Concerns have been raised by GPs in Surrey about the validity and accuracy of the INR assay at the East Surrey Hospital laboratory. The issue concerned three patients who, in late April (two in one week from the same practice), had blood collected for INR from a primary care setting. The samples were analysed at the East Surrey Hospital laboratory and were found to be high (>20 in most cases) but on repeat testing of a fresh sample taken at a later time were found to be within therapeutic limits.
The Haematology department has conducted a thorough investigation into the performance and accuracy of the analysers and methodology. In all three cases, the initial high result was confirmed by repeat testing before being reported (this is our Standard Operating Procedure). The reproducibility of the results indicates that the readings were not due to a processing error. Both analysers are functioning correctly and performance on Internal Quality Control is within stated tolerances showing no significant bias or random error. External Quality Assurance data demonstrates good correlation of results with similar models of analyser in other laboratories. Resulting from this investigation the Trust wishes to reassure all GP colleagues that there is no systemic problem with INR analysis at the East Surrey laboratory.
A data search of the pathology computer system for the six months to end of April 2013 identified eight patients with INR results >20 out of 66,246 tests carried out. Four samples were from within the hospital and four from primary care, of which three from primary care were within two weeks of each other. This is an incidence of approximately 1 in 8,300. The fact that these unexpectedly high results are so rare means that they should be interpreted with caution and in the light of the clinical picture. If the result does not correlate with the clinical findings, repeat sampling is advised before deciding the patient management.
Whilst it is impossible to be completely certain, it is well known that INR analysis is very susceptible to poor sampling technique and order of draw (the INR sample should be drawn first). This is the most likely cause of repeatable but erroneous results on the same sample, and is the most likely explanation for the cases in question.
In future, and after discussion with the Consultant Haematologists, all results where the INR is >20 will be phoned in accordance with current laboratory practice but additionally reports will contain the following comment:
“INR result on this sample confirmed. However, please treat this result with caution and in the light of other clinical findings. If in any doubt please repeat.”
We hope that this has provided reassurance about the safety and quality of our INR service. As always, we are happy to discuss any further concerns you may have.
Dr Bruce Stewart
Chief of Service for Clinical Support Services
Tel: 01737 231904