B G Blijenberg1, H Lilja, H Neels, U H Stenman. 1. Department of Clinical Chemistry, Erasmus Medical Center, Rotterdam, the Netherlands. blijenberg@erasmusmc.nl
Abstract
OBJECTIVE: To assess the application of a quality control scheme for total prostate-specific antigen (PSA) as used for participants of the European Randomized Study for Screening of Prostate Cancer (ERSPC) during 1996-2002. METHODS: From 1996, the first complete year being 1997, an external scheme was organized by the Dutch Quality Assessment Foundation especially for the ERSPC. This scheme consists of one control round every 2 months with two different human serum samples and is only meant to compare the recovery of methods. From 1998 an internal scheme was also applied by adding two distinct samples to every round. RESULTS: Initially there was a wide variation (coefficient of variation of +/- 15% at threshold PSA of 4.0 ng/mL) among all ERSPC participants who were all using the Tandem assay (Hybritech Inc, USA). After introducing the internal scheme the performance of some intra-ERSPC group comparisons for PSA and the introduction of the completely automated Beckman-Access analyser in 2001 there was state-of-the-art precision for PSA of +/- 5% in the 2002 surveys. CONCLUSIONS: The ERSPC group measurements of PSA have considerably improved since 1996 because of the application of a quality-assessment scheme and with the introduction of complete automation of the PSA assay. Both findings are in line with earlier developments in clinical chemistry.
OBJECTIVE: To assess the application of a quality control scheme for total prostate-specific antigen (PSA) as used for participants of the European Randomized Study for Screening of Prostate Cancer (ERSPC) during 1996-2002. METHODS: From 1996, the first complete year being 1997, an external scheme was organized by the Dutch Quality Assessment Foundation especially for the ERSPC. This scheme consists of one control round every 2 months with two different human serum samples and is only meant to compare the recovery of methods. From 1998 an internal scheme was also applied by adding two distinct samples to every round. RESULTS: Initially there was a wide variation (coefficient of variation of +/- 15% at threshold PSA of 4.0 ng/mL) among all ERSPC participants who were all using the Tandem assay (Hybritech Inc, USA). After introducing the internal scheme the performance of some intra-ERSPC group comparisons for PSA and the introduction of the completely automated Beckman-Access analyser in 2001 there was state-of-the-art precision for PSA of +/- 5% in the 2002 surveys. CONCLUSIONS: The ERSPC group measurements of PSA have considerably improved since 1996 because of the application of a quality-assessment scheme and with the introduction of complete automation of the PSA assay. Both findings are in line with earlier developments in clinical chemistry.