Supporting Patients Using DIY and Open-Source AID Systems

Open-source automated insulin delivery (AID) systems are increasingly recognized for their unique customizations, rapid innovations, and supportive communities. Several varieties exist, each with its own features and algorithms. Although these programs are not FDA-approved, emerging studies and real-world outcomes suggest they can be safe and effective—when used by engaged and informed individuals. For clinicians, particularly DCESs, supporting patients who choose these tools requires a shift from traditional top-down education toward patient-guided, collaborative care.

Setting Expectations

Not every person with diabetes will be a good candidate for an open-source AID system. The ideal user should be well-informed, technologically comfortable, and willing to take an active role in their care. Key characteristics include:

• Demonstrated diabetes self-management skills

• Experience with CGM and insulin pump therapy

• Comfort with smartphone technology

• A willingness to learn and troubleshoot

• Ability to trust the recommendations of an automated system

Solid Start

Before starting an AID system, candidates need realistic expectations and a clear understanding of how the system is built and operates. Since these systems are not commercially supported, users are responsible for installing, updating, and maintaining the apps themselves—or identifying a dependable support system. DCESs can guide patients by:

• Identifying and securing compatible hardware

• Reviewing basal and bolus settings

• Coaching through common technology hiccups

• Connecting users with peer communities and educational resources

Legal and Ethical Considerations

Because open-source AID systems are not FDA-approved, healthcare professionals should be clear that their role is supportive, not prescriptive. Documenting patient conversations, reviewing informed decision-making, and clarifying the limits of clinical responsibility help protect both patients and providers. While clinicians are not endorsing or prescribing these systems, acknowledging and supporting safe use helps patients achieve better outcomes.

If you are not comfortable supporting patients who choose to use open-source AID systems, it is best practice to refer them to a healthcare professional who is familiar with these technologies. This approach ensures patients receive informed, safe guidance while maintaining trust in the care team.

Continued Support

Once an open-source AID system is running, the clinician’s role continues. Users benefit from guidance on troubleshooting, interpreting data, and exploring advanced features. It is important to emphasize that while open-source AID systems automate insulin delivery, they still require hands-on engagement and ongoing adjustments. Platforms like Nightscout and Tidepool are commonly used for data sharing and analysis, giving HCPs opportunities to teach patients how algorithms function and when to self-adjust settings.

Building HCP Competence

For many clinicians, open-source AID systems may feel outside of traditional training. Building confidence begins with exploring trusted resources such as LoopDocs, Loop and Learn, and professional case studies. Participation in peer learning networks, webinars, or continuing education can also prepare DCESs and other HCPs to provide meaningful support. Familiarity with these systems allows clinicians to meet patients where they are, even without prescribing or directly managing the technology.

At a minimum, an HCP should understand these core points to feel comfortable guiding patients who use open-source AID systems:

1. What open-source AID systems are
   – They are user-built, non-FDA-approved apps that connect existing pumps and CGMs to automate insulin delivery.
   – They differ from commercial AID systems in that patients maintain and update them independently.

2. Their role as an HCP
   – Supportive, not prescriptive: clinicians don’t prescribe or install these systems but can help patients use them safely.
   – Document discussions to protect both patient and provider.

3. Who makes a good candidate
   – Strong diabetes self-management skills, comfort with technology, willingness to troubleshoot, and reliable backup plans.

4. What to review with patients before starting
   – Importance of accurate basal/bolus settings.
   – Need for realistic expectations: these systems automate insulin delivery but are not “set it and forget it.”
   – Awareness that support comes mainly from peer communities, not traditional customer service.

5. What ongoing support looks like
   – Help patients interpret CGM and pump data (often through platforms like Nightscout or Tidepool).
   – Coach them on algorithm logic, self-adjustments, and troubleshooting common issues.
   – Encourage ongoing learning through community resources and training.

6. Where to find resources
   – Familiarity with trusted guides like LoopDocs, Loop and Learn, and AAPS Docs.
   – Awareness of continuing education opportunities or peer-to-peer learning for HCPs.

If an HCP can grasp these essentials, they don’t need to be a technical expert to provide meaningful guidance, they just need to know how to assess readiness, set safe expectations, and point patients to reliable resources.

Practice Tips

1. DIY doesn’t mean do-it-alone. Users build and maintain the system, but support from healthcare professionals is essential.

2. Start with the basics. Accurate basal rates, insulin sensitivity factors, and carb ratios are critical; open-source systems amplify correct settings—and expose weak ones.

3. 1-800 social media. These systems lack traditional customer service. Patients need to know where to find reliable online help.

4. Readiness is more than motivation. Interest in technology matters, but so do confidence, trust, and problem-solving skills.

Readiness Checklist

Patient Readiness
✓ Strong diabetes self-management and troubleshooting skills
✓ Experience with CGM and insulin pump therapy
✓ Comfortable with technology
✓ Capable of following detailed instructions
✓ Access to compatible hardware and support resources
✓ Willing to build and maintain the system—or has reliable help
✓ Backup plan in place for downtime or failure
✓ Understands that the HCP role is supportive, not prescriptive

HCP Readiness
✓ Basic training or access to learning resources on open-source AID
✓ Familiarity with peer communities and online support tools

Resources

• Loop and Learn

• Loop Docs

• Trio Docs

• AAPS Docs