Statistical issues in tumor marker studies

Context.- Inferences from tumor marker studies are complicated by a variety of statistical concerns, which can make proper interpretation of results d ifficult. This article focuses on important issues that should be addressed when designing, conducting, and analyzing tumor marker studies. Objective.-To highlight the importance of considering statistical significa nce, risk ratios, statistical power, reproducibility, multiple testing, con firmatory studies, and missing data in the design of marker studies used fo r prognosis. Results.-Suggestions are provided for more effectively conducting marker st udies. These include more careful attention to adequacy of the number of su bjects for a marker study and improved documentation and standardization of assay methods. The importance of complete reporting of study results and d escription of the statistical analysis methods used is also emphasized. Conclusion.-Cooperation among clinicians, laboratory scientists, and statis ticians will be required to conduct statistically sound tumor marker studie s and to facilitate prioritizing markers for new and confirmatory studies i n an environment of limited patient and specimen resources.