Tools, Technologies and Training for Healthcare Laboratories

Guest Essay

Mass Diabetes Diagnoses in the UK: Lessons for all labs

55,000 NHS patients in the UK were misdiagnosed with diabetes due to instrument bias. Is there anything labs can learn from this?

Lessons for Human and Veterinary Laboratories from a Recently Human Laboratory Problem

 

February 2026

Kathleen P Freeman, DVM, BS, MS, PhD, DipECVCP, FRCPath, MRCVS
European Veterinary Specialist in Clinical Pathology
RCVS Specialist in Veterinary Pathology (Clinical Pathology)

 

The Problem

In the United Kingdom, there was a shift in results with the Trinity Biotech assay for HbA1c (positive bias) resulting in estimated mis-diagnosis of 50,000-55,000 people as either pre-diabetic or diabetic. These diagnoses alter insurance premiums (higher!) and may result in driving license restrictions and/or loss of license in some cases, as well as causing stress, significant interventions and treatment costs. The positive bias was detected and reported to the Medical Health Regulatory Agency in EQA results and in individual laboratory statistical QC as an ‘intermittent issue’ as early as mid-April and mid-July 2024.

Background

Two common methods for evaluation of HbA1C in venous blood from human patients in the UK are the Trinity Biotech system (Menarini) andthe  Tosoh system.
Both are HPLC systems with Tosoh measuring HbA1C directly and Trinity Biotech measuring all glycated hemoglobins with an algorithm for HbA1C determination. Both are licensed for use in UK and assured to provide equivalent results.

News Reports

Sky News picked up on the problem and highlighted it in September 2024 ((Link: Thousands of patients contacted over 'wrong' blood test results that may have misdiagnosed some as diabetic | UK News | Sky News)

However, major news headlines from multiple sources only appeared in Autumn of 2025 as ongoing investigations of root causes highlighted the complexity of the problem:

 

Further Investigations in the UK

Three Field Safety Notices (FSNs) were posted by Trinity Biotech in 2024..

Copies of the FSNs issued by Trinity Biotech are available at this link: Field Safety Notices: 30 September to 4 October 2024 - GOV.UK

The 30 September 2024 FSN indicated that 8 complaints of positive bias had been received and that root cause investigation was ongoing. Investigation focused on (1) Sample carryover and (2) Calibrators. The FSN emphasized the need for preventive maintenance at 15,000 injections or 6 months. Advice included that the probe needle guide should not be used for whole blood samples and that whole blood samples should not be pre-mixed. It recommended bracketed QC at start, end and every 50th sample during a run/shift. Continuing review of calibrator performance was indicated to be underway. There was a statement that: Since A1c test is not the only test used for glucose monitoring in a patient serious harm due to this issue is not expected.

The 07 October 2024 FSN indicated that sample carryover was being addressed through preventative maintenance and, where required, customer training. After an extensive review of customer and our own internal data, Trinity Biotech was validating a potential root cause and contributory factors for discrepant results observed in some customer labs. Additional guidance was included for service and semi-annual preventive maintenance at 15,000 injections or 6 months and annual preventive maintenance at 30,000 injections or every 12 months. Trinity Biotech reported completing testing of retained samples of non-expired calibrator lots: all lots passed release criteria when running the IFCC HbA1c Calibrator sets. A caution was included not to pre-mix whole blood samples.

The 25 October 2024 FSN that Information about possible bias in EQA and positive bias in patient sample results, as well as disagreement in results between 2 Trinity Biotech analyzers in the same facility had been received. Root Cause Investigations indicated a combination of factors:

  1. Probe needle guide should be removed –its presence promotes carryover
  2. Lot-to-lot variation in calibrator value assignment: value assignments for all HbA1c calibrator lots have met Trinity Biotech Quality Control specifications. Value assignments for recent calibrator lots have trended towards the positive limit whereas previously they were trending towards the negative limit. The change in negative to positive trend may be detected by some labs when the Premier Hb9210™ HbA1c analyzer is not in optimum condition.
  3. Trinity Biotech will implement statistical process control (SPC) improvements in our HbA1c Calibrator and HbA1c Control value establishment processes. Trinity Biotech will coordinate with A. Menarini Diagnostics to roll out a comprehensive training program for the UK.

