Abstract
In the specimen of free PSA in the low concentration, the result in % bias from our institution and comparable evaluation institution was -33.7% which is exceeded % bias ${\pm}20%$ ; however, it was the domestically allowable limit recommended by the laboratory accreditation commission for specimen at the low concentration. In this paper, the cause was accredited by instability of free PSA substance within the specimen, and the specimen stability test was performed according to CLSI documents GP29-A2. After the low and high concentration specimen were made, and rapidly cooled down in a deep freezer with $-30^{\circ}C$, serum of two concentrations was measured for 10 consecutive days with 3 times a day by Architect i2000 and observed a change in the mean value. As the results of two groups, there were changes in the established target value, and a change level was evaluated by calculating it with % bias. The low concentration specimen had no significant reduction until the 4 day lapse in cold storage. However, % bias were reduced by -17.5% from the 5 day lapse, by 21.5% after the 7 day lapse, and by -26.9% after the 9 day lapse. The frozen specimen had only intra-day variation for 10 days. In the high concentration specimen, bias began to show as -12.2% from the 3 day lapse in cold storage. There was reduction by -28.9% from the 5 day lapse, by -39% after the 7 day lapse, and by -42.9% after the 9 day lapse. In the frozen specimen, there was only intra-day variation like the low concentration specimen in cold storage.