Do-It-Yourself (DIY) Closed Loop Systems: Perspectives of an Endocrinologist

Abstract

Do-It-Yourself closed loop systems are becoming increasingly popular as an alternative choice for managing type 1 diabetes. In this commentary, the issues that are faced by both the clinician and the person with diabetes are discussed from my perspective.

Technology plays a vital role in the care of people with diabetes (PWD). The development of hybrid closed loop systems that automate basal insulin delivery independent of user input is a step closer to replicating beta cell function. The profile of Do-It-Yourself (DIY) closed loop systems has increased in the last five years, emerging as an alternative to commercial systems for individuals that are not willing to wait for mainstream technology to catch up to their needs. My foray into this area was spurred on by a blog post that mentioned the use of a self-built artificial pancreas system. I am writing this from the perspective of a person interested in technology, a person who has friends with type 1 diabetes and as a clinician who cares for many people with type 1 diabetes. I do not have diabetes and do not loop, nor do I claim to know what it is like to live with diabetes but have learned a lot from my patients and friends who do live with diabetes.

Living with diabetes can be challenging at many levels. While there are many facets to address, a major aim is to normalize glucose levels to reduce the risk of developing diabetes-related complications. In recent times, the shift has been to focus more on time spent in target range rather than on HbA1c alone.¹ This needs to be undertaken safely with individualized treatment targets taking into consideration factors such as age and risk of hypoglycemia. While health care professionals (HCPs) strive to offer the best health advice toward this goal, the reality is that most individuals spend very few hours each year in consultation with HCPs. In addition, not all HCPs are equally experienced in the latest technology. Self-management is thus, key to success, and equipping PWD with the best available tools (education, technology, and equipment) is paramount.

Mimicking the role of the endocrine pancreas is physiologically difficult. Beta cell destruction in Type 1 diabetes results in the loss of ambient glucose sensing and deficient insulin production. Closed loop systems attempt to bridge this gap by integrating a glucose sensor, an algorithmic engine, and an insulin delivery device to try and replicate beta cell function. The manual administration of insulin at fixed doses is no match for a well-tuned closed loop system (DIY or otherwise) that continuously self-adjusts insulin delivery to maintain normoglycemia.

At present, DIY systems have yet to be approved by regulatory bodies such as the FDA (US Food and Drug Administration), and this has been challenging to the medical fraternity, insulin pump, and Continuous Glucose Monitor (CGM) manufacturers, as well as to the regulatory establishments themselves. Regulatory approval for a product is usually obtained by a backing entity with commercial interest, but this is not the case in the DIY space. This is significant for HCPs as recommending the use of unapproved systems for diabetes management raises both legal and ethical issues that are not easily solved. This also provides fuel to the fire for sceptics of DIY systems. Although the number of users of these systems is not astronomical (>1500 individuals)² and represents a minority of the population with type 1 diabetes, the popularity of these systems continue to grow. Do-It-Yourself systems are here to stay, and it is important for HCPs to be aware of it.

How does the average HCP learn about these DIY systems? The principle underlying this movement is that it is DIY and there is no “company” that is actively promoting its use among HCPs. Education needs to be self-driven. The ecosystem has its support system based around the internet and unless one (a) forays into such groups online, (b) encounters a person using such technology, or (c) is told about this by word of mouth, it is unlikely that you will be exposed to the DIY systems. Having said this, there has been increasing discussion of DIY systems in medical circles, and scientific meetings have begun to incorporate topics around DIY systems into their programs. There also appears to be a growing interest in mainstream media although the reporting is not always accurate. Becoming a member of various Facebook groups dedicated to the DIY cause is another avenue for education. A recent publication entitled “Automated Insulin Delivery: How artificial pancreas ‘closed loop’ systems can aid you in living with diabetes”³ is also a useful resource. Another means of learning about such systems is to procure the necessary hardware and attempt building one (or more) of these systems yourself. This process has been personally rewarding and provides some insight into the world of looping. While I am not technologically inept, the support of others in this space has been invaluable in this process. The difference here is that this support does not come from a telephone helpline or company but rather from other users of DIY systems who are located somewhere in the world. Responses through the internet are rapid and distance is not a factor.

