Month: September 2024

Looping/Automated Insulin Delivery: Does it Work? Commercial vs Open Source, Affordability, and How Professionals Can Support the Lived Experience

leontribeLeave a comment

I recently presented at the Australian Diabetes Educators Association NSW Branch Conference and wrote an article for the Australian Diabetes Educator journal on looping and how diabetes educators can support people with type 1 diabetes (PWDs) using looping systems, both open source and commercial. This is not a verbatim reproduction of the presentation or the ADEA article but does cover some of the key points which may be useful to PWDs and Health Care Professionals (HCPs). While the focus of the article is the Australian experience, there is value for international audiences as well.

What is Looping?

There are many terms for technologies allowing a Continuous Glucose Monitor (CGM) and an insulin pump to make blood glucose management decisions without the intervention of the PWD. Originally, when such technologies were ‘hacked’ together on the internet, it was referred to as looping. Now, with both open source (internet) and commercial versions available, and recognition in the academic literature, the more formal term Automated Insulin Delivery systems (AIDs) is used. For the rest of the article, this is the term I will use. Similarly, for simplicity, open source/internet systems I will refer to as ‘Open Source’ and commercial systems as ‘Commercial’. There are nuances which are lost in this broad classification but, for the purposes of this article, they are not overly important.

For a brief history of how the internet made the first AID, you can go to my other blog. These days, there are a variety of options for PWDs. In Australia, there are now four main Open Source systems: 

For Commercial systems, at the time of writing, Australia has three commercial looping systems available: 

  • Medtronic 780G + Guardian 4 Sensor 
  • Tandem t:slim X2 with Control IQ + Dexcom G6 
  • YpsoPump with CamAPS FX + Dexcom G6 

Open Source vs Commercial Systems

These days, I use a Commercial system (Ypsopump with CamAPS FX + Dexcom G6) but my first loop was Android APS using both the Libre 2 and Dexcom G6 CGMs and Omnipod pumps. Neither is better than the other but each system has their own features which are more or less appealing to each PWD.

Open Source Commercial 
Built by the wearer/user Commercially available software 
(Generally) The code which it is based on is available for review Proprietary i.e. the code is not available and how it works is often a trade secret
Very ‘tweakable’ Less flexibility/features 
Not formally approved by a government bodyApproved by a government body i.e. the Therapeutic Goods Administration

There is nothing illegal about using an Open Source AID. The lack of government approval means while they cannot be sold commercially, the law in Australia allows the personal use of unapproved medical technologies and devices. This means, while the software code is usually openly available, it is the PWD who must compile it (not as technically challenging as it sounds these days). The advantage of Open Source is that, because it does not have to go through formal approval processes every time a change is made, innovation comes faster with Open Source AIDs.

The advantage of a Commercial system is that it is largely ‘plug and play’. While PWDs need formal training on the use of the CGM and pump, the commercial software is often simpler to use and administer. Some consumers prefer to be hands-on with their technology while others prefer simplicity.

Do AIDs Work and What Are Their Limitations?

AIDs continue to improve but they do have their limitations. While it is possible for a PWD to modify their diet e.g. eat a Low Carb or Very Low Carb diet, and minimise their intervention with the AID, most PWDs will need to bolus for meals and take care during exercise. For tips on dealing with AIDs in exercise, refer to my exercise article.

My experience has been, being a LADA with some residual pancreatic function, I can skip a lot of bolusing and the AID is very good at keeping me in range. These days I bolus for very carby meals (say 45g or more) but do not need to declare/bolus the rest. Everyone’s experience will vary. One thing I never do are bolus corrections. If I do not declare, the AID keeps adding insulin to manage the carbs and automatically manages the ‘landing’, preventing a hypo. If I manually bolus too much, the AID will suspend basal insulin to ‘soak up’ the excess. No corrections means no stacking and while it can be tempting to bring my levels down quicker after a meal, I always try to let the AID do the work. Similarly, whatever my levels are when I go to bed I can, almost without fail, be assured I will wake up in range. As long as my pump has enough insulin and my CGM does not expire, I can be assured a good night’s sleep.

The other great thing about AIDs is they do not stop you from pulling other ‘levers’ in managing diabetes. If you want to eat a very low carb diet with an AID you can. I touch on this in my “Practical Diabetic Solution” article. In short, when the PWD is asleep or otherwise occupied, the AID takes care of glucose management, ensuring superior results compared to relying exclusively on the PWD to make all decisions 24-7.

