Multivariate analysis and genetic-morphological method for predicting ventral hernias

Abstract

Background. Development of prognostic models of disease development allows identification of variables influencing the risk prediction of disease occurrence or recurrence and use of these multiple risk factors in a systematic, reproducible manner in accordance with evidence-based methods. As an example of one of the problems discussed among surgeons is the reduction of the number of recurrences of anterior abdominal wall hernias, as well as prevention of this disease, which can be achieved by early diagnosis of predisposition using methods of analysis of etiopathogenetic features of this multifactorial disease.

Objective: to analyze risk factors and develop a prognostic model of ventral hernia.

Methods. The study was performed on 579 patients. The study included: correlation analysis, ROC analysis and analysis of the importance of etiologic risk factors using an artificial neural network. Morphological factors were studied in skin biopsy and using skin autofluorescence. PCR analysis of rs3809060 WT1 and rs2009262 EFEMP1 was performed.

Results. Based on the results of the study, 1 genetic (rs2009262 EFEMP1) and two morphological factors of connective tissue changes were included in the prognostic model. Subsequently, using logistic regression, a prognosis was made for the occurrence of a hernia, as well as its recurrence, after which patients underwent various types of hernioplasty.

Conclusion. Thus, this study once again proved the major role of genetic and morphological factors associated with changes in connective tissue, determining the surgeon’s tactics.