11.07.2015 Views

Clinical Trials

Clinical Trials

Clinical Trials

SHOW MORE
SHOW LESS
  • No tags were found...

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

■■❚❙❘Chapter 18Significance Tests andConfidence IntervalsDuolao Wang, Tim Clayton, and Hong YanThe primary objective of a clinical trial is to provide a reliableestimate of the true treatment effect regarding the efficacyand/or safety of an investigational medicine or therapeuticprocedure. Three major factors can influence the observedtreatment difference away from the true treatment effect.These are bias, confounding, and chance/random error.Assuming no bias or confounding exists, statistical analysisdeals with chance; by providing statistical estimation andtesting (inference), it assesses whether random variationcould reasonably explain the differences seen. While statisticalestimates summarize the distribution of a measured outcomevariable in terms of point estimate (eg, mean or proportion)and measure of precision (eg, confidence intervals), statisticaltesting involves an assessment of the probability of obtainingan observed treatment difference or more extreme differencein the outcome variable, assuming there is no difference in thepopulation. In this chapter, we introduce the ideas underlyingthe principles of statistical inference and describe two statisticaltechniques (hypothesis testing and confidence intervals), withemphasis on their interpretation and application.185

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!