A criticism sometimes made of AIDs, especially in the Commercial systems, are their high blood glucose targets (the Tandem, for example, does not allow for a target below 6.1mmol/L (110mg/dL)). The truth is not all AIDs are equal. Broadly speaking the Open Source systems allow for a more aggressive target than the Commercial systems but the system with the lowest possible target is the CamAPS system whose target can be made to go as low as 4.4mmol/L (80mg/dL) although I generally set mine to 5.5mmol/L (99mg/dL) as this is the average resting glucose of a person without diabetes.

To the question of how well these systems work, there is now a strong body of evidence that they do and, from a clinical perspective, outperform any other alternative. For example, here are the times in range for the Medtronic system across various regions of the world. As we can see, it is typical to be at least 70% in range (70-180mg/dL) with minimal lows. These results are consistent across the variety of AID systems and, as mentioned before, do not prevent even better results through the use of adjunct therapies such as medications and diet.

How Do People Decide on the Right System for Them?

There are four key factors to consider when choosing an AID:

  • Availability: Are the pump, CGM, phone and related consumables available? 
  • Accessibility: Are the pump, CGM, and phone usable by the individual? 
  • Adaptability: Can the AID system be extended, as desired? 
  • Affordability: Is the AID system affordable? 

In terms of affordability, Australia is unique in being ‘half-pregnant’ with their AID subsidies. For people with type 1 diabetes, CGMs are funded with a small co-payment by the PWD whereas pumps are effectively unfunded. To get an insulin pump, the PWD must either get private health insurance or pay the, say, $8,000 for the pump out of their own pocket. There is a petition urging the Australian government to change this inequity which I encourage you to sign.

For the other factors, considerations include:

  • More choice of glucose targets.: As mentioned, different systems have different target ranges which may be important to some 
  • Readability: For a user with vision impairment, a high-contrast screen or management via a mobile phone may be desirable. 
  • Connectivity to other devices: Being able to access functions via a watch or the internet may be useful. 
  • Form factor: Device size, whether it is tubed or tubeless, and whether it has an IP rating for water resistance can be key decision factors for some wearers

The available combinations are also important as it is likely a PWD is already using a specific CGM and may want to stay with that CGM as part of their AID system. In fact, there are limited combinations available for AIDs.

 Tandem t:slim X2 YpsoPump Omnipod Dash Medtronic 780G 
Dexcom G6 Android/iPhone (CGM only) Android via CamAPS Android/iPhone  
Medtronic Guardian    Samsung/iPhone (Monitor only) 
Abbott Libre 2   Android  

Above we can see the available Open Source and Commercial combinations in the Australian market for pumps and CGMs approved for use in Australia (Green = Commercial, Blue = Open Source). As we can see, for someone using a Libre 2, the only option available is an Open Source one. Similarly, while the next version of the Omnipod has an AID system (approved in Australia but not widely available), the main Omnipod version available in Australia, at the time of writing, does not with the gap being filled, again, with Open Source.

How HCPs Can Help?

There is understandably some nervousness with HCPS about supporting or recommending Open Source systems to PWDs. These are, almost by definition, unofficial/unapproved systems. On the other hand, to stay relevant to a PWDs care, they cannot remove themselves completely and let the internet and social media be a PWD’s only source of information. Thankfully, there is some guidance for HCPs out there.

Locally, guidance is limited. Diabetes Australia has a position statement on “People with type 1 diabetes and Do It Yourself (DIY) technology solutions” but it is over five years old, light on specific guidance and, arguably, out of date. In its statements around commercial systems it suggests “(In Australia) there is no approved product that is a fully closed-loop system”, and “there is currently no subsidy for the CGM component for adults over the age of 21 years”. Both statements are no longer correct.

Internationally, there are three published documents providing guidance to HCPs and applicable to the Australian context. These are:

These documents are comprehensive in their guidance and freely available to view and download at the above links. Their main points are: 

  • While there are legal complications in recommending Open Source systems, users still require support. Therefore, it makes sense for the HCP to learn about these systems to provide effective guidance. 
  • If the HCP is unable to learn and provide guidance, they should collaborate with individuals who can. 

Conclusions

AIDs provide a lot of value to PWDs and both Commercial and Open Source systems are valid options, depending on the PWD’s needs. While they have their limitations, in terms of bolusing and exercise, AIDs are proven to provide benefit in all global settings in terms of time in range and overnight management.

While it can be daunting for an HCP to provide guidance to PWDs using AIDs, especially if it is Open Source, it is important to provide a voice of reason when there can be a variety of information sources which are less informed or reliable. Thankfully, there are multiple peer-reviewed international guides for the HCP to provide precisely the support PWDs need.