Project Overview
A UK cancer centre treating 4,200 patients per year wanted to build a clinical decision support (CDS) tool for...
Technology Stack
Compliance & Standards
The Challenge
A UK cancer centre treating 4,200 patients per year wanted to build a clinical decision support (CDS) tool for oncology MDT (Multi-Disciplinary Team) meetings — aggregating patient data from 6 clinical systems (EPR, radiology PACS, pathology LIS, genomics system, clinical trial management, and pharmacy) into a single pre-MDT briefing document, with AI-powered treatment pathway recommendations based on NICE clinical guidelines. DCB0129 clinical safety, UK GDPR Article 9, and DTAC were mandatory. Budget: £140,000.
Our Approach
Clinical Data Aggregation
- HL7 FHIR R4 integration with EPR (System C Medway).
- DICOM API integration with PACS (radiology images via DICOMweb).
- HL7 v2 parsing for pathology LIS.
- Custom REST API integration for genomics system.
- The aggregated patient profile is compiled 24 hours before MDT meeting — oncologist reviews the auto-generated briefing.
Highest risk in the Hazard Log
- incorrect treatment recommendation (mitigated by displaying as "suggested pathway for discussion" — never as a decision).
- Clinical Safety Officer (CSO): all AI outputs labelled "Clinical Decision Support — not a clinical decision."
Human review
MDT clinician must actively accept or modify the recommendation before it enters the patient record.
NICE Guideline Integration
NICE Technology Appraisals and Clinical Guidelines for 8 cancer pathways structured as JSON rule sets.
Pathway recommendation engine
treatment pathway matched against patient profile (stage, histology, genomic markers, comorbidities, prior treatment).
Display
matching NICE guideline reference, relevant clinical trial options (from CTMS integration), and contraindications.
UK GDPR Article 9 and Clinical Governance
- Special category health data (cancer diagnosis and treatment data).
- Explicit consent already obtained at oncology registration — CDS falls within the same processing purpose.
Article 22
- automated recommendation clearly labelled, human oversight mandatory, decision audit trail.
- IG toolkit compliance review by Trust DPO.
The Results
Live at 20 weeks, £132,000 — under budget.
DTAC approved all 5 domains (first submission).
MDT preparation time: 4.5 hours per consultant per week → 1.2 hours (74% reduction).
NICE guideline adherence: automated check catches 100% of cases where proposed treatment deviates from NICE pathway (previously estimated 23% of deviations missed).
Clinical trial enrolment: 18% increase (system surfaces eligible trials automatically).
DCB0129: zero clinical safety incidents in 12 months of operation.
“The DTAC first-submission pass on a clinical AI system was genuinely unprecedented in our experience. The DCB0129 approach — treating the AI recommendation as a discussion prompt, never a decision — was the right clinical safety design. The 74% reduction in MDT prep time has given our oncologists back half a day every week." — Clinical Director, UK Oncology Centre (name withheld)”
Project Details
Related Case Studies
View AllNHS-Connected Mental Health App — DTAC Approved
NHS-Connected Mental Health App — DTAC Approved. A UK mental health technology company needed to build a patient-facing digital therapy support app connected to NHS syst...
Workforce Management SaaS — UK Healthcare Staffing Agency
A UK healthcare staffing agency placing 2,800 NHS-registered nurses, doctors, and allied health professionals ...
NHS Outpatient Referral System — FHIR R4 Integrated
An NHS Foundation Trust needed to replace a 2007 paper-based outpatient referral system. GPs referred patients...
Ready to Transform Your Business?
Let's discuss how our technical expertise can help you achieve remarkable results.