Sample Size Determination for Binary Data in two Independent Groups of Clinical Studies

Abstract:
Backgound and
Objectives
Nowadats, it is very common to investigate the effect of a new method or medicine by using a comparison between two independent groups in many clinical studies, but unfortunately the sample size part has not been done based on scientific formulas but on practical issues in some researches of this type of cinical studies, and this makes some unreliable results. Therefore, this paper studies how to obtain the sample size for binary data when the research aim is to compare two independent groups, and some tables are prepared to reach this aim.
Methods
In this applied study, how to obtain the sample size for binary data to compare two proportions is studied. Chi-square test and exact Fisher test as well as chi-square test with yates correction are used in this paper, and some formulas and tables are presented to the aim of calculating the sample size with the significant level of 0.05 and power of 80% and 90%. SAS 9.2 is used to obtain the sample size in different scenarios.
Findings: Some applied and medical examples are presented to get familiar with the right way of obtaining the sample size, and it is argued how to use the formulas and tables to calculate the minimum size of an appropriate sample.
Conclusion
it is strongly suggested to use the formulas and tables presented at this paper to obtain the appropriate sample size in such researches that their study plan is based on the comparison of two independent groups with binary data.
Language:
Persian
Published:
Razi Journal of Medical Sciences, Volume:23 Issue: 4, 2016
Pages:
52 to 61
https://www.magiran.com/p1574496