Description of Event or Problem · 1
THE EQUIPMENT, CEQER SOFTWARE, AND VC-1 CARTRIDGES ARE USED AS PART OF THE CEQER ASSAYS TO DETERMINE THE CONCENTRATION OF RESIDUAL WHITE BLOOD CELLS IN LEUKOCYTE REDUCED PLATELET RICH PLASMA, RED BLOOD CELLS, AND WHOLE BLOOD. THE CEQER ASSAYS ARE USED AS A QUALITY CONTROL TEST FOR LEUKOREDUCED BLOOD PRODUCTS. THE RISK ASSOCIATED WITH THE FAILURE IS ACCEPTANCE OF BLOOD PRODUCT THAT DOES NOT MEET SPECIFICATIONS FOR LEUKOREDUCED LABELING. STANDARD LABORATORY PRACTICE IS TO PERFORM VALIDATIONS OF THE ASSAY INCLUDING PRECISION AND COMPARISON STUDIES. GENERALLY, A CARTRIDGE THAT IS NOT FUNCTIONING PROPERLY WOULD BE FOUND DURING VALIDATION. HOWEVER, IF THE LABORATORY DID NOT PERFORM A PROPER VALIDATION WITH THE CARTRIDGES, IT IS POSSIBLE THAT A FAULTY CARTRIDGE WOULD GO UNDETECTED. ON JULY 9, 1999, A REPORT WAS RECEIVED INDICATING THAT ERRONEOUS RESULTS DURING A VALIDATION HAD BEEN TRACED TO A SINGLE VC-1 CARTRIDGE. THIS CARTRIDGE, AS WELL AS NINE OTHER CARTRIDGES, WERE REC'D BY BIOMETRIC IMAGING ON JULY 30, 1999. THE RETURNED CARTRIDGES WERE INSPECTED AND ONE WAS FOUND TO BE OUT OF SPECIFICATION FOR DEPTH OF THE CAPILLARY WELL AT THE DISTAL END. THIS WAS THE SAME ONE IDENTIFIED IN THE REPORT. TESTING WITH CONTROL PRODUCT CONFIRMED THAT THE CARTRIDGE GAVE ERRONEOUS CELL COUNTS WHEN COMPARED TO IN-HOUSE KNOWN GOOD CARTRIDGES. THE RESULTS OBTAINED WITH THIS CARTRIDGE WAS 20 PERCENT BELOW THE REFERENCE, HOWEVER NO ERROR MESSAGE WAS GENERATED BY THE CEQER SOFTWARE. THE INSPECTION AND TESTING WAS COMPLETED ON AUGUST 13, 1999. NO RESULTS WERE USED FROM THIS CARTRIDGE. ON AUGUST 26, 1999, ANOTHER REPORT WAS REC'D, ALSO DURING A VALIDATION. RESULTS WERE NOT USED. ON SEPTEMBER 9, 1999, BMI REVIEWED THE DATA GENERATED DURING THE COMPLAINT INVESTIGATION. BMI DETERMINED THAT THE CAUSE OF THE ERRONEOUS RESULTS WAS DUE TO THE FOCUS DEPTH BEING OUT OF RANGE FOR THE INSTRUMENT BECAUSE OF THE WELL DEPTH OF THE VC-1 CARTRIDGE. IT WAS DETERMINED THAT THE CURRENT SPECIFICATION AND INSPECTION PROCEDURE FOR THE VC-1 CARTRIDGE WELL DEPTH IS NOT ADEQUATE TO CONTROL THE DEPTH. A NEW SPECIFICATION HAS BEEN SET AND VALIDATION FOR THE CARTRIDGES PERFORMED. A FIELD CORRECTION WAS IMPLEMENTED ON SEPTEMBER 28, 1999. BMI PERSONNEL NOTIFIED CUSTOMERS BY PHONE AND FAXED CUSTOMER NOTIFICATION LETTERS. BMI PERSONNEL VISITED US SITES TO TEST VC-1 CARTRIDGES AND EQUIPMENT USING THE CEQER PRP ASSAYS WITH CONTROL CELLS AND SOFTWARE. ANY DEFECTIVE CARTRIDGES WERE REPLACED IMMEDIATELY. TWO SITES THAT HAD AUTOFOCUS ERRORS WERE DETERMINED TO HAVE A FAULTY PLATTER IN THE IMAGN, WHICH WAS REPLACED. AT LEAST TWO SITES HAD AUTOFOCUS ERRORS THAT COULD BE ATTRIBUTED TO OPERATOR ERROR BECAUSE OF INCORRECT INSERTION OF CAPILLARIES IN VC-1 CARTRIDGES. THESE SITES WERE INSTRUCTED IN THE PROPER USE OF CAPILLARIES. ON SEPTEMBER 28, 1999, BMI IDENTIFIED AT LEAST ONE SITE THAT HAD AUTOFOCUSING ERRORS DUE EITHER TO THE VC-1 CARTRIDGES, THE PLATTER IN THE IMAGN, OR OPERATOR ERRORS, THAT WERE NOT DETECTED IN VALIDATION EXPERIMENTS. WITHOUT AN ERROR MESSAGE IN THE SOFTWARE, CO DEEMED THIS TO BE SUFFICIENT INFO REGARDING THE POTENTIAL FOR PRODUCTS TO BE MISLABELED AS LEUCOREDUCED TO DETERMINE THAT A MEDWATCH REPORT WAS WARRANTED.