Based on Moore, V. and Quigley, M., ‘Open-Source Diabetes Technologies: On Clinician Liability and Patient Safety‘, published in Patient Safety Law, Ethics and Practice in the NHS (Routledge, 2026).
In previous work we examined how GMC guidance on prescribing and consent applies to clinicians caring for OS AID users, concluding that nothing in it should be read as prohibiting clinicians from discussing or prescribing these systems. In this new book chapter, we turn to the question underpinning much of that anxiety: could clinicians actually be found legally negligent? Our overall conclusion is that it is highly unlikely.
The Legal Framework
For a clinician to be found negligent, a claimant must establish duty of care, breach, and causation. Breach is assessed under the Bolam test: a clinician is not negligent if they acted in accordance with a practice accepted as proper by a responsible body of medical practitioners. This is qualified by Bolitho, which requires that body of opinion to withstand logical analysis. We apply this framework to the following four main scenarios.
Informing patients about OS AID
Where a clinician proactively discusses OS AID risks, the duty of disclosure in Montgomery v Lanarkshire Health Board applies. A clinician who provides accurate, balanced information, including about the unapproved status of these systems, is unlikely to be found negligent. Clinicians should be aware that they may be expected to have some familiarity with OS AID given its prevalence, and should be honest with patients about the limits of their knowledge.
Failing to mention OS AID
Could a clinician be negligent for not raising OS AID with a patient who hasn’t asked? The Supreme Court’s 2023 judgment in McCulloch v Forth Valley Health Board clarified that whether a treatment constitutes a ‘reasonable alternative’ requiring disclosure is determined by the Bolam/Bolitho professional practice test. Given the absence of regulatory approval and the still-developing evidence base, it is unlikely a court would consider a decision not to mention OS AID to be illogical. Clinicians are not under a legal obligation to raise it proactively; though, as we argued previously, there are strong ethical reasons why they should be free to do so.
Advising patients about OS AID
Discussing OS AID with a patient who raises it is not automatically negligent. The risk arises if a clinician makes a negligent misstatement (careless advice on which the patient relies and which causes harm). The practical message is clear: be accurate, acknowledge the limits of your knowledge, and do not speculate about technical aspects of systems you may be unfamiliar with. Even then, establishing causation is very difficult: a patient would need to show that correct information would have changed their decision, a high bar for a community of users who have actively sought out these systems.
Prescribing the component devices
Prescribing the licensed CGM and insulin pump components is unlikely to be negligent where NICE guidelines are followed. Even if breach of duty could be established, the patient’s independent decision to install the software, configure the system, and manage it on an ongoing basis would likely constitute an intervening act breaking the chain of causation. The multiple steps between clinical decision and adverse outcome make it exceptionally difficult to establish the clinician’s prescription as the material cause of any harm.
Patients’ own responsibility
Where patients have been open and harm occurs, two legal doctrines become relevant. The first is contributory negligence: where a patient’s own actions have contributed to the harm they suffered, a court can reduce the damages they recover proportionately. Given that OS AID users take a proactive, hands-on role in building and operating their systems (and in some cases do so against clinical advice it would not be unreasonable for a court to find that a patient bore some responsibility for any harm they experienced. The second is volenti non fit injuria – broadly, the principle that someone who voluntarily accepts a known risk cannot then seek compensation for harm arising from it. Given that OS AID users typically build and operate their systems in full knowledge of their unapproved status, this defence has some relevance. In practice, courts tend to prefer contributory negligence over volenti: rather than removing liability entirely, it allows damages to be reduced to reflect each party’s share of responsibility.
It is also worth keeping perspective on the broader picture. OS AID users are a highly engaged, technically informed group who are well aware they are taking on personal responsibility for an unregulated system. The likelihood of legal action being brought at all is low. The strengthening evidence base means a clinician’s decision to engage thoughtfully with these systems is increasingly unlikely to lack the ‘logical basis’ required by Bolitho.
The anxiety driving the hands-off approach is understandable, but it is not well-founded in law. Existing legal principles offer clinicians considerably more protection than many may realise.
The full chapter is available on request.
Written by: Victoria More
Funding: Work on this was generously supported by a Wellcome Trust Investigator Award in Humanities and Social Sciences 2019-2026 (Grant No: 212507/Z/18/Z)