[Decision analysis of congenital toxoplasmosis in the absence of exact knowledge of the treatment benefits and secondary effects].

Abstract

Decision analysis seemed the appropriate method to bring out the interest of toxoplasmosis serology in children born from mothers who seroconverted during pregnancy. In particular it provides the possibility to choose between IgA serology, recently introduced in hospital practice, and IgM serology that is the reference test. We study a series of 96 children suspected of having congenital toxoplasmosis who were followed up until the age of one. A decision analysis is thereafter conducted by synthesizing data about clinical efficacy with the RBNCN ratio that is the net benefit of treating an affected person divided by the net cost of treating an unaffected person; these costs and benefits are clinical. The effectiveness of the treatment (sulphonamides in most cases) was not quantified. Therefore, we used the fact that the RBNCN ratio is equal to the number of healthy children likely to be treated unnecessarily in order not to leave untreated a contaminated child. Although data available in the literature and data from our study still remain insufficient to draw a final conclusion, IgA antibodies are significantly more sensitive than IgM antibodies (p < 0.01). Conversely, IgA antibodies are significantly less specific than IgM antibodies (p < 0.001). The sensitivity analysis shows that the IgA test is preferred to the IgM test if the mother is contaminated during the third trimester. On the other hand, for low prevalences, as observed in the first trimester, the IgM serology is more useful.

Cite this paper

@article{cochard1996DecisionAO, title={[Decision analysis of congenital toxoplasmosis in the absence of exact knowledge of the treatment benefits and secondary effects].}, author={Ren{\'e} {\'E}cochard and Gilles Landrivon and Cyrille Colin and P V Cr{\'e}ac'h and Martine Wallon and F. Gandilhon and François Peyron and Yves Matillon}, journal={Revue d'épidémiologie et de santé publique}, year={1996}, volume={44 2}, pages={145-54} }