- 
                Notifications
    You must be signed in to change notification settings 
- Fork 1.4k
Description
Is your feature request related to a problem? Please describe.
I am a patient with congenital hyperinsulinism (endogenous hyperinsulinism) who is treated under close medical supervision with somatostatin analogs (e.g., Sandostatin) delivered through an Omnipod Dash pump.
Currently, the loop app is designed for diabetes and assumes that high glucose requires increasing insulin delivery.
In my case, this logic is reversed I need the system to increase delivery when glucose is low or dropping, and possibly reduce delivery when glucose is stable or rising.
This limitation prevents me from safely using the automatic loop features of loop, even though my setup is medically supervised and would greatly benefit from automation.
Describe the solution you’d like
I would like to suggest adding an optional “Reverse Logic” mode (or “Custom Pharmacodynamic Mode”) where the algorithm’s response to glucose trends can be inverted or manually defined.
For example:
• When glucose is falling, the app could increase the delivery rate or trigger a bolus of my medication (somatostatin analog).
• When glucose is rising, the app could decrease or suspend delivery.
This feature could be implemented as a configurable toggle under advanced settings, available only to users who enable it manually and confirm they are under medical supervision.
Describe alternatives you’ve considered
I have considered manually delivering boluses and adjusting basal rates, but this requires continuous attention throughout the day and night, which is unsustainable long term.
I also explored modifying open-loop targets to simulate a reverse response, but the system still fundamentally treats low glucose as a danger signal rather than as a trigger for action.
Additional context
This use case is for patients with endogenous hyperinsulinism using somatostatin analog therapy, not for diabetes.
Such a feature could potentially help other patients under similar medical guidance.
All medication dosing parameters would still be fully user-defined according to medical instructions the system would only automate timing and trend response.
Technical Details
From a technical standpoint, this could be implemented by:
• Allowing an optional inversion of the control algorithm’s sign (e.g., reversing the PID controller’s correction direction).
• Exposing a customizable “response curve” editor for advanced users to define how delivery should change relative to glucose trend.
• Keeping existing safety limits (max/min delivery, bolus caps) active for protection.
User Impact
This would enable safe semi-automated therapy for a small but important group of medically supervised patients using the same hardware infrastructure (CGM + Omnipod Dash) for a different metabolic condition.
It would also make loop more flexible and medically inclusive