Guest Essay

Dr. Freeman and Alison Farr talk more about their 2008 papers on veterinary quality requirements and the Sigma performance metrics they calculated for their laboratory.

Quality Requirements,QC Validation and Sigma Metrics for Veterinary Laboratories

February 2009

Alison Farr BvetMed MRCVS
Kathleen P. Freeman, DVM, BS, MS, PhD, DipECVCP, FRCPath, MRCVS

We have been invited to present this essay about our experiences with quality requirements, QC Validation and Sigma Metrics for Veterinary Laboratories based on our paper:

Quality control validation, application of Sigma metrics, and performance comparison between two biochemistry analyzers in a commercial veterniary laboratory Alison J Farr and Kathleen P. Freeman. J Vet Diagn Invest . September 2008;20(5):536-44.

The definition of quality requirements is central to quality planning in the laboratory. In the veterinary laboratory the regulatory requirements which apply in human laboratories do not exist. To our knowledge there are no published quality requirements for common domestic species. In order to perform quality control validation within our laboratory it was therefore necessary to define a set of quality requirements for the analytes in the commonly encountered species, namely cat, dog and horse.

Quality requirements were chosen based on a review of the veterinary literature as well as on the clinical use of data for clinical decision making based on internal clinical pathology discussion within the authors’ laboratory. Quality requirements are likely to differ between veterinary laboratories based on clientele, species of interest and use of the laboratory data.

Rather than defining requirements for each species for each analyte, we selected the species with the most stringent quality requirement for a particular analyte. For example, in the horse there was a need to detect a smaller change in CK than in the dog or cat because of the propensity of this species to the development of exertional rhabdomyolysis. In this way we were able to define the single most stringent quality requirement for each analyte and use this as a basis for QC Validation.

Quality requirements were determined for high and/or low clinical decision limits, depending on the analyte. For example low levels of liver enzymes are not generally of clinical interest and so a quality requirement was only determined for high values. For many analytes, such as calcium or glucose, both low and high levels are of interest and were therefore determined.

In our study the results from 2 Olympus analysers (AU2700 and AU640) over a six month period of known stable performance were used. The internal quality control and external quality assurance data were used to generate CV and bias for each analyte. This information, in conjunction with the quality requirements, was then used to calculate a sigma value for each analyte.

In the future quality requirements may be defined for many additional species and across a more comprehensive range of analytes.

We found that:

1. The setting of quality requirements was an intensive exercise, but one that provided a clear standard for establishing and maintaining quality through QC Validation and a strategic approach to QC based on the capabilities of the analysers.
   
   
2. The differences in performance between analysers was not anticipated and underscored the necessity for QC validation for each analyte on each analyser separately.
   
   
3. The sigma metrics for the various analytes gave a good indication of the capability of analysers and reflected the approach to statistical and nonstatistical QC.
   
   

We hope that other veterinary laboratories will take on board the concepts that Dr. James Westgard has helped us understand and apply, as presented in our paper.

If you are interested in visiting more with us about veterinary applications of QC concepts and QC validation for veterinary laboratories, please feel free to contact either of us!

Alison J. Farr - alilomas AT yahoo.co.uk

Kathleen P. Freeman – kathy-freeman AT talktalk.net