Kathleen Freeman
Photo courtesy of Dr. Kathleen P. Freeman
A veterinary clinical pathologist with expertise in laboratory quality control, Dr. Kathleen P. Freeman, pictured with a family dog, Nikita, recommends that veterinarians take time to understand the limitations of diagnostic tools.
Listen to this story.
Few tools are as fundamentally important to the practice of veterinary medicine as diagnostic tests. Yet the crucial results they provide — often in life-and-death scenarios — aren't always as trustworthy as this clinical pathologist would hope.
Many practitioners might be surprised to learn how lightly regulated the veterinary diagnostics field is compared with that of human medicine.
Veterinary laboratories, whether inside clinics or external "reference" laboratories, aren't subject to the Clinical Laboratory Improvement Amendments, which require all human-health laboratories in the United States to be certified. Similarly, instruments and devices marketed for veterinary use aren't subject to regulations and testing requirements enforced by the U.S. Food and Drug Administration for human medicine. The situation is similar in many other countries.
Veterinarians are left, essentially, to trust what product manufacturers and service providers tell them.
Some accreditation programs for veterinary hospitals encourage or require the testing of in-clinic laboratory equipment. But such testing isn't mandatory for all clinics, regularly monitored or standardized.
The lack of oversight may make it difficult for veterinarians to determine the capability and quality of devices, whether existing or new, and how to use them correctly. Even veterinary reference laboratories face validation challenges, since staff may lack formal training in laboratory techniques and quality, or have been persuaded to disregard such training by managers with limited expertise or enthusiasm.
Still, regulatory oversight is not a magic solution. Mandatory programs invariably add cost that is passed on to the veterinarian and, ultimately, to the owners of their patients. Enforcement of regulations usually is aimed at excluding only the most obviously inferior laboratory products from the market, and compliance alone is not the solution to achieving the level of quality expected by veterinarians and their clients worldwide.
In veterinary medicine, we often look to the comparative literature about human medical conditions and practices and how it can be applied to veterinary species. However, despite abundant literature about human laboratory testing and quality assurance, insufficient progress has been made in applying quality systems within veterinary in-clinic and reference laboratories.
In instances where equipment used in veterinary medicine originally was developed for human use, manufacturers usually have a technical support team or department that provides technical expertise to veterinarians. Sometimes, though, that guidance is not readily accessible, is backed by limited support capabilities or does not have an educational emphasis. Manufacturers accustomed to laboratories for human medicine may assume that a certain knowledge level is present (without any evidence) or that "others" will take care of educational efforts.
Specifically, veterinary diagnostic testing is characterized by:
- Lack of sufficient training of veterinarians and veterinary technicians/nurses about in-clinic laboratory testing and laboratory quality system planning, implementation and maintenance.
- Lack of explicit instructions from product and service providers that are suitable for veterinary users with limited knowledge of testing systems.
- Lack of information provided by manufacturers about their products (reagents, kits, instruments, devices), including peer-reviewed published validation studies and explicit information about the population and species for which validation has or has not been conducted.
- "Creative use" of instruments or devices by veterinary in-clinic laboratories with specimens, species or patient age groups for which use has not been validated.
Consequences of these failings include delays in making accurate diagnoses, unnecessary costs to animal owners and potential risks to patient health. Case examples based on real-life scenarios are provided further down.
The need to step up
Veterinarians and manufacturers both must accept responsibility for improving the existing situation. Manufacturers must contribute to better validation studies and better communication with veterinarians. Veterinarians need to learn more about laboratory testing, and the strengths and limitations of various laboratory products.
Veterinarians also must incorporate in-clinic quality assurance and quality control to ensure the accuracy and reliability of their results in making diagnoses, monitoring disease progression and response to treatment. Even veterinary clinical pathologists may be deficient in their knowledge of what constitutes laboratory quality.
