Терапевтический архив (Apr 2016)
Multivariate analysis of genealogical markers in adults with primary immunodeficiencies
Abstract
Aim. To study the genealogy of adult patients with primary immunodeficiencies (PID) and to analyze its data in the relatives of these patients. Subjects and methods. A genealogical analysis was carried out in 74 adult patients with PID and 200 adults without this condition, by examining groups of signs in the relatives in at least 4 generations, the genealogical markers were an atypical infectious and inflammatory process; allergic diseases; autoimmune diseases; the presence of relatives with cancers; cases of reproductive dysfunction; deaths from infectious diseases and/or cancers in children; and congenital malformations. The percentage of relatives with the above genealogical markers of the total number of the relatives in 4 generations was used as an indicator. The analysis applied nonparametric methods, such as quartile analysis, Spearman’s correlation coefficient, and Mann-Whitney test to verify the statistical significance of differences between the independent groups. Multifactor prediction models were based on the decision theory (Wald-Bayesian analysis) and classification trees. At Stage 1, the investigators made a univariate analysis, the data of which were used to perform a correlation analysis of the indicators. Multifactor prediction models were based on the decision theory (Wald—Bayesian analysis) and classification trees. Results. The genealogical markers were identified and analyzed using different statistical methods and conclusions were made on prognostically significant indicators. Conclusion. The findings may be recommended for practical use in order to enhance the efficiency of work with patients having various immunopathological syndromes.
