Evaluation of five enzymic kits for determination of triglyceride concentrations in plasma

We evaluated five commercial enzymic kit methods (methods A-E) for the determination of triglyceride concentrations in plasma, by comparing each with the automated fluorometric procedure of the Lipid Research Clinics Program. Aliquots of 9 to 97 samples were analyzed with each kit and with the comparison method. The results obtained with all five were highly correlated with those of the comparison method (r = 0.97-0.99). Group mean triglyceride concentrations as determined with kits A and D did not differ significantly from those assayed with the comparison method. The other three kits gave values 6.5-7.5% higher than those by the comparison method (p <0.001). The relative biases between the kit and reference methods did not vary with triglyceride concentration for kits B, D, and E; kits A and C had concentration-dependent biases. Kits D and E were the most nearly precise and had coefficients of variation (CVs) of 3.1% or less. Kit B had CVs of 3.8-4.4%; the remaining two kits had CVs of 5.4-6.7%. On the basis of these results, we judged kits B, D, and E to perform most satisfactorily, from the standpoint of their precision and their concentration in dependent relationships to the comparison method.