Dionisio Acosta: Clinical decision support systems
-
Upload
nuffield-trust -
Category
Health & Medicine
-
view
685 -
download
1
description
Transcript of Dionisio Acosta: Clinical decision support systems
Delivering Clinical Decision Support System for Cancer Multidisciplinary Meetings: MATE
Dionisio Acosta, CHIME, UCL
Vivek Patkar, Cancer Institute, UCL
Mo Keshtgar, Department of Surgery, Royal Free Hospital
John Fox, Department of Engineering Science, University of Oxford
Outline
• Introduction• Approach• Results & Discussion• Conclusion
IntroductionCancer MDM
Cancer Multidisciplinary Meeting (MDM) is a widely endorsed mechanism for ensuring high quality evidence-based cancer treatment.
However, in a context of increased demand and accountability there are shortcomings in the current conduct of MDMs that have made them a priority of the National Cancer Action Team (NCAT) and the National Health Service (NHS).
IntroductionMDM Shortcomings
Inadequate documentation of both patient data and MDM decisions.
Missed opportunities to screen patients suitable for clinical trials.
Lack of appropriate tools for auditing and monitoring the MDM performance.
IntroductionChallenges
Harmonize multiple clinical guidelines in one framework
Generate individual treatment recommendations
User-centred application design
Integrate prognostications tools
Approach
Clinical Knowledge Base (17 CPGs)
PROformaDecision Engine
GraphicalUser Interface
ElectronicPatient Record
Audit &Performance
AccurateMDM Documentation
ClinicalTrial Screening
PrognosticationTools
MATE Middle-ware
ResultsMultidisciplinary Assistant & Treatment sElector (MATE)
Results
Currently evaluated in a randomised control trial at the same institution.
Successfully piloted at the Royal Free Hospital Breast MDM in over 1000 patients.
Concordance with clinical guidelines in 97% of cases.
Identified 60% more eligible patients for clinical trials.
DiscussionPotential Impact
• The methodology, know-how and the underlying technology can be directly applied to other cancer MDMs.
• Reducing unwarranted variations in cancer care– Promoting adherence to best practices.
– Minimizing unjustifiable deviations.
• Supporting the life-cycle of clinical practice guidelines– Documenting deviations and their corresponding justifications.
– Dissemination and implementation.
DiscussionLimitations
• Personalised patient recommendations:– Does not account for patient preferences.– Does not seamlessly integrate with prognostic
calculators.– Does not capture (local) treatment outcomes.
• Communicating Risk– Does not depicts patient pathways.– Does not harmonize recommendations with prognostic
calculators.– Not designed for patients.
DiscussionRisks for Technology Uptake
• Integration with EHR– Clinical document standard architectures (HL7 CDA,
EN13606).– Controlled terminologies.– Governance.
• Knowledge base updates– Centralised vs. distributed approach.– Governance: Who?, when?, where?, how?
• Understanding and communicating economic impact.