An application on meta analysis and meta regression in clinical studies

With this study, the presentation of meta-analysis and meta-regression methods as useful solutions to the researchers is aimed. It has been conducted as a practice of nine studies which are on celiac disease and cancer cases that are observed in different years and countries. The regression and multiple regression methods are used in order to evaluate the relationship between one or more covariates (independent variables) and one dependent variable. In this study, a meta-analysis and post-meta-regression were conducted to evaluate the strength of the differently reported existing evidence for the association of celiac disease with cancer. Examining the variability or heterogeneity in the results of studies is essential for evaluation of the conclusion. Firstly, the present studies on celiac disease and cancer were combined with meta-analysis, and then meta-regression method was performed. As a result, it is marked that the incidence of malignancy is slightly increased in patients with celiac diseases. The existence of regression was also tested with effect size in the meta-regression method. However, no relation has been found as a result of the test. According to the related results; the heterogeneity level of the obtained data in the study on the relationship between celiac disease and cancer was found to be moderate. During the process of meta-analysis, if the number of included studies are inadequate or the confidence intervals of the studies are too wide (%95), it should be taken into consideration that a false negative result can be obtained in terms of heterogeneity. One of the most common mistakes in meta-analysis is the heterogeneity in patient population or in study designs. By using usual and classic regression methods, the important perspectives of the meta-regression method used in meta-analysis to examine the effect of the moderator variables (a third variable that affects the strength of the relationship between dependent and independent variables) on the study effect size were presented in the study. In addition, a practice of the data has been conducted. It should be noted that the low number of studies and explanatory variables in meta-regression can be stated as potential causes of heterogeneity.