I get an odds ratio of 50 millions with binomial logistic regression. I tried with single predictor but the result remains the same. The independent variable is categorical and there is no missing data. Please advise, thank you.