ACQUITY

Description of Event or Problem

CUSTOMER REPORTS THAT THE WATERS 3777C SAMPLE MANAGER INJECTED A SAMPLE FROM THE INCORRECT VIAL LOCATION. THIS OCCURRED WHEN THE CUSTOMER PAUSED AN ACTIVE BATCH OF PATIENT SAMPLES IN ORDER TO ADD SOME PRIORITY SAMPLES, WHICH WERE ADDITIONAL CONTROLS MISSING FROM THE ORIGINAL BATCH. AT THE TIME THE LIST WAS PAUSED THE NEXT SAMPLE IN THE BATCH HAD ALREADY BEEN DRAWN INTO THE INJECTOR USING THE SINGLE SAMPLE OVERLAP FUNCTION. WHEN RESTARTING THE RUN, THIS PRE-DRAWN SAMPLE WAS INJECTED BUT RECORDED IN THE DATA FILE FOR THE FIRST OF THE PRIORITY SAMPLES. AFTER THE PRIORITY SAMPLES WERE RUN THE SAMPLE LIST RESUMED AND THE PRE-DRAWN SAMPLE WAS INJECTED AGAIN. IN THIS CASE, THE INCORRECTLY INJECTED SAMPLE WAS A CONTROL AND NOT A PATIENT SAMPLE. THERE IS NO ALLEGED MISREPORTING OF PATIENT DATA.

Additional Manufacturer Narrative

WATERS TESTING HAS REPLICATED THE REPORTED CONDITION BY THE CUSTOMER AND HAS INITIATED ACTION TO CORRECT. WATERS CORRECTIVE AND PREVENTATIVE ACTION PLAN UNDER DEVELOPMENT.

Additional Manufacturer Narrative

INVESTIGATION AND ROOT CAUSE: THE RENATADX SYSTEM (WHICH IS A XEVO TQD WITH A PAL 3777C AND MASSLYNX SCN1005IVD) USING THE SINGLE SAMPLE OVERLAP FEATURE, THE USER THEN PAUSES THE RUN LIST AFTER IT HAS STARTED ACQUIRING AND ADDS MORE SAMPLES TAGGING THEM AS PRIORITY SAMPLES (USING THE PRIORITY SAMPLES FEATURE IN THE SOFTWARE), THEN RESTARTS THE RUN LIST. THE 3777C HAS ALREADY EXTRACTED THE NEXT SAMPLE FROM THE PAUSED SAMPLE RUN LIST AND DOES NOT FLUSH IT OUT OF THE SYRINGE, SO THE SOFTWARE THINKS THAT THIS SAMPLE IS THE FIRST PRIORITY SAMPLE AND STORES THE DATA AS PRIORITY SAMPLE 1 DATA. HOWEVER, THE SAMPLE ANALYSED WAS A SAMPLE FROM THE PRE-PAUSED RUN LIST. THE DATA STORED IS AN INCORRECT RESULT FOR THAT SAMPLE. ALL SAMPLES AFTER THIS ARE ANALYSED AND STORED CORRECTLY. THIS IS A PROBLEM WITH THE SOFTWARE, THE 3777C DOES NOT KNOW WHICH SAMPLE IS HELD IN THE SYRINGE WHEN A RUN LIST HAS BEEN PAUSED AND IT DOES NOT KNOW TO FLUSH THIS OUT AND START A NEW EXTRACTION ON RESTARTING THE RUN. WATERS MASSLYNX SOFTWARE IS TESTED BASED ON THE GENERAL PRINCIPLES OF SOFTWARE VALIDATION; FINAL GUIDANCE FOR INDUSTRY AND FDA STAFF WHICH STATES THAT "SOFTWARE TESTING IS VERY LIMITED IN ITS ABILITY TO SURFACE ALL LATENT DEFECTS IN SOFTWARE CODE. FOR EXAMPLE, THE COMPLEXITY OF MOST SOFTWARE PREVENTS IT FROM BEING EXHAUSTIVELY TESTED". IN THIS CASE THE USE OF BOTH THE SINGLE SAMPLE OVERLAP FEATURE AND THE PRIORITY SAMPLE FEATURE WAS NOT INCLUDED IN THE TESTING SCHEDULE BECAUSE USING BOTH THESE FEATURES TOGETHER AFTER PAUSING A SAMPLE LIST WAS NOT AN EXPECTED WORKFLOW OF THE WATERS CUSTOMER BASE. IF THESE FEATURES ARE AVAILABLE FOR USE AT THE SAME TIME, ONCE THIS DEFECT HAS BEEN FIXED, THEN THIS WORKFLOW WILL BE A KNOWN ISSUE AND CAN BE ADDED TO THE TEST SCHEDULE. RISK ASSESSMENT: CURRENTLY THE RISK REMAINS LOW DUE TO LOW PROBABILITY OF USE OF PRIORITY SCHEDULING OR DELETING BATCH IN ORDER TO TEST DIFFERENT PATIENT SAMPLES THAT MAY REQUIRE IMMEDIATE ACTION. A CHANGE TO MASSLYNX THAT WOULD ALLOW USE OF PRIORITY SCHEDULING, PAUSING, DELETING BATCHES AND SINGLE SAMPLE OVERLAPPING RESULTS AND WOULD RESULT IN THE INJECTION OF THE CORRECT SAMPLES, WOULD ELIMINATE THE RISKS COMPLETELY. CORRECTIVE ACTION: THIS ISSUE REQUIRED THE FIELD ACTION. IN THIS CASE, THE FIELD ACTION WOULD TAKE THE FORM OF A CUSTOMER NOTIFICATION OUTLINING THE STEPS THAT SHOULD BE TAKEN TO PREVENT THIS EVENT AND REITERATING THE RELEVANT PORTIONS OF THE OVERVIEW AND MAINTENANCE GUIDE. A NEW MASSLYNX UPDATE SCN1035IVD IS BEING RELEASED TO RESOLVE THE SPECIFIC 3777C ISSUE.