In November 2024 a new calibrator was issued.

Additional investigations by various committees found a number of issues thought to be underlying or contributing to the ongoing production of erroneous HbA1c results:

POSSIBLE UNDERLYING OR CONTRIBUTING CAUSES FOR ONGOING PRODUCTION OF ERRONEOUS HbA1c RESULTS

  • Lack of clear responsibility for stopping testing when errors have been identified and reported to the manufacturer
  • Reliance on a single test with correlation with other clinical and laboratory results
  • Lack of laboratory-specific Key Performance Indicators using patient data Insufficient characterization or specifications for lot-to-lot variation evaluation by manufacturer and/or laboratory
  • Lack of use of third party independent control materials
  • False sense of security with accreditation (ISO 15189) Lack of use of median patient results data for short, medium and long-term monitoring Insufficient information for some EQA programmes
  • Lack of adherence to maintenance recommendations of the manufacturer
  • Lack of or inadequate training of laboratory personnel
  • Lack of rerunning patient samples following corrective action when an out-of-control result was obtained
  • Lack of communication between laboratories, manufacturer, MHPRA, and EQA providers
  • Lack of criteria for evaluating and maintaining harmonization of assays across NHS Trusts
  • Lack of rapid followup, notification and retesting of some patients
  • Lack of rapid investigation for possible positive bias in other laboratories using the same instrument
  • Lack of clarity in definition of ‘patient harm’ (death or disability?)

An interesting retrospective evaluation of patient coding within the NHS records found an unexpected spike in diagnoses of pre-diabetes and diabetes around July-August 2024, coinciding with this problem. When separated out according to the National Health Trust in England and the instruments used in their associated laboratories, it was evident that the spike was associated with the Trinity Biotech assay, rather than the Tosoh assay.

 

Implications for Veterinary Laboratories

The complexity of the problem and implications for veterinary laboratories are ominous since veterinary laboratories have similar or more severe problems with laboratory quality management, as found in the Westgard 2017 Survey of Veterinary Laboratory Quality Management and from one of the authors (KF) ongoing conversations with and experience in teaching the VIN Laboratory Quality Management Course, provided in conjunction with the European College of Veterinary Clinical Pathology Laboratory Standards Committee.

Some ongoing issues in veterinary medical laboratories (in-clinic, as well as university, commercial and reference laboratories) are summarized below:

ONGOING ISSUES IN VETERINARY MEDICAL LABORATORIES THAT MAY LEAD TO REPORTING ERRONEOUS LABORATORY RESULTS

  • Common absence of use of third party independent control materials
  • Lack of or infrequent calibration verification
  • Frequency of QC usually limited to once per day (at startup)
  • Manufacturer advertising and advice that contradicts expert body opinions and recommendations
  • Infrequent use of QC validation to ensure high probability of error detection with QC
  • Lack of or infrequent evaluation of lot-to-lot variability of reagents and/or calibrators
  • Infrequent use of clinical pathologist consultation and/or human expert models for veterinary laboratory quality by in-clinic and commercial laboratories
  • Lack of manufacturer transparency about on-board Quality Control and its ability to detect errors
  • Absence or infrequent availability of EQA programmes (commercial or self-designed) using species-specific or commutable materials
  • Lack of or variable oversight by regulatory agencies or EQA (varies by country)
  • Lack of centralized instrument evaluation and licensing for use in veterinary medicine
  • Ability to obtain hospital or clinical accreditation without acrural of points for in-clinic laboratory quality systems, despite requirements for some inclinic laboratory testing
  • Lack of knowledge/education about inclinic laboratory testing and quality systems
  • Lack of ISO accreditation for MEDICAL laboratories (15189) – instead 17025 is applied (for calibration laboratories)
  • Lack of or infrequent use of Key Performance Indicators using patient data
  • Lack of knowledge or application of instrument performance assessment(partial validation) prior to implementing inclinic and/or laboratory testing
  • Laboratory quality is not yet ‘on the radar’ of state boards of veterinary practice, AVMA accreditation of education or the AVMA
  • Lack of published peer-reviewed information about various laboratory tests, artificial intelligence and other aspects of laboratory testing
  • Common provision of consensus guidelines that provide strict cut-offs without acknowledgement of ‘gray zone’ results, effects of bias, possible interferences and absence of harmonization across different instruments, methods and laboratories
  • Lack of rerunning patient samples done before a current QC failure (out-of-control QC)

