Pathology Automation

Pathology Automation

80+ biomarkers. 17 test categories. Gender and age-specific reference ranges. Intelligent filing with zero compromise on accuracy.

OVERVIEW

What is Pathology Automation?

The Pathology Automation software is a Class I Software as a Medical Device (SaMD) designed to automate the administrative review and filing of pathology results determined to be 'normal'. It operates as an autonomous Robotic Process Automation (RPA) bot that securely logs into your Electronic Health Record system, analyses patient pathology data using strict rule-based logic incorporating numerical thresholds, trend analysis, and keyword verification — then performs one of two actions:

File Results meeting all criteria for 'normal' are filed with the comment "Normal - No action needed"
Skip Results that don't meet criteria are left in the clinician's workflow for manual review

With 80+ biomarker tests configured across 17 test categories, the system filters high volumes of normal results, allowing qualified healthcare professionals to focus their attention on results that are abnormal, ambiguous, or otherwise require clinical judgement. The result is faster turnaround, safer workflows, and dramatically reduced admin burden for your clinical team.

Get in Touch

Key Benefits

6 key features that make our Pathology Automation service essential for your practice.

  • 80+ Biomarkers Covered Comprehensive coverage across 17 test categories including FBC, LFTs, Renal Function, Thyroid, Lipid Profile, Bone Profile, and many more.
  • Gender & Age-Specific Ranges Evaluates results against reference ranges tailored to patient gender (e.g. Haemoglobin, Creatinine, Gamma GT) and age (e.g. PSA thresholds for ages 40-49, 50-59, 60-69, 70-79).
  • Intelligent Routing Three clear actions: File Without Comments for normal results, Skip for results needing doctor review, and Sample Unavailable when no valid sample is detected.
  • Sample Quality Detection Automatically identifies 6 sample quality issues: No sample, Contaminated, Unable to provide, Regret error, Haemolyzed, and Duplicate specimens.
  • Trend Detection Detects clinically significant trends by comparing against previous values -- such as significant drops in Haemoglobin or spikes in Creatinine -- and flags them for review.
  • 100% Accuracy Clinically reviewed workflows maintained by a Clinical Safety Officer, ensuring every result is evaluated correctly and routed appropriately.
COMPREHENSIVE COVERAGE

17 Test Categories. 80+ Biomarkers.

Every major pathology test category is covered, with gender-specific and age-specific reference ranges applied where clinically appropriate.

Full Blood Count (FBC)
15 tests
WBC, Haemoglobin, Haematocrit, Platelets, RBC, MCV, MCH, MCHC, Neutrophils, Lymphocytes, Monocytes, Eosinophils, Basophils
Liver Function Tests (LFT)
8 tests
Albumin, Bilirubin, ALP, ALT, Globulin, Total Protein, Gamma GT
Lipid Profile
6 tests
Non-HDL, HDL, LDL, Total Cholesterol, Triglycerides, Cholesterol/HDL Ratio
Metabolic / Vitamins / Minerals
8 tests
Folate, Ferritin, B12, Vitamin D, Uric Acid, Iron, Transferrin, Magnesium
Bone Profile
7 tests
Calcium, Phosphate, ALP + others
Renal Function
5 tests
GFR, Creatinine, Potassium, Sodium, Urea
Inflammatory Markers
3 tests
CRP, Plasma CRP, ESR
Thyroid Function
2 tests
TSH, Free T4
Clotting Tests
3 tests
APTT, INR, Fibrinogen
GI Tests
3 tests
Faecal Calprotectin, Helicobacter Screen, FOBT
Cardiac Biomarkers
2 tests
NT-proBNP, BNP
PSA (Prostate)
Age-based logic
Custom thresholds for ages 40-49, 50-59, 60-69, 70-79
HbA1c (Diabetes)
1 test
Glycated haemoglobin for diabetes screening
ACR
1 test
Albumin:Creatinine Ratio
Rheumatology
1 test
Rheumatoid Factor
Tumour Markers
1 test
CA125
Muscle
1 test
Creatine Kinase

How It Works

A six-step processing pipeline that evaluates every lab report with clinical precision and routes it to the right action.

  • 1. Submit Lab Report EMIS-format pathology report is received by the processing service via API.
  • 2. Check Sample Quality The system checks for 6 sample quality issues -- no sample, contaminated, unable to provide, regret error, haemolyzed, or duplicate -- and flags them immediately.
  • 3. Extract Biomarker Values Numerical values are extracted from unstructured lab report text for each biomarker present in the report.
  • 4. Evaluate Against Reference Ranges Each value is compared against gender-specific and age-specific reference ranges to determine normal, low, or high status.
  • 5. Apply Custom Logic Advanced rules are applied: age-based PSA thresholds, previous value trend detection for Haemoglobin and Creatinine, and keyword filtering for clinical terms.
  • 6. Return Structured Results The service returns a structured response with a recommended action: File Without Comments (normal), Skip (needs doctor review), or Sample Unavailable.
CLINICAL OVERSIGHT

The Merlin Dashboard

The device has no direct operational user interface — its functions are fully autonomous. The primary human interaction point is the Clinical Oversight Dashboard, designed for monitoring and auditing purposes.

