Skip to content

roadmap.md

Latest commit

 

History

History
164 lines (115 loc) · 5.3 KB

roadmap.md

File metadata and controls

164 lines (115 loc) · 5.3 KB
Raw

ABC of ECL - Enhancement Roadmap

Target Audience: Clinical informaticians
Approach: Open exploration tool
Focus: Medication data, real-world prescribing/dispensing scenarios

1. ECL Query Anatomy Panel (Depth of Explanation)

Add an interactive breakdown panel that explains each part of the generated ECL in real-time.

What it does

When a query is built, show a collapsible "Explain this query" section that breaks down each component (operator, concept ID, attribute, value).

Why it helps

Informaticians learn why << is used vs <, why certain attributes are selected, and how refinements chain together.

Example

For << 763158003 |Medicinal product| : 127489000 |Has active ingredient| = << 387517004 |Paracetamol|, show:

Component Explanation
<< Includes the concept AND all descendants (subtypes)
763158003 Root concept for medicinal products in SNOMED drug model
: Refinement operator - filters products by attribute
127489000 This attribute links a product to its active substance
= Value constraint - the attribute must equal this value
<< 387517004 Paracetamol and all its subtypes (salts, combinations)

Effort

Medium

2. Clinical Scenarios Section (Real-World Use Cases)

Add a dedicated Scenarios page with guided, practical use cases.

Planned Scenarios

Scenario ECL Focus
Dose-based prescribing Finding products with specific presentation strengths (e.g., "500mg tablets of paracetamol") using strength numerator/denominator attributes
Prescribing → Dispensing link Navigating VTM → VTMP → AMP → AMPP hierarchy; showing how a prescriber's intent maps to dispensable products
Substance allergy exclusion "Patient is allergic to penicillin - find all products to exclude" using Has active ingredient = << 764146007 |Penicillin|
Therapeutic substitution Finding equivalent products by therapeutic class or mechanism
Formulary membership Using ^ (member of) to query reference sets representing approved formularies

Scenario Structure

Each scenario would include:

  1. Clinical context - Why this matters in practice
  2. The ECL approach - Which operators and attributes to use
  3. Step-by-step query building - Guided construction
  4. Interpretation of results - What the output means clinically

Effort

Medium-High

3. Multi-Server Support (National Applicability)

Architect the app for multiple national terminology servers.

Features

  • Server selector in the UI (dropdown or settings panel)
  • Server status indicator showing connectivity and available content
  • Graceful handling of edition differences (e.g., UK dm+d vs AU AMT vs International drug model)

Pre-configured Endpoints

Country Server Endpoint
🇦🇺 Australia CSIRO Ontoserver https://r4.ontoserver.csiro.au/fhir
🇬🇧 UK NHS Terminology Server TBD
🇮🇪 Ireland HSE NMPC FHIR Server TBD
🇨🇦 Canada Infoway Gateway TBD
🌍 International SNOMED Snowstorm API TBD

Considerations

  • Different national extensions have different drug models
  • Some servers require authentication
  • Response times and rate limits vary

Effort

High

4. Query Comparison Tool (Deepen Understanding)

Allow users to run two queries side-by-side and compare results.

Use Cases

  • Compare < 763158003 vs << 763158003 (with vs without self)
  • Compare <! (children) vs < (all descendants)
  • See overlap/difference between two ingredient queries
  • Understand AND vs OR behavior

Visual Output

  • Result counts for Query A, Query B, and intersection
  • Simple Venn-style indicator
  • "Only in A", "Only in B", "In both" result tabs

Learning Value

Informaticians immediately see the practical impact of operator choices through concrete results.

Effort

Medium

5. ECL Playground with Live Validation (Expert Mode)

Enhance the "Custom Query" feature into a full playground for power users.

Features

Syntax Highlighting

Colorize operators, concept IDs, and terms in the editor.

Live Syntax Validation

Catch errors before execution - show inline warnings/errors as user types.

Auto-complete Suggestions

  • Attribute names after typing :
  • Operators after typing <, >, ^
  • Recent/common concept IDs

Query History

Save and recall previous queries within a session.

Shareable Links

Generate URL with encoded ECL for sharing queries with colleagues.

Effort

Medium-High

Priority & Implementation Order

Priority Enhancement Rationale
1 ECL Query Anatomy Enhances every existing template immediately
2 Clinical Scenarios High educational value, addresses core user needs
3 Query Comparison Deepens understanding of operators
4 ECL Playground Power-user feature, builds on existing custom query
5 Multi-Server Support Foundational change, requires research on endpoints

Status

  • ECL Query Anatomy Panel ✅ (v1.3.0 - December 2025)
  • Clinical Scenarios Section
  • Query Comparison Tool
  • ECL Playground
  • Multi-Server Support

Last updated: December 2025