To address the problems and information gaps, manufacturers, general practice veterinarians and expert veterinary groups such as the American Society for Veterinary Clinical Pathology (ASVCP) and European Society for Veterinary Clinical Pathology should collaborate to:
- Develop widespread training programs for veterinarians and veterinary support staff about in-clinic laboratory testing and laboratory quality system planning, implementation and maintenance.
- Provide to veterinary teams detailed instructions in the use of instruments and other products, including information such as applications for which they are not suitable.
- Provide up-to-date peer-reviewed publications and peer-reviewed "white papers" presenting information about various laboratory products.
- Encourage publication of peer-reviewed validation studies.
Heartening developments
At a veterinary conference in March called Animal Health, Nutrition and Technology Innovation Europe 2023, diagnostics company Siemens Healthineers sponsored a roundtable forum entitled "Immunoassay application in veterinary diagnostics." Attended by manufacturers of veterinary laboratory products and veterinary laboratorians, among others, the discussion highlighted the need for validation studies and the preference that these be published in peer-reviewed veterinary journals. For their part, manufacturers highlighted their need for guidance on conducting validation studies and the process for getting studies published.
Now, Siemens Healthineers and some veterinary clinical laboratory specialists are forming a collaborative group that will provide guidelines for manufacturers, veterinarians and veterinary laboratory professionals. The guidelines will detail the process of analytical and clinical validation of their products for veterinary use, and advise on the preparation of manuscripts for peer-reviewed publication.
Separately, the Veterinary Information Network, an international online community for the profession, has been a leader in providing veterinary continuing education, including a variety of courses aimed directly or indirectly at quality in laboratory testing and interpretation of laboratory results. An existing course in veterinary laboratory quality, attended primarily by veterinary clinical pathologists and veterinary clinical pathology residents, is offered regularly. A course addressing principles of veterinary in-clinic laboratory testing and its quality, tailored for veterinary technicians/nurses and general practitioners, is in development. (Disclosure: I work for VIN as a clinical and anatomical pathology consultant.)
Few other organizations offer courses on laboratory techniques and interpretation that deal specifically with laboratory quality and quality systems. Although veterinary technicians/nurses often receive training in obtaining, preparing and doing some types of laboratory testing, there is little or no formal training in laboratory quality and quality systems.
In summary
The use of veterinary laboratory testing continues to grow throughout the world but education about the need to provide quality systems to ensure the accuracy and reliability of results has not kept pace. Veterinarians and support staff with in-clinic laboratories should take it upon themselves to request and undertake additional training to ensure they know what they're doing.
In short, rigorous oversight — whether via total quality management or laboratory quality management systems — is needed to ensure that results attain a level of trustworthiness that veterinarians, clients and patients deserve.
Case example 1
A glucometer was used in a veterinary practice to assess whole-blood glucose in a dog with clinical signs that could be attributable to low blood glucose. The glucometer result was low. However, when a blood sample was tested on an in-clinic biochemistry analyzer, the reading was normal. Despite this clue that one of the results must be erroneous, an expensive investigation for possible causes of hypoglycemia was undertaken.
But hypoglycemia wasn't the issue. The patient, it turned out, was polycythemic (overproducing red blood cells) due to a pre-existing cardiac problem, and that increased ratio of red blood cells to total blood volume, or hematocrit, can interfere with glucometer results, producing a falsely low result.
The manufacturer's package insert gave specifications for a workable range of hematocrits but did not clearly communicate this limitation of glucometer testing. Unnecessary investigations for this patient, unnecessary expense for the owner, and waste of the veterinary and technical support time could have been avoided if the in-clinic laboratory had been aware of this limitation.
Solution: The manufacturer should provide clear information for veterinarians regarding the circumstances under which glucometers should not be used. Staff at the in-clinic laboratory would have found information about such limitations if they had done due diligence. A search online turns up various publications about glucometers, including guidance from the American Society for Veterinary Clinical Pathology.