Real-Time Action Log A live log of every action taken by the bot — which results were filed, which were skipped, and when.
Decision Transparency View the specific rules and logic applied to each result, so clinicians can understand and verify every decision.
Full Audit Trail Complete traceability for clinical governance — every filed and skipped result is logged with reasoning.
PROCESSING LOGIC

General Processing Rules

The core algorithm is a rule-based decision engine that categorises results based on pre-defined clinical criteria. Every rule is clinically approved during a formal hazard workshop with each organisation's Clinical Director.

1
Age Restriction Do not process results if patient age is under 12.
2
Textbook Naming Blood test names follow textbook definitions; terms like 'plasma' or 'serum' do not affect interpretation.
3
Suffix Handling Suffixes such as 'TSH (Beckman)' are ignored; only textbook definitions determine processing.
4
Error Detection If no numerical value is found or if any error (e.g., Sample Haemolysed, Duplicate result, Regret Error) is present, do not process.

Example: PSA Clinical Logic

PSA is produced by prostate cells and is specific to males. Levels increase with age due to benign enlargement. PSA is used for cancer screening and monitoring, with age-dependent reference ranges essential for accurate interpretation.

Age 40-49< 2.5 ng/mL
Age 50-59< 3.5 ng/mL
Age 60-69< 4.5 ng/mL
Age 70-79< 6.5 ng/mL
ROLES & RESPONSIBILITIES

Intended Users

The Pathology Automation software is an autonomous system that operates at scheduled times without direct user interaction. The "users" are defined by their roles and responsibilities in relation to the system:

System Configuration & Approval

The rules and logic for automated filing are established and clinically approved by senior healthcare professionals (Medical Director, designated clinicians) during a formal clinical hazard workshop for each organisation.

System Oversight

The autonomous operation is monitored by qualified healthcare professionals and trained administrative staff via the Merlin Dashboard, which provides full visibility of the bot's actions.

Clinical Responsibility

Qualified healthcare professionals (GPs, Specialist Doctors, Clinical Pharmacists, Practice Nurses) are solely responsible for the manual review, interpretation, and clinical management of all results skipped by the automation.

Safety & Compliance

TRUST & COMPLIANCE

Safety First, Innovation Second

Our pathology automation is built on a foundation of clinical safety and regulatory compliance.

UKCA Class I Medical Device

Rule 11(a) + MDCG 2019-11 guidance. Non-invasive SaMD that does not by itself drive diagnosis or therapy.

DCB0129 Compliant

Strict clinical risk management with documented hazard logs and safety cases.

Clinical Safety Officer

Safety cases approved and maintained by a registered CSO with ongoing clinical oversight.

ALARP Verified

All residual risks reduced to As Low As Reasonably Practicable. Clinical hazard workshops conducted with each organisation's Clinical Director.

THE MECHANISM

Conservative by Design

Our system prioritises safety over sophistication. Every guardrail is adjustable per practice needs, ensuring clinical governance is maintained at every step.

1
Trend Analysis (Safety) Detects clinically significant changes between results
  • Skips results if Creatinine increases ≥1.5x from previous result (AKI risk)
  • Skips results if Haemoglobin drops >10 units from previous value
  • Ensures emerging clinical trends are caught before they escalate
2
Context & Exclusions Age-based rules, PSA logic, and diabetes checks
  • Patients under age 12 are automatically excluded from auto-filing
  • PSA results use age-specific thresholds (40-49, 50-59, 60-69, 70-79)
  • Diabetes-related biomarkers flagged with appropriate clinical context
3
Safety Keywords Clinical keyword detection and flagging
  • Detects AKI (Acute Kidney Injury) keywords and flags for immediate review
  • Identifies CKD (Chronic Kidney Disease) indicators
  • 'Not Detected' rules ensure negative results are handled appropriately
4
Lab Errors & Quality Sample quality detection and error handling
  • Detects haemolysed samples and prevents filing
  • Catches null values and incomplete results
  • Identifies lab processing errors and duplicate specimens

Clinical Governance Based On

NICE CG181 Cardiovascular disease: risk assessment and reduction, including lipid modification
NICE NG131 Prostate cancer: diagnosis and management
NICE NG28 Type 2 diabetes in adults: management
NICE NG8 Anaemia — iron deficiency
KDIGO 2012 Clinical Practice Guideline for CKD Evaluation and Management
BCSH Guidelines British Committee for Standards in Haematology
BSG 2011 British Society of Gastroenterology: Management of iron deficiency anaemia
Tietz (8th ed.) Fundamentals of Clinical Chemistry and Molecular Diagnostics
BHF British Heart Foundation: Understanding Cholesterol
Synnovis Guy's & St Thomas' NHS Foundation Trust Laboratory Medicine Handbook
Oxford Handbook Oxford Handbook of Clinical Medicine (10th ed.)
DCB0129 Clinical Risk Management: Application in Manufacture of Health IT Systems
DCB0129 COMPLIANT ALARP VERIFIED CLASS I SaMD

Clinical Safety & Risk Management

The Pathology Automation system has undergone a comprehensive clinical risk management process in accordance with DCB0129. Hazard identification workshops, risk assessment, and mitigation controls ensure that all identified risks are managed to an acceptable level.