 

Laboratory Philosophy

In the final reckoning, the quality culture of the laboratory will be the factor that most effects whether or not there is a low, moderate or high risk of production and/or reporting of erroneous results.

The quality literature for human laboratory testing is vast. In veterinary internal medicine, comparative pathology and pathophysiology is important in understanding the differences in species and assay applications. The veterinary internal medics frequently peruse the human literature to understand and become knowledgeable about the latest human research and developments. The veterinary laboratory community does not appear to do comparable parallel research regarding the art and science of laboratory quality.

 

The Importance of Laboratory Testing to Veterinary Medicine

Laboratory test results provide information that impacts a large number of cases in diagnosis, treatment, and monitoring of veterinary patients

Laboratory testing requires special knowledge that is not widely taught, if at all, in veterinary schools

Laboratory testing and quality systems may not be addressed at multiple levels due to insufficient knowledge, insufficient staffing and/or insufficient consultation opportunities

In veterinary medicine, we do not currently have a regulatory framework for ensuring veterinary laboratory quality. This may or may not be a ‘good thing.’

We acknowledge that owners want the same attention to patient safety and result accuracy as that for their own laboratory testing. Owners assume that laboratory results will be accurate and reliable and that veterinary laboratories have a quality system for in place to ensure this.

No laboratory will advertise that they provide mediocre or poor laboratory quality, regardless of the evidence
provided or lack of evidence to support high quality claims.

Many resources exist for quality for human laboratory testing and can be directly applied and adapted to veterinary testing.

We acknowledge that laboratory quality has associated costs. These costs are considered by the authors to be manageable compared to costs associated with litigation, and unnecessary or high risk treatment, or ongoing monitoring based on interpretation of erroneous results. The ‘proof of quality’ lies with the laboratory. Show the documentation and data that supports your claims for quality.

 

Conclusions 

The recent HbA1c problem in human medical laboratories in the UK will undoubtedly have repercussions for years to come and has resulted in formation various committees and boards to investigate and assure that similar problems are not repeated.

There are many lessons that can be learned from this example for veterinary laboratories. There is great variation in the quality systems applied in veterinary in-clinic, university, commercial and reference laboratories. Current accreditation programs for veterinary hospitals/clinics and for veterinary university, commercial and reference laboratories do not address many aspects of veterinary laboratory quality. Accreditation alone does not provide assurance of accurate and reliable laboratory results. Laboratory quality is not yet ‘on the radar’ of veterinary state or national boards, university laboratory accreditation, practice/hospital accreditation programs or ‘the public’. A need for ongoing education about laboratory quality exists.

There is a need for experts in laboratory quality to consult with and advocate for quality systems in all laboratories producing results for veterinary patients. Veterinary education and continuing professional development/continuing education should prioritize provision of information about laboratory quality systems and laboratory tests to develop the expertise of general practitioners and specialists with regard to laboratory testing. There needs to be better communication between laboratories, manufacturers of instruments, reagents and kits, and users of these.

We hope that this communication will provide ‘food for thought’ for veterinarians, veterinary clinical pathologists, laboratorians and the organizations within which they are employed regarding laboratory quality systems and their application in veterinary laboratories.

 

 

 

Let us know what you're interested in!

Please use this form to request more information about.

Westgard Products and Services.

Invalid Input
Invalid Input
Invalid Input
Invalid Input