RESUMO
Screening for blood cholesterol levels has become an essential need in many countries. The aim of this study was to evaluate the performance of Reflotron analyzer in screening for blood cholesterol in terms of its precision and accuracy. We have compared the results obtained by the Reflotron with those obtained by a standard hospital laboratory procedure for the same samples. An optimal precision that met the 1992 LSP standards for precision was observed (CV = 2.27%). On the other hand, a suboptimal accuracy of Reflotron was detected, where more than 80% of the Reflotron values differed from the reference laboratory values by more than 5%. As a consequence, 21% of individuals were misclassified as normal due to the very low sensitivity in spite of the optimum specificity observed for the Reflotron. Using the receiver operating characteristic curve analysis, the optimum cut-off points giving the best validity values for the test were 170 mg/dl and 190 mg/dl (4.4 mmol/L and 4.9 mmol/L) for the border-line and true hypercholesterolemia, respectively. At these levels, sensitivity was 0.89 and 0.86, specificity 0.93 and 0.93, and positive predictive value was 0.91 and 0.82, respectively. Considering some limitations of the study, it is concluded that the Reflotron could be of more potential validity for the use in cholesterol measurement surveys when the recommended cut-off points are used.
