factors to ongoing engagement, with lack of trust in the devices and irritation with technological failure cited as key reasons for cessation of use. These factors may contribute to the low uptake of CGM use at around 6% at best.
So, the new self-monitoring of blood glucose meters have a lot to live up to. There is no doubt that SMBG is a crucial aspect of diabetes self-management for people with type 1 diabetes and many people with type 2 diabetes, however previous research shows that SMBG is associated increased anxiety and depression for some users. How do we know that this will not be the case with the new SMBG systems and flash monitoring devices? The ability to simply ‘swipe the device over the sensor’ is clearly very tempting and enables users to frequently check their blood glucose without having to conduct a finger prick test. I wonder whether this ability for more frequent, non-invasive checking rather than engaging in the recommended testing frequency of 4-6 times a day is reassuring or whether it contributes to increased anxiety and worry for users. The truth is that currently we are unable to answer that question. Nor will we be able to do so until robust clinical trials have been conducted to find out.
Whilst I still maintain that the future is very bright for technologies in the support of diabetes self-management, I equally believe that these must fill a real need. They must be evidence-based and theory driven, developed with people with diabetes as equal members of the clinical research team and meet the needs of the end users from an engineering, biomedical and psychosocial perspective. Only when we can genuinely say that we have achieved this goal will a device be truly fit for purpose.
The Grumpy Pumper 
Lots to think about there thank you. I would also say the low uptake on CGM is probably as much down to cost as well. £48-50 a sensor that officially lasts 7 days is out of range of a lot of people cost wise.
I’ve never really thought about the psychological impact of diabetes. Not until I heard that is it 2/3rds end up on anti-depressants? Shocking. Then I realised it was alright to get fed up about type 1.
Thanks again.
Great article. I was about to make the same comment about CGMs – I would have thought the low uptake of CGMs is more to do with prohibitive expense.
It sounds like the documented reasons for PWDs stopping using CGMs relate to the population of PWDs who have already used them under clinical supervision and not the broader population of PWDs. Of the PWDs I know, if they aren’t already self-funding CGMs, cost is the barrier.
Fundamentally I agree that an evidence base is required for correct funding allocation, but without greater CGM use, how will insights into how CGMs best benefit different populations of PWDs be gleaned?
Thank you. I agree, cost is prohibitive and CGM is not currently reimbursed on the NHS so very difficult to access through clinic. Depression 2-3 times more prevalence in people with diabetes (T1 and T2) than in the general population, so around 25-30% of PWD will suffer depression and up to around 45% will suffer sub-clinical depression, low mood and diabetes-related distress. It’s a really tough gig and 100% OK to get fed up with it. Best wishes, Kath.
Agree that the cost is prohibitive for most. Our son, now 14, was diagnosed T1 just after he turned 6. After the initial two then MDI injection regimes for a couple of years, we tried the Abbott Freestyle CGM and were amazed by the analytical data. The CGM only lasted a few days, plus it wasn’t linked to the pump, meaning he had to carry two devices around.
Then Medtronic came out with their combined VEO pump and CGM. Out clinic approved the pump and we paid for the sensors.
We spent the next two years trying to make the CGM work consistently. But with calibration errors and sensors only lasting two days max, plus the cost, plus very painful insertion and if you could get it working, proved too much for us and our son. So we gave it a rest.
Ufortunately even win the pump, his HBA1c has been typically 8+. Not great. But if we hadn’t had the pump it may well have been higher. I was still looking for a reliable, robust CGM.
We tried the Dexcom CGM for 1 week. In the UK, Dexcom have a distribution partner that will loan you (at a significant cost) the Dexcom receiver and sensor. Its ability to provide 5+ days of sensor data reliably without calibration errors gave us what we were looking for. But again, the receiver was a separate device to carry.
I knew the benefit of CGM in providing enough analytical data probably could help us get the HBA1c down and the cost was something we just had to prioritise. We have been lucky we can make that priority decision.
After 4 years on the Medtronic, our pump was due for changing, so we moved to the Animas Vibe since it provides an additional benefit of being waterproof and had a built-in CGM we knew was reliable.
So we are again trying the CGM.
Due to its reliability, and longer sensor life, means we have a CGM solution we can work with.
It’s still a high cost. £186 for 4 sensors. Plus the transmitter costs some £500 and needs changing after 9 months.The lack of salary increases over the years has meant we now need to prioritise even harder. T1 is the highest priority. We are working towards a future benefit of lower HBA1c we can only read about. That’s a priority and benefit you have to believe in. It saps and drains all of us.
So now we have found a device that can give you almost real-time data. But data is only good if it can be analysed and results in a lower HBA1c.
Our son is an active 14 year old. Anyone reading this will know T1 is unpredictable. So just when you have collected a weeks worth of data you think you might be able to use and make “longer term” pump basal or bolus ratio changes – the next week might turn out completely the opposite. Now lets throw in the odd post virus rash, a sleepover (not), scout camp, school bus ride home with the odd Mars bar he “forgot” to bolus for, the school PE session, football training and football match – I’m sure there’s more, but you get the picture.
Ongoing “real-time” analysis is critical to safe pump management.
With our son being 14, he has started studying statistics at school and so is able to understand and recognise the value of the data. We have started looking at the CGM graphs and trying to find trends and figure out that the forgotten mars bar might account for the high 20’s.
Sometimes you can’t see the wood for the trees in the data. You have to stand back and have someone else analyse the numbers and find the patterns and make the basal changes you dare not make for fear of creating the hypo that night that might affect him in so many ways you dare not imagine for too long. But analysing data, finding patterns, and understanding the unpredictable nature of T1 is a black art. You need to have the support of T1 specialists that can work with you and support you.
What’s still needed and desperately missing is training for the young diabetics to manage their condition independently. We cannot be with them 24 hours a day. They make decisions without us.
My ending positive note. My son just completed a two day cycle ride with an overnight camp – over 50 miles. Lots of lows. Not many highs. But that CGM proved it was worth every penny. I’m extremely proud of my son for managing his diabetes and achieving this. It’s a long journey and together we will work through it.