Abstract: PUB086
Role of Artificial Intelligence (AI) in Renal Diet
Session Information
Category: Augmented Intelligence, Digital Health, and Data Science
- 300 Augmented Intelligence, Digital Health, and Data Science
Authors
- Fichadiya, Harshil, Mayo Clinic Minnesota, Rochester, Minnesota, United States
- Patel, Vraj, Canton Medical Education Foundation, Canton, Ohio, United States
- Craici, Iasmina, Mayo Clinic Minnesota, Rochester, Minnesota, United States
Background
Patients with advanced CKD and ESRD experience hypertension, hyperkalemia and hyperphosphatemia and frequently struggle with maintaining a low sodium, potassium and phosphorous diet. Part of this struggle is from lack of understanding of foods containing higher amount of these electrolytes and lack of an organized diet plan. Consultation with a renal dietician may help them create a dietary plan, however the high CKD patients to renal dieticians ratio and subesequent difficulty with frequent monitoring of dietary plans make this option less feasible for many patients. An AI tool that can monitor daily dietary electrolyte intake and suggest dietary plans per patient preference may be more affordable and accessible to majority of patients.
Methods
We used ChatGBT- 4 to
i) Assess electrolyte content of various foods and suggest a renal friendly alternate
ii) Devise a daily renal dietary plan for vegan, vegeterian and meat based diet (K<60 Meq, Na <2g, Phos <1g/kg/day) for 70 kg patient
iii) Assesed if Chat-GBT was able to modify the dietary plan based on patient preference for episodic food craving while maintaining compliance with renal diet.
Results
Chat GBT 4 devise a wonderful vegan, vegeterian and meat based dietary plan (including breakfast, lunch and dinner) with renal electrolyte restriction as mentioned above. It was able to accomodate a small serving of high potassium food high phosphorous food (orange & cookie), when asked to modify diet to include episodic cravings that our patients may feel. It was also able to suggest alternates to high potassium foods, eg apple juice, cranberry juice or grape as an alternate to orange juice.
Conclusion
Though simple, this intervention has the potential to be highly effective in helping advanced CKD and ESRD patients maintain compliance with a renal friendly diet. While daily access to a renal dietician may be difficult, this intervention will allow patients to devise daily renal friendly dietary plan based on their food preference while accomodationing for episodic non-renal friendly food craving in real time. Long term effective compliance with a renal friendly diet will translate into better bone health, slower CKD progression and reduction in episodes of dangerous hyperkalemia between dialysis sessions. In summary AI will complement the role of a renal dietician by facilitating effective implementation of dietary goals set by the dietician.