Evaluation of within-day precision of serum cholesterol measured by a portable analyzer

The use of dry-chemistry analyzers for the measurement of cholesterol in mass screening settings has received considerable attention. Less is known about the efficacy of these types of analyzers in the clinical office or laboratory setting. Thus, we investigated the within-day precision of cholesterol measurements performed by the Reflotron method with specific comparisons made between two trained technicians and two instruments. Forty serum specimens were analyzed eight times (twice by each technician on both instruments) with the identity of the specimen blinded from the technician. Each specimen was also analyzed by two different methods commonly accepted in the clinical laboratory to estimate the accuracy of the Reflotron cholesterol measures. Small, significant (P less than 0.05) main effect differences were observed in cholesterol concentrations between technicians (1.8%) and instruments (0.8%). The overall coefficient of variation (CV) of the serum cholesterol measures was 2.5%, which meets the Laboratory Standardization Panel's "ideal" goal. However, three individual cases (one specimen's eight analyses) had CV greater than 5%, and three other cases had CV greater than 3%. Most of these cases could be traced to one outlier value. Review of all 320 Reflotron analyses revealed that only 10 (3.1%) were outliers (greater than 5% from specimen mean value). When operated in a laboratory with regular quality control procedures, the Reflotron method can meet national standards for precision. In this setting, differences between technicians and instruments are not of clinical importance.