101: Basic Evaluation with Visibility Rules
Overview
This tutorial walks through building a clinical decision support evaluation using the URECA (Adult Urinary Retention Evaluation & Catheterization Algorithm) as the working example. The evaluation uses 7 questions and 5 conditionally displayed results to guide clinicians through a branching decision pathway.
The tutorial demonstrates the five-step development methodology — from initial component breakdown through testing — showing how structured planning makes EVAL evaluation development faster and more reliable.
Evaluation Structure
- 7 questions with branching logic
- 5 results displayed conditionally based on user responses
- Decision pathways determined by symptom assessment and clinical criteria
Five-Step Development Methodology
Step 1: Break Down Components
Start by identifying the workflow elements. Map out the questions users will answer and the results the evaluation must produce. For the URECA example, this preliminary analysis reveals the question/result structure needed to support clinical decision-making.
Key principle: Simplify choices. Use yes/no responses when the choices don't drive distinct logic paths. Minimize the number of options wherever possible.
Step 2: Diagram the Process Flow
Convert your workflow diagram into logic formulas that express if-then relationships between user inputs and outcomes. For each result, write out the conditions that must be true for it to display.
Key principle: Every component (question, choice, result) must be uniquely identifiable so that visibility rules can reference them accurately.
Step 3: Peer Review
Before building in EVAL, share your component breakdown and process flow diagrams with a clinical colleague for review. Early peer review catches:
- Translational inaccuracies between the clinical source and your workflow
- Logic gaps (outcomes with no decision path leading to them)
- Computational errors in scoring formulas
- Design flaws in the question flow
Investing time in peer review before building saves significant rework later.
Step 4: Build in EVAL
Using your validated diagrams as a blueprint, build the evaluation in the EVAL Builder:
- Create the evaluation and fill in the General tab metadata
- Add sections as needed
- Create each question with its choices and values
- Define result formulas
- Apply visibility rules to sections, questions, and results based on your process flow diagrams
Step 5: Conduct Testing
After building, test every possible logic pathway:
- Use EVAL's built-in test scenarios to define input combinations and expected outputs
- Run each scenario and verify the correct results display
- Manually validate edge cases where logic paths converge or diverge
- Confirm that all visibility rules behave as intended
Key principle: Account for every possible outcome combination. Gaps in decision pathways can lead to clinical errors.
Design Principles Summary
| Principle | Guidance |
|---|---|
| Simplicity | Minimize choice options; prefer yes/no where distinct logic isn't needed |
| Unique identification | Every component must be clearly distinguishable for accurate logic connections |
| Completeness | Every possible input combination must lead to a defined outcome |
| Early peer review | Review diagrams before building to catch errors cheaply |
Tutorials
End-to-end walkthroughs that teach you to build clinical evaluations, manage patient populations, and configure your organization in EVAL.
Build a PHQ-9 depression screener
Create a complete PHQ-9 Patient Health Questionnaire from scratch — with sections, scored questions, a severity formula, and conditional results.