The system does not handle abnormal results or make clinical decisions — its function is limited to filtering and filing normal results. Abnormal or ambiguous results are always left for clinician review.

Identified Hazards & Controls

H-01 Medium → Low
Unauthorised Access to Sensitive Data

Unauthorised access to the system could lead to breaches of patient confidentiality and compromised data integrity.

Control Measures
  • Robust authentication protocols with encrypted credential vaults
  • Multi-factor authentication and regular security training
  • Appointed Data Protection Officer (DPO) with regular policy reviews
H-02 Medium → Low
Inadequate Access Controls & Permissions

Insufficient access controls might allow unauthorised data alterations, risking data integrity across the system.

Control Measures
  • Refined access controls — only authorised personnel access sensitive information
  • Encryption of all stored credentials and clinical rules
  • Regular security audits and access reviews
H-03 Medium → Low
Poor Data Quality Leading to Inaccurate Filing

Inaccuracies in digital patient records could lead to medical errors arising from incorrect automated filing decisions.

Control Measures
  • Practices encouraged to adopt robust quality checking processes for digital records
  • Regular audits and feedback mechanisms to improve data accuracy
  • System handles naming variations (serum/plasma) and detects lab errors automatically
H-04 High → Medium
eGFR Misclassification

Filing an eGFR result as normal when it is not could delay CKD diagnosis or miss acute kidney injury indicators.

Control Measures
  • Files only if eGFR >86; skips if creatinine increases ≥1.5x from previous result
  • Keyword detection for "AKI" and "CKD" — presence triggers automatic skip
  • Logic aligned with NICE and KDIGO guidelines for CKD staging
H-05 Medium → Low
Misclassification Due to Simplified Logic

The simplified File/Skip logic could potentially lead to misclassification of results or overlooked clinical conditions.

Control Measures
  • Comprehensive keyword and trend checks for creatinine, haemoglobin, HbA1c, thyroid, PSA, Helicobacter, and HLA-B27
  • Clinical dashboard provides transparent display of all decision logic
  • Regular clinical audits and feedback loops to ensure continued accuracy
H-06 Medium → Low
Data Breach via Unauthorised Process Access

A data breach of the automation process due to unauthorised access could expose sensitive patient information.

Control Measures
  • Regular data security and protection training for all staff
  • System operates within the secure environment of the healthcare organisation's network
  • Bot securely logs out and resets session after each run
Clinical Safety Officer Statement

Following rigorous evaluation, the updated system (Version 1.2) demonstrates significantly improved clinical safety through simplified yet highly robust automated pathology management logic. With stringent keyword and trend checks, clear clinical oversight via dashboards, and regular clinical audits, patient safety and clinical efficacy are comprehensively safeguarded. This version aligns with best clinical practices and meets DCB0129 compliance effectively.

Important This automation does not handle abnormal test results, which are flagged for clinician review. The system processes normal blood test results based on predefined criteria agreed with the Clinical Director of each organisation. The use of this automated system must always be overseen by qualified healthcare professionals who are responsible for interpreting results and making clinical decisions.

Need the Full Safety Documentation?

Request the complete Clinical Safety Case Report, Hazard Log, and Process Definition Document for your ICB or procurement process.

Questions about clinical safety? Email our Clinical Safety Officer

80+
biomarker tests configured
17
test categories covered
6
sample quality checks
100%
filing accuracy maintained
WHY US?

Why Choose ApolloIQ's Pathology Automation?

  • Clinically Validated: Workflows reviewed and maintained by a Clinical Safety Officer.
  • Reduces Admin Burden: Automatically files normal results, freeing clinician time for patients needing advanced care.
  • Gender & Age Aware: Reference ranges adapt to patient demographics -- not a one-size-fits-all approach.
  • Trend Detection: Catches clinically significant changes between current and previous results before they reach a doctor.
  • Sample Quality Screening: Identifies six distinct sample quality issues so they never get filed as valid results.
  • Keyword Filtering: Checks for specific clinical terms to ensure results are routed appropriately.

Unlike rigid systems or slow manual processing, our automation is flexible, safe, and constantly refined by real NHS clinicians. You get faster turnaround, safer workflows, and reduced pressure on your clinical team -- all from day one.

PRICING

Simple, Transparent Pricing

No hidden fees. No complex contracts. Just straightforward pricing that saves you thousands.

Per Patient Per Year
£0.30
billed annually
Example: 10,000 patients x £0.30 £3,000/year
That's just £250/month
  • 6-month initial contract
  • Then flexible month-by-month
  • No setup fees
  • Includes full support
Get Started