The customer stated that they received an erroneous result for one patient sample tested for nh3l ammonia (nh3) on a c501 analyzer.
The sample initially resulted as 461.
0 umol/l and this value was reported outside of the laboratory to a physician.
The patient was treated based on the initial result, but no details could be provided regarding the treatment that was provided.
A second sample was collected from the patient for monitoring and this sample resulted as 28.
9 umol/l accompanied by a data flag on (b)(6) 2016.
The second sample was diluted manually at a times three dilution, resulting with a raw value of 13.
5 on (b)(6) 2016.
When accounting for the dilution factor, the diluted sample resulted with a final value of 40.
5 umol/l.
The physician questioned the value obtained with the original sample since the result from the second sample was normal.
The original sample was repeated on (b)(6) 2016, resulting as 30.
9 umol/l accompanied by a data flag.
The original sample was repeated a second time, resulting as 50.
8 umol/l.
A third sample was also collected from the patient and tested on (b)(6) 2016, resulting as 34.
8 umol/l accompanied by a data flag.
The third sample was diluted manually at a times three dilution, resulting with a raw value of 12.
0 on (b)(6) 2016.
When accounting for the dilution factor, the diluted sample resulted with a final value of 36.
0 umol/l.
The repeat analysis of the original sample and results from the second and third samples were believed to be correct.
It was asked, but it is not known if the patient was adversely affected due to receiving treatment.
No adverse events were alleged to have occurred with this patient.
The nh3 reagent lot number and expiration date were asked for, but not provided.
The customer stated that one level of control was outside of range prior to the event.
The field service representative could not find a cause for the event.
He checked the operation of the instrument.
The customer ran precision studies, calibration, and quality controls; all tests passed.
During investigations, it was stated that the issue is likely related to pre-analytic sample handling since the affected sample was said to be only 1/4 full.
|