JMIR Publications

ABSTRACT

Background:

45% of older adults living in long-term care (LTC) have some form of malnutrition. Several methods of tracking food and fluid intake exist, but are limited in terms of their accuracy and ease of application. Thus, an easy to use, objective, accurate, and comprehensive intelligent food intake system designed with LTC in mind may provide additional insights regarding nutritional support systems and nutritional interventions.

Objective:

The purpose of this research was to conduct a multi-stage participatory iterative design sprint of a Goldilocks quality horizontal prototype for the Automated Food Imaging and Nutrient Intake Tracking (AFINI-T) system. Specific design objectives included: (1) identify practice-relevant problems and solutions through user-centered participatory design, (2) mitigate feasibility-related barriers to uptake, and (3) employ user-centered technology development.

Methods:

A six-stage iterative participatory design sprint was developed and executed. Thirty-eight participants and advisors representing 15 distinct roles (e.g., personal support worker, nurse, dietitian etc.) were engaged in the design sprint. Trust, subjective workload (RTLX), subjective usability scales, and a modified Ravden checklist were used to assess project advisors’ perceptions of the AFINI-T system prototype compared to the current method of food and fluid intake charting.

Results:

The top three priorities for this system were identified as: (1a) ease of use, (1b) high accuracy, (2a) system reliability, (2b) ease of maintenance, and (3) requirement of integrating with the current PointClickCare system. Project advisors informed design decisions leading to a Goldilocks quality horizontal prototype of the AFINI-T system. Compared to the current food and fluid intake charting system, AFINI-T was perceived to have: less time demands (P < .001), less effort (P < .001), and less frustration (P = .002) with a lower perception of system deception (P = .005) and wariness (P = .006). Usability ratings of the AFINI-T prototype were high with a subjective usability score mean of 89.2 and the highest ratings on a modified Ravden usability checklist of “very satisfactory” for 7/8 sections.

Conclusions:

Based on the design process, the concept of the AFINI-T system as a tool for an intelligent food and fluid intake appears to have good practice-relevance. Feedback from evaluation of the AFINI-T prototype suggests that many feasibility-related barriers to uptake could be removed with an improvement over the current system and that advisors are keen to try the AFINI-T system. Design decisions were informed through the application of a user-centered participatory iterative design sprint; we present this approach in the context of our case-study to provide tangible examples of how this method can be applied in the development of novel needs-based application-driven technology.