Table 4 Human-centered design principles for HRS
From: Human-centered design of a health recommender system for orthopaedic shoulder treatment
I-C-IT Design Features | Human Factors for System Design | Design Principles |
|---|---|---|
- Familiar and utilitarian data elements for treatment decisions. - Appropriate feedback on action. - Accuracy and reliability in data analytics. | User trust in data evidence | Predictability |
- Physicians search history learning (content filtering) - Similar search learning (content and collaborative filtering) - Patient cohort learning (collaborative filtering) - Patient’s treatment outcome preference-based risk benefit profiling of treatment options. | User preference integration | Personalization |
- Clear data cohort selection with data classifiers on top of visualization pane. - Number of patients in selected cohort (N) dynamically displayed - Information icons to define visualized concepts and metrics. | Increased interpretability and explainability of visual data elements | Transparency |
- Dynamic searches and cohort selection - Instant visualization update - Quick screen loading - Capability for EHR integrated auto-search for the current patient’s record. | Instant screen updates and intuitive navigation | Latency |
- Context aware and personalized filtering of data cohort and data visualizations. - Optional search elements for tailored cohort selection. - User (physician) knowledge integration in data elements design. - Intuitive mapping of element labels to real world practice. | Relevant data elements selection and display | Accuracy |
- Comprehensive visualization of data elements. - Patient provider communication enhancement with visual data evidence for treatment options. - EHR integration capability for increased utility. - User-centered evaluation of design. | Reduced burden of gathering data evidence | Usefulness |