Proposal of automation of candidate reference method for the accurate serum cholesterol assay in clinical laboratories

Background: In 2001, Japan Society of Clinical Chemistry (JSCC) recommended the cholesterol dehydrogenase (CD)-UV method as a comparative method for serum cholesterol measurement in Japan. Although the CD-UV method is intended to standardize routine laboratory tests, it requires complex manipulations, and has been difficult to use for the evaluation of clinical laboratories. We therefore attempted to automate this method using reagents specified by JSCC and developed a simple automated method. Methods: We evaluated the simple automated method using 2 general instruments (JCA-BM12 and H-7170S). The linearity was confirmed for the range over 15.52 mmol/l. The coefficients of variation for 20 measurements of serum containing 2.5 and 6.1 mmol/l of cholesterol were < 1.0%, respectively. No interference by bilirubin, ditauro bilirubin, hemoglobin and chylomicrons was observed in this method. When measurement data with JCA-BM12 were compared with those using the comparative method, the correlation coefficient was 0.999 (n = 23), the regression formula was y = 0.992x - 0.0036 (mmol/l), and the bias was 0.8%. A similar data was obtained with H-7170S. Thus, in both comparisons, the bias was within the target (± 3.0%). Conclusions: This automated method provides a valid means of implementing the serum cholesterol measuring method recommended by JSCC.