It was determined that in a group of 100 births, given women with risk factors, there were 30 premature births, while among the 100 births, given by women without risk factors, there were 5 premature births. What method of statistical data processing is optimal, if a physician needs to assess the validity of such differences in compared groups?