Country reports from Japan: external quality assessment scheme in Japan.

Several external quality assessment schemes (EQAS) have been conducted in Japan. Results obtained from nation-scale EQAS reveal the current quality of laboratory testing in each laboratory. The largest nation-scale EQAS in Japan is that conducted by the Japan Medical Association. The numbers of participants and of items evaluated have increased in EQAS by JMA over its history of 32 years. Improvement in inter-laboratory differences has been observed for most items in EQAS in recent decades. In 1998, about 2,500 laboratories from throughout the country participated in this surveillance, and 47 items were evaluated. The coefficient of variations for the group of all participants was less than 5% for about one third of all test items. On the other hand, very high variations over 20% were observed for 6 items. Also, inter-method differences exist for many items, which may be or may not be related to matrix effects. Retrospective evaluation of all EQAS data suggests that there is still room for improvement in inter-laboratory differences.

External quality control results for hormones, tumor markers and CRP testing.

Most hormones, tumor markers, C-reactive protein, and rheumatoid factor (RF) are measured immunologically. Immunological methods based on the antigen-antibody reaction have certain specific problems, including their principle of determination, character of antibodies used, reaction conditions, and others. Free thyroxine (FT4) and thyroid stimulating hormone (TSH), as well as alpha-fetoprotein (AFP), carcinoembryonic antigen (CEA), prostate antigen, carcinoantigen 19-9 (CA 19-9), and CA 125 are very commonly measured in the routine medical laboratory. Authentic materials can be obtained for hormones and CRP, and efforts to improve quality control and standardization have been made for years. Results of surveillance for FT4, TSH, and AFP were not poor, but inter-laboratory differences for CEA, CA 19-9, and RF were not insignificant.

Quality control, quality assessment of laboratory tests, lipid profile.

The first external quality control survey implemented by the Japan Medical Association was performed in 1967, and now. 2.400 or more laboratories are participating in this survey. The survey for total cholesterol was performed from the first survey, triglyceride survey was started in 1973, and survey of HDL-cholesterol was started in 1981. CVs of results for above items in the surveys were greatly improved after the 1980's. CVs in the 1970's were 8 to 10 per cent. but from the middle of the 1980's they were about 3 per cent or less. The main reason for this improvement is the introduction of automated analyzers in laboratories. An additional reason for this improvement is the requirement for accuracy of cholesterol determination by physicians, because the Japan Society of Atherosclerosis decided the consensus value of cholesterol for the initiation of treatment in 1987. Supply of standard material for cholesterol since 1995 is also a reason for this improvement. In 1995. CVs of cholesterol assays by were less than 3% when using most determination kits. Two determination kits, however, were found to give very high CVs. Both of them used dry chemistry, and the reason for these discrepancies was matrix effects of the survey material.

Quality control, quality assessment of laboratory tests, HbA1c.

HbA1c level is known to be highly correlated with blood sugar level for a while after the maximum blood sugar level was observed, and is now considered more reliable than blood sugar as a long-term index of blood sugar control in diabetic patients. The Diabetes Control and Complication Trial in United States of America reported that mean HbA1c was the dominant predictor of retinopathy progression and that 10% lower HbA1c (e.g., 8 vs. 7.2 %) is associated with a 43 % to 45 % lower risk of progression of retinopathy. This report indicated that only a 1% difference in HbA1c level has a critical effect on progression of diabetic retinopathy. It also indicates the need for high precision and accuracy of hemoglobin A1c determination for control of diabetes patients. There are various methods to measure HbA1c. but now most laboratories use the high-performance liquid chromatography (HPLC) method; immunoassay is the second most popular, followed by affinity chromatography. It has been reported that each of these methods has its own characteristics, and that their results can vary considerably. In addition, there are two types of HbA1c, one a stable component and the other a labile component, and the measured HbA1c values were influenced by the contents of labile component included in the sample. The Japan Diabetes Society took a leading role and organized a committee on inter-laboratory standardization of HbA1c Determination in 1994. The HPLC method, which eliminates labile components, is used in most laboratories. Also, the number of laboratories using the immunoassay has gradually increased, while use of the other methods has rapidly decreased. In conclusion, since the proposal of a reference method and initiation of supply of reference materials by the Japan Diabetes Society, inter-laboratory differences in measured values have become much smaller, mainly due to the methods used for measurements. We need to continue efforts toward further standardization of HbA1c determination.