Case example 2
A recent publication, Validation of a cage-side agglutination card for Dal blood typing in dogs, compared a novel blood-typing card for dogs with an established gel-column method. The authors described their research as a validation study. But there was no information about repeatability — the process of retesting the same sample to gauge accuracy — or other work that's needed to validate a new method of blood typing.
Validation studies are multi-faceted and include tasks such establishing confidence intervals, assessing the effectiveness of quality control in detecting errors and identifying possible interferents, to name just a few. A letter to the editor, submitted by Dr. Urs Giger and myself to the Journal of Veterinary Internal Medicine, points out that this publication made it through a peer-review process without correction of this claim. True validation studies require multiple steps and should address both the analytical and clinical validation of the test.
Solution: The authors of the paper should have investigated more thoroughly the abundant literature about studies needed for validation of a qualitative test such as the blood-typing card. A good review of validation studies is provided by Westgard QC, Basic Method Validation and Verification.
Case example 3
A veterinarian had an unusual automated hematology result, with an unexpected lymphocytosis (high number and proportion of lymphocytes, a type of white blood cell). They were ready to investigate the lymphocytosis but decided to sound out a VIN consultant first. The consultant suggested that the in-clinic analyzer might have misclassified cells, and advised examining the blood smear to determine whether the cells present reflected the analyzer cell counts. It turned out that the automated hematology analyzer had classified cells as lymphocytes when they were not.
Solutions: While the mistake was caught in time to avoid an unnecessary investigation of a lymphocytosis, the in-clinic laboratory could have followed ASVCP recommendations for point-of-care testing and reviewed the blood smear as part of quality assurance to confirm the automated hematology results.
Further, a search of the veterinary literature would have shown that misclassification of some types of cells is not uncommon with automated hematology analyzers, and that users should correlate automated results with blood smear morphology. The manufacturer's user manual recommends correlation with dot plots/graphs and blood smear morphology to confirm automated results; this recommendation should not be ignored.
Case example 4
Results of a blood test on a healthy older dog as part of an examination for renewal of heartworm medication came back from an outside laboratory showing that the dog might have polycythemia (too many red blood cells). Concerned, the veterinarian posted about the case on a VIN message board. The responding VIN consultants pointed out that the proportion of red blood cells in the blood — indicated by a hematocrit reading of 92% — was outside that expected in a healthy animal free from clinical signs.
Further examination of the report showed an unusual feature. It included an instrument "flag" indicating a spectrophotometric failure. This flag had not been investigated, resolved or commented on by the reference laboratory, and the veterinarian had not recognized that the reported flag could be influencing the results. Inquiries with the reference laboratory indicated that the spectrophotometer in the hematology instrument had a faulty light source that needed replacing. Once it was replaced and stable, accurate performance was confirmed by running a control material. The patient was retested and had results that were normal.
If the veterinarian had not contacted VIN about treatment options for the case, this patient may have undergone unnecessary treatment based on an erroneous result.
Solutions: The reference laboratory should have done appropriate investigation of the extremely high reported hematocrit value and instrument flag prior to releasing the results for reporting. Investigation of instrument flags is a part of any laboratory quality system. The veterinarian should have recognized the laboratory result was inconsistent with the clinical appearance of their patient, noted the instrument flag in the report and called the laboratory for an explanation and a rerun of the sample. Veterinary schools and postgraduate education programs need to put greater emphasis on "lab errors" as a differential diagnosis and instill more skepticism of in-clinic and reference laboratory results.
Dr. Kathleen P. Freeman is a U.S.-trained veterinary clinical pathologist with experience as an educator, laboratory director and senior clinical pathologist. She has worked in academia and commercial veterinary laboratories in the U.S. and United Kingdom. Freeman currently is a clinical pathology consultant for the Veterinary Information Network, working remotely from her home in Scotland. She has special interests in laboratory quality, cytology, equine medicine and pathology, and veterinary medical specialty competency-based education.