Abstract: Objective To analyze the method variation and actual allowable error range for tumor markers in Shanghai External Quality Assessment Scheme, and to compare the difference between international standards and biological variation requirements in order to investigate the suitability of allowable total error (TEa) for tumor markers. Methods Based on the data of Shanghai External Quality Assessment Scheme from 2010 to 2012, 30 samples were analyzed to calculate synthetic coefficients of variation (CV)of different methods for tumor markers, for 6 times. The actual TEa of tumor markers in Shanghai External Quality Assessment Scheme was calculated by synthetic CV and compared with the Rilibak quality control guide in Germany and the Royal College of Pathologists in Australia (RCPA) quality and biological variation requirements. Results In 6 times of Shanghai External Quality Assessment Scheme, the actual TEa of alpha-fetoprotein(AFP), carcinoembryonic antigen(CEA), carbohydrate antigen(CA) 125, CA153, CA19-9 and prostate-specific antigen(PSA) were 22.63%, 22.43%, 19.30%, 21.56%, 27.43% and 19.55%, which were closed to the Rilibak TEa requirements and biology variation TEa desirable requirements, but was superior to the TEa of RCPA requirements. Conclusions In Shanghai, the actual TEa of tumor markers in Shanghai External Quality Assessment Scheme is closed to the requirements of German Rilibak and slightly below the specification of the more strict Australian RCPA TEa, and TEa in Shanghai External Quality Assessment Scheme can meet the current quality level.

Key words: External quality assessment, Allowable total error, Tumor marker