Clinical Decision Support (CDS) Models and Frameworks

1. Introduction

• Definition: CDS is a process enhancing health decisions with clinical knowledge and patient info to improve healthcare.
• CDS System (CDSS): Implementation of CDS interventions.
• Background: Despite potential, CDS has not fulfilled its promise.
• Commentary Development: Written by experts invited by the organizer (RG).

2. Purpose of Commentary

• Origin: Request to comment on Medlock et al.'s model.
• Medlock et al.'s Model:
  • Two-stream info flow model: clinical stream and cognitive-behavioral stream.
  • Measures effectiveness based on clinical care impact and user impact.
• Discussion Focus: Broader aspects of CDS, not just Medlock et al.'s model.
• Field Evolution: Since late 1950s, varied approaches but limited adoption.
• Challenges Identified:
  • Inadequate care process focus.
  • Poor clinical workflow integration.
  • Human factors and patient specificity issues.
  • Knowledge sharing difficulties between EHR systems.

3. One Model or Framework vs. Many

• Framework Value: Testing hypotheses and describing CDS dimensions.
• Workflow Integration: Not the sole focus; numerous other impediments exist.
• Example Frameworks: PROforma, Stanford EON, Asgaard/Asbru, GLIF, OMG models.

3.a. Integration/Adaptation to Workflow

• Granular Understanding of Workflow:
  • Importance of characterizing workflow at a granular level.
  • CDS Five Rights framework: Right information, people, formats, channels, times.
• Models for Smooth Integration:
  • Efforts to model workflows with care pathways and guidelines.
  • Context-awareness and dynamic context adaptation.

3.b. Construction of CDS Artifacts

• Knowledge Representation:
  • Importance of formal models for decision rules and guidelines.
• Knowledge Use:
  • Aligning CDS with user tasks and making "the right thing the easy thing".
• Cognitive Support:
  • Organizing specific actions and groupings to support cognition.

3.c. Knowledge Management and Interoperability

• Challenges:
  • Limited innovation in proprietary EHRs.
  • Interoperability barriers between EHR systems.
• Current Approaches: SMART on FHIR, CDS Hooks for external CDS access.

3.d. Cognitive Tasks and Reasoning Processes

• Decision Making Models:
  • Various models like Bayesian and hypothetico-deductive.
• Behavioral Economics: Leveraging for effective CDS intervention.

3.e. Health System Priorities and Implementation

• New Models: Evolving care models focus on coordination, continuity, predictive analytics.
• Implementation Frameworks: Need for a framework addressing purpose, approach, and evaluation.

3.f. Quality Improvement Impacts

• CDS and Quality Improvement:
  • Integrating CDS with broader quality improvement efforts.
  • CDS 5 Rights framework as a guide.

3.g. Evaluation of CDS Effectiveness

• Evaluation Frameworks: Need for detailed evaluation methods.
• Intention-Oriented Design: Assessing process and outcome intentions.

4. Conclusions

• Broader Approach Needed: Need for multiple models addressing different CDS aspects.
• Lifecycle Approach: Iterative development and evaluation cycle.
• Future Work: Addressing aspects like requirements elicitation and data quality.

Conflict of Interest

• Authors disclosed conflicts of interest, if any.

References

• Comprehensive list of references supporting the commentary.