Page | 27 VIEW POINT Use of statistical agreement to run logistic regression

Objective: To run logistic regression after merging two variables of raters depending on percent of agreement. Methodology: Percent of agreement identifies quantitative expression of observer variation between two raters. Correct identification of abnormal and normal subjects depends on terms of sensitivity and specificity. Merging of two variables to have one variable is a crucial step to identify the only diseased/abnormal subjects by two raters. Thereby, logistic regression takes place to evaluate whether risk factors are associated with a defined condition. Conclusion: Using of statistical agreement to run logistic regression can help researchers to assess strong risk factors for one disease diagnosed and agreement by two raters for the same subjects by one step