Early Patient Input to Improve Safety, Usability, and Dosing Accuracy in a Diabetes Insulin-Dosing App
Source
Medical Device Innovation Consortium
Link
Supporting Patients in Making Treatment Decisions
Description of Case Study
Roche Diabetes Care, Inc., engaged patients in the development of a diabetes insulin-dosing app to ensure the product reflects how patients make insulin decisions in practice and to reduce the risk of user errors associated with manual calculation and cognitive burden. The research and development (R&D) team, including human factors researchers and risk management experts, collaborated with patients through direct input, usability testing, and iterative feedback across multiple study rounds. Patient input informed app features, usability improvements, and safety controls designed to support more accurate and personalized insulin-dosing decisions.
Quick Facts Organization name: Roche Diabetes Care, Inc. |
Product Type(s)
digital health technology:
- care support
Product Medical Specialty Area(s)
- other: endocrinology
Stakeholders
- medtech company: internal discovery/R&D teams
- patients and caregivers, patient organizations
- providers (clinicians, hospital/health system administrators)
- other: human factors researchers, risk management experts
Methods and Sources of Patient Input
- interviews/focus groups/surveys
- human-centered design methodologies
- other: usability testing with patients
Impact of Patient Engagement
Reported areas of impact include:
- Design focused on patient cognition and behavior: App features were designed with patients’ perception, cognition, and actions in mind, including storing therapy parameters provided by health- care professionals to reduce memory burden.
- User interface design to reduce use errors: The interface was designed to prevent certain actions and inform patients of hazardous situations, addressing risks associated with manual calculation and unclear information.
Lessons Learned
This case study demonstrates the following lessons learned:
- Patient input informed multiple stages of development: The case demonstrates that patients were engaged from initial input gathering through iterative design refinement. Human factors researchers interacted with patients both directly and indirectly, and design changes were made based on analysis of usability study data. This suggests patient input was not a one-time requirement but an ongoing part of the development process.
- Usability study data directly influenced risk mitigation strategy: The case explicitly states that the team assessed patient outcomes from usability studies to determine what risk-mitigating controls were needed. This indicates that observations of actual patient use and errors informed which safety measures would be implemented.
- Understanding patient cognition and behavior was central to design: The case states that app features were designed with patients; perception, cognition, and actions in mind. The resulting design, particularly the storage of therapy parameters and simplified input requirements, addressed specific cognitive limitations (memory recall, mental calculation) that could lead to errors.
- Multiple parties need to engage with patient data: The case involves human factors researchers, R&D teams, and risk management all working with patient feedback from usability studies. This cross-functional approach meant that insights about how patients actually used the app informed decisions about design, safety, and risk controls.