As a clinician, the safety of DIY systems is paramount. Insulin is effective treatment but can result in significant harm when administered incorrectly or inadvertently, regardless of delivery device. Safety is dependent on (a) hardware reliability, (b) the safety margin of the algorithm, and (c) end-user factors such as equipment maintenance, parameter adjustments, and life events. It is important to remember that problems can arise irrespective of your chosen system whether it is with a closed loop system (commercial or DIY) or with multiple daily injections. In an ideal world, DIY systems would be built using the latest equipment but a significant proportion of loopers use older, out of warranty pumps because of their open communication channels that enable looping. There will always be a concern about equipment reliability when older equipment is used, but this appears to be an acceptable risk to the users given the anticipated benefits of closing the loop. Unless there is actual failure of the pump, any loss of communication between the hardware components results in defaulting to open loop, resuming insulin delivery as originally setup in the pump. Given the difficulties obtaining a loopable pump, not everyone has a backup system capable of looping should their primary systems fail but the generosity and support of this niche DIY community to troubleshoot often astounds.

In all the DIY instructions, it is explicitly stated that the end user is responsible for building and maintaining their systems and accepts the associated risks. My understanding is that DIY systems are developed with safety in mind and the source code is available for review. After all, it would seem fool hardy for the developers to build something that they did not feel safe to use themselves. Conventionally, clinical trials are required to ensure that a particular “medication” or “treatment” is safe and efficacious. Although there are many hours of safe use and better glycemic control² reported with DIY systems, what is needed in the literature are published trials involving larger cohorts of individuals using DIY systems that examine their safety, efficacy, and effects on the user quality of life. This is a necessary step that will likely result in better acceptance of this technology among HCPs. There is activity in this area, such as that promulgated by the OPEN (Outcomes of Patients’ Evidence with Novel, Do-it-Yourself Artificial Pancreas Technology)⁴ consortium in Europe, the clinical study being undertaken by Dr Martin de Bock in New Zealand focusing on “Automated insulin delivery for type 1 diabetes utilising open source technology”⁵ and also work by Dana Lewis, on fostering and supporting research among the DIY community.⁶

As more people adopt DIY systems to manage diabetes, more HCPs will encounter them in consultation. Health care professionals are concerned about their professional liability when offering treatment advice around these unregulated systems and PWD are concerned whether their HCPs will support them in their use of a DIY kit. Should the use of a DIY system change the interaction dynamics with a HCP? As the person’s diabetes has probably not disappeared, they have a clinical reason to be supported regardless of their choice of treatment. After all, this situation is akin to the PWD that has just moved into the country and is using a form of medication (or treatment) that is not available in the new country. What is necessary though is a conversation between the two parties to delineate the boundaries of their therapeutic relationship. This is what I have employed and it appears to work reasonably well. As the HCP is unlikely to have access to the glycemic data via standard pump uploads, it would be useful for the PWD to provide this for the consultation. Other clinical aspects of management such as sick day management, exercise, screening for complications, and a mental health review could still be explored, and routine tests should still be undertaken at appropriate intervals. Having some understanding of various DIY systems is useful for the interaction; this does not mean that there is an obligation to provide advice about the DIY system.

Are DIY closed loop systems suitable for everyone with Type 1 diabetes? Probably not, but to be fair, this answer requires qualification. Apart from the improvement in glycemic control, loopers report an improved quality of life, especially in terms of sleeping well overnight.⁷ Although there are good instructions available on how to build a system, the PWD needs to be sufficiently motivated to invest enough time and effort to maintain and fine tune their system. Some individuals report time in range results that are superior to commercial systems, but others struggle. Some may be uncomfortable with the “unregulated” status and may prefer an “off the shelf” product that comes with a warranty. The cost of CGM, fatigue from alarms, and the “technological burden” are factors that sometimes result in the cessation of all pump therapy but this is not exclusive to DIY systems.

In the end, what is offered by the DIY systems is choice for the PWD. It is not so much about trying to compete with commercial systems as both systems have their pros and cons. The uncertainty about the regulatory status of such systems will be an ongoing issue for both the HCP and the PWD. It is time for various professional bodies to step up, obtain some legal advice, and work with PWD to develop some guidelines so that nothing is ambiguous for both parties when it comes to DIY looping.