Back to Swatches

NHS - Blood Test Lab Report Processing

Fabric Score 3.6
PublicHealthNational Health Service UKNon-GenAI modelUser / Operator / Front-line Staff ×1

Workflow Diagram

NHS - Blood Test Lab Report Processing workflow diagram

Fabric Score

ValueEfficacySecurityRiskExternalitiesEfficiency4.53.75.02.33.23.13.6Fabric Score

Task

Patient undergoes blood test. Once blood test report is available, the AI system uses RPA to identify trends/abnormalities. If bloods are classed as ‘normal’ based on guidelines, it will file away the report and can automatically inform patients via SMS.

Intent

The AI system attempts to reduce clinicians having to review ‘normal’ bloods and filing it away, reducing workload.

AI Workflow

Input

Blood test report.

Process

Based on decision trees developed with clinical input, the AI system will class a result as ‘normal’ or ‘abnormal’.

Output

The ‘normal’ lab reports will be filed to patient notes and patient informed of outcome.

Human Oversight Level

Conditionally Autonomous AI

Institutional Oversight Examples

  • •Data protection regulation
  • •Clinician oversight
  • •MHRA - medical device regulation

Risk

A ‘normal’ result may be filed away - but this may not be normal for the patient.

Output Modification Telemetry

  • Factual errors / hallucinations90%
  • Poor usability (unclear language)10%

Transversal Metrics

Grouped by Fabric dimension.

Efficacy

Accuracy60 %

Efficiency

Modification Rate30 %
Modification Time15 min/output
Verification Time5 min
Rejection Rate10 %
Operational Friction77 hrs/wk
Implementation Overhead200 hrs
Governance Overhead200 hrs
Time to Launch10 wks

Value

Effort Reduction90 %

Risk

Reliability2 incident/month
Autonomy99 %