Introduction: An equivalence model which allows comparison of blood cholinesterase values, measured by Lovibond (semiquantitative technique), and Michel, EQM, Monotest (erythrocyte and plasma cholinesterases) values measured by quantitative techniques is required.
Objective: The performance of Lovibond (Edson tintometric and Limperos & Ranta techniques) were compared with quantitative techniques.
Materials and methods: The experimental design was descriptive, cross-sectional, and prospective. From a working population (18-59 years) in Valle de Aburrá and Near East of Antioquia. 827 representative samples were chosen for their lack of exposure to cholinesterase-inhibiting plaguicides and affiliated to the Social Security System.
Results: (1) 827 workers were classified by Lovibond in four categories: 821 values with 75% of cholinesterase activity or greater (categories 75, 87.5 and 100%) and 6 with cholinesterase activity smaller than 75%. (2) With each quantitative method, the mean values of erythrocyte and plasmatic cholinesterase corresponding to the four values obtained with Lovibond were statistically different to each other. (3) The mean values of each quantitative technique increased when increased the tintometric method value. (4) Lovibond classified the low enzymatic erythrocyte activity very poorly (61-73%), but the classification of the low enzymatic plasma activity was almost completely in error (94-96%).
Conclusion: The values of erythrocyte or plasma cholinesterase were adequately estimated by both the quantitative techniques of Michel and EQM and by Lovibond, but only when the enzymatic activity is normal. Lovibond, however, had a poor capacity to designate as "low" the values that were low according to the quantitative tests.