Abstract

CARDENAS DELGADO, Víctor Manuel; NUNEZ, Myriam  and  NUNEZ TAQUIA, Guillermo Gabriel. Validation of solid phase immunoassays. Acta bioquím. clín. latinoam. [online]. 2018, vol.52, n.3, pp.339-345. ISSN 0325-2957.

Determination of the levels of specific antibodies by semiquantitative ELISAs requires the construction of a 4 parameter logistic regression (4PL) calibration curve. However, alternative models are often employed to adjust and estimate calibration data. The aims of this work were to determine the model that best adjusts calibration data, and which alternative model generates a lower relative percentage error (RPE) in the prediction of calibrator concentrations. An IgA anti-gliadin ELISA was used. The optical density values (OD) of calibrators were adjusted with the following mathematical models: i) Exponential, ii) Boltzmann, iii) Boltzmann semilog, iv) Deming, v) Linear regression, vi) 4PL, and vii) Quadratic. Results indicated that the best adjustment is given by models v) (R2=0.9914), i) (R2=0.9652) and iii) (R2=0.9650). However, models i) and iii) performed better in the reciprocal adjustment procedure: RPE values were ≤20% for all the calibrator levels analyzed (0-100 IU/mL). The linear models iv) and v) had high RPE values. To sum up, the exponential (i) and Boltzmann (iii) adjustments present R2 and RPE values similar to those of the 4PL, the use of any linear adjustment being inappropriate.

Keywords : Calibration; Semiquantitative ELISA; Immunoassay; Logistic model; Linear model; Validation; Comparison; Reciprocal adjustment; Relative error; Estimation.

        · abstract in Spanish | Portuguese     · text in Spanish     · Spanish ( pdf )