Stratified analysis of multivariate clinical data: application of a Mantel-Haenszel approach.

Laboratory determinations on children aged 6 to 10 years obtained over a 5-year period are analysed by a method described in detail for differentiating between children from exposed and control areas of Seveso, Italy. In the analysis, stratification is employed to distinguish the separate days of laboratory measurement. The analyses permit the study of differences based on all variables simultaneously as well as on each variable separately. Results are obtained for each individual year and for longer intervals of 2 successive years, 3 successive years, 4 successive years, and all 5 years. The results and some statistical aspects of the analysis are discussed.