Hospital Digital Management System with Constraint-Aware Personalized Diet Recommendation

Hospital food management has to coordinate patient information, dietary restrictions, food and meal preparation and delivery. Traditional workflows might cause a division of patient management from food-service activities, which reduces the possibilities of centralized coordination and personalization. The aims of this work are to create a Hospital Digital Management System which will enable patient management, recommendation of dietary suggestions, generation of dietary chart, preparation of the pantry and meal delivery tracking. Prototype: Backend services were created with Django REST Framework; the database is SQLite; the presentation layer is written in React.js and mounted with a role-based authentication system using JSON Web Token. A constraint-aware rule-based recommendation mechanism, which takes medical condition, allergy information, dietary preferences, meal requirements, food restrictions, etc., is used to help the generation of four meals daily diet plans. Simulated patients: Six patient scenarios were used for the prototype. Meals delivered: There were 24 instances of meals delivered. Coverage of diet-charts was 100%, and 22 out of 24 deliveries were done (91.67%). In the prototype environment, the implemented recommendation function was able to provide the displayed diet plans in less than 3 seconds. Conclusions: It has been shown that it is possible to combine individualized dietary support with the organization of the food-service process in the hospital. The current implementation is a rule-based system that is constrained based instead of a trained predictive machine learning model. There is a need for clinical validation and larger scale evaluation before implementing in clinical care. Introduction: As the healthcare industry goes digital, the need for unified systems to handle patient data, dietary needs, food preparation, and delivery is becoming more pressing. With medical conditions, food allergies, dietary preferences and food restrictions, personalized dietary recommendation can help meet the individual's needs. Objectives: This study seeks to create an integrated hospital digital management system that integrates patient management, constraint-aware dietary recommendation, diet-chart generation, pantry preparation, and meal-delivery tracking. The system also intends to analyse the prototype based on cover of diet chart, completion of delivery, response time for recommendation and operation of workflow in the roles. Methods: Web-based prototype developed with React.js, Django REST Framework and SQLite with JWT-based role authentication. A constraint-aware rule-based recommendation mechanism is used to create 4 meal diet charts based on patient medical condition, allergy, food preference, food requirement and food restriction. A total of 24 meal delivery instances were used with six simulated patient profiles to evaluate the prototype. Results: The patient charts were generated for all six simulated patient profiles, with 100% coverage of diet-chart. A total of 24 meal deliveries were tracked, with 22 deliveries (91.67%) being completed. The prototype took less than 3 seconds to produce the diets shown. Results are prototype observations and not clinical validation. Conclusions: The prototype highlights the potential for incorporating patient-specific dietary restrictions into hospital food-service processes, such as the creation of meal menus, preparation of meals in the pantry, and monitoring the delivery of meals. The current system is based on constraints-aware rule-based system instead of a trained predictive machine-learning model. Before clinical use, clinical validation and evaluation on a larger scale are needed.