AMPLILINK SOFTWARE V3.3.5 WITH CAP/CTM 48 SYSTEM

Additional Manufacturer Narrative

THE INVESTIGATION INTO THIS ISSUE IS ON-GOING. THE CONCLUSION FROM THE INVESTIGATION WILL BE PROVIDED THROUGH A FOLLOW-UP REPORT. (B)(4).

Additional Manufacturer Narrative

DATE OF REPORT 02/10/2015. DATE RECEIVED BY MANUFACTURER 02/10/2015. FOLLOW UP REPORT 1. ADDITIONAL INFORMATION / DEVICE EVALUATION. DEVICE EVALUATED BY MANUFACTURER YES. (B)(4). A CUSTOMER FILED A COMPLAINT STATING THAT 12 (B)(6), GENERATED WITH THE COBAS AMPLIPREP / COBAS TAQMAN 48 SYSTEM, WERE MIS-ASSIGNED TO WRONG PATIENTS. THE 12 RESULTS WERE IDENTICAL TO 12 RESULTS PREVIOUSLY RUN WITH DIFFERENT SPECIMENS ON THE SAME THERMAL-CYCLER OF THE COBAS TAQMAN 48 (CTM 48) ANALYZER. THE INVESTIGATION DETERMINED THAT THE CUSTOMER'S PHYSICAL NETWORK SETUP WAS INCORRECT AND THE HARDWARE WAS FAULTY. BOTH THE LABORATORY LOCAL AREA NETWORK (LAN) AND THE INSTRUMENT LAN WERE CONNECTED TO THE SAME NETWORK HUB AND THE HUB WAS MOST LIKELY DEFECTIVE, WHICH CAUSED FREQUENT DISCONNECTIONS OF BOTH CTM 48 ANALYZERS THAT WERE CONNECTED. THE DISCONNECTION EVENTS WERE VISIBLE IN THE RUN LOG OF THE ANALYZERS. AFTER CORRECTING THE NETWORK LAYOUT AND INSTALLATION OF A SWITCH TO REPLACE THE FAULTY HUB, NO FURTHER LAN DISCONNECTION EVENTS OCCURRED. THE NETWORK MIS-CONFIGURATION AND FAULTY HARDWARE AT THIS CUSTOMER SITE CAUSED THE INITIAL RAW DATA DELETION ON THE CTM 48 ANALYZER, REQUESTED BY THE AMPLILINK SOFTWARE, TO FAIL FOR THERMAL CYCLER B. WHEN THE NEW RUN WAS PERFORMED ON THERMAL CYCLER B, RAW DATA WAS DELETED BEFORE THE RUN STARTED, BUT THE DELETION WAS SLIGHTLY DELAYED. A SOFTWARE TIMING ISSUE ALLOWED FOR AN EXTREMELY LIMITED TIMEFRAME WHERE OLD RAW DATA THAT STILL EXISTED ON THERMAL CYCLER B WERE ABLE TO BE USED BY AMPLILINK. NEW RAW DATA GENERATED FOR THE CURRENT RUN ON THERMAL CYCLER B WAS NOT CONSIDERED BY AMPLILINK SINCE RAW DATA FOR ALL CYCLES HAD ALREADY BEEN RECEIVED. AMPLILINK CALCULATED RESULTS FOR THE RUN ON THERMAL CYCLER B USING INCORRECT RAW DATA FROM THE PREVIOUS RUN, WHICH WERE INCORRECT. AFTER THE NETWORK SETUP WAS CORRECTED AND THE POTENTIALLY FAULTY HUB REPLACED, FAILURE OF THE INITIAL RAW DATA DELETION EVENT DID NOT RECUR. THE ISSUE WOULD BE EXTREMELY UNLIKELY TO HAVE OCCURRED OR TO RECUR IF NOT FOR THE IMPROPER LAN CONFIGURATION AND THE FAULTY HUB HARDWARE. THE ISSUE WOULD REQUIRE AN EXTREMELY UNLIKELY PREREQUISITE SEQUENCE OF EVENTS AND TIMING TO OCCUR. THIS WAS THE FIRST OCCURRENCE OF THIS ISSUE REPORTED IN OVER TWELVE YEARS SINCE THE RELEASE OF THE AMPLILINK SOFTWARE. THERE WAS NO HARM REPORTED FOR THIS COMPLAINT. (B)(4)

Description of Event or Problem

A CUSTOMER SITE IN (B)(6) FILED A COMPLAINT ALLEGING THAT WRONG RESULTS WERE ASSIGNED TO PATIENT SAMPLES WHEN USING THE COBAS AMPLIPREP / COBAS TAQMAN 48 (CAP/CTM) SYSTEM WITH AMPLILINK SOFTWARE VERSION 3.3.5, RUNNING THE CAP/CTM (B)(4) TEST, V2 . RESULTS WERE ASSIGNED TO 12 SPECIMENS ON A SAMPLE RACK THAT ARE IDENTICAL TO THE FIRST 12 RESULTS ON AN UNRELATED RACK WHICH WAS PREVIOUSLY RUN WITH DIFFERENT SPECIMENS. INCORRECT BUT BELIEVABLE RESULTS WERE GENERATED FOR 11 SAMPLES. THE CUSTOMER STATED THAT (B)(6) TREATMENT FOR ONE PATIENT WAS NOT STARTED DUE TO THE ORIGINAL (B)(6) RESULT REPORTED ON (B)(6) 2014 OF 23 IU/ML. WHEN REPEATED ON (B)(6) 2014, THE (B)(6) RESULT WAS 580,000 IU/ML.