|
A customer in (b)(6) notified biomérieux of false positive results associated with bact/alert® 3d instrument.The customer reported that when the bottles flagged positive by the instrument the laboratory performed a subculture using bold macconkey, chocolate agar incubating the plates +/- co2 at 37°c for 48 h.The customer reported that the concerned bottles were unloaded and incubated further in the external incubator together with the subcultures.If there was no growth on the subculture after 48 hours, the laboratory inoculated another sample from the bottle and incubated it again for 48 hours.This step has been repeated until the end of the incubation time defined for the bottles (seven days).Patient results: false positive result.Delay in reporting of results.The bact/alert results have been released to the physician, when the bottles were flagged positive.After having the subculture results, the customer informed the physician about the false positive results.There is no evidence regarding inappropriate patient treatment.An internal biomérieux investigation will be initiated.
|
|
An internal biomérieux investigation was performed with results as follows: backup review for bottle id [nrjgy0c6] bottle movement history nrjgy0c6 - the bottle was loaded in cell 2d28 when the instrument initially determined the positive result.The 3d backup data shows that the customer randomly moved the bottle twice after the bottle was initially reloaded from the positive result.The bottle readings extracted from the 3d instrument shows that the graph's bottle appears to have biphasic growth (downward reflectance on graph) but this occurred after the bottle was determined positive and has no impact on the bottle testing or result.There is no final result listed for the bottle as it was still under test when the backup was created.Table 1: bottle history (b)(6) algorithm results graphs nrjgy0c6 -, the instrument determined the bottle result was positive by (res pos how 2 "acceleration") as shown in the table above.The bottle readings were processed through the bact algorithm tool which determined the same result "acceleration" as the instrument.Fault codes - the bottle was loaded in cell 2d28 and there were no fault code recorded in incubation module 2 during the bottle test time.There were no temperature or power events recorded on the system.There was on fault code 80 recorded on incubation module 1 and this had no impact on the bottle testing or result.Fault code 80 - reading gap detected operator error - there are three operator error codes recorded during the testing time.The operator error code that was recorded on the system are detailed below had no impact on the bottle testing or result.Operator error code 800 - printer fault operator error code 911 - invalid cell unload, these codes in cell 1a49 and 1a50, the bottle was loaded in cell 2d28 operator error code 921 - barcode scan not allowed in this mode.Table 8: error codes (see attached root cause investigation form) bottle loading and unloading events the backup shows the customer did load 10 bottles at once during the load of the bottle id nrjgy0c6.There were no large unload of bottles at once from the instrument.Large loading and unloading events can cause a heat loss within the racks.The loading of 10 bottles did not cause a temperature fluctuation.Sample issue customer indicates: customer stated that the white blood cell count [wbc] is 8860, which is considered normal.The fill volume was not verified by the customer as the bottles were no longer available.Conclusion the backup was reviewed from the site, there were no bottle movements, fault / operator codes, or algorithm issues found with the instrument that caused the false positives to occur.The backup did show that the customer loaded multiple bottles (10 bottles) in the rack which could cause a drop in temperature; however temperature fault codes did not occur, therefore, loading multiple bottles can be ruled out as a root cause.To prevent possible temperature fluctuations, the customer should activate the cold bottle compensation filter (see 514483-8en1 setup and configuration version b.50) to alleviate potential false positives when loading and unloading in bulk.Currently, the backup shows this feature is disabled.The probable root cause is sample related as all known issues with the instrument have been reviewed and not determined to be the root cause of the false positives.Biomérieux recommends to review the best practices in the user manual 514818 appendix c to help prevent false positives with the instrument and sample.Also recommend to follow the ifu (instructions for use) when handling false positives.
|
