To save content items to your account,
please confirm that you agree to abide by our usage policies.
If this is the first time you use this feature, you will be asked to authorise Cambridge Core to connect with your account.
Find out more about saving content to .
To save content items to your Kindle, first ensure firstname.lastname@example.org
is added to your Approved Personal Document E-mail List under your Personal Document Settings
on the Manage Your Content and Devices page of your Amazon account. Then enter the ‘name’ part
of your Kindle email address below.
Find out more about saving to your Kindle.
Note you can select to save to either the @free.kindle.com or @kindle.com variations.
‘@free.kindle.com’ emails are free but can only be saved to your device when it is connected to wi-fi.
‘@kindle.com’ emails can be delivered even when you are not connected to wi-fi, but note that service fees apply.
The random]ized clinical trial is widely accepted as the optimal approach to evaluating the safety and efficacy of medical treatments. Resistance to randomized treatment assignment arises regularly, most commonly in situations where the disease is life-threatening and treatments are either unavailable or unsatisfactory. Historical control designs, in which all participants receive the experimental treatment with results compared to a prior cohort, are advocated by some as more ethical in such circumstances; however, such studies are often highly biased in favor of the new treatment and frequently yield misleading results. Alternative controlled designs motivated by the desire to maximize the number of patients with the treatment ultimately determined to be superior have been proposed, but have been challenged on both methodological and ethical grounds. Debates about appropriate and ethical study designs recurred during the recent Ebola Virus Disease (EVD) epidemic in West Africa. Despite its devastating nature, the EVD epidemic showed the ongoing necessity of conducting randomized trials to obtain convincing evidence of the safety and efficacy of therapeutic interventions.
The goal of a controlled clinical trial is to compare the effects of interventions on outcomes of interest. This chapter considers the methods to limit bias and random error at each stage of a clinical trial-design, conduct, analysis and interpretation of results. Many aspects of study design relate to control of bias. The one of greatest importance is the method of assignment to treatment. Study assessments that incorporate some element of subjectivity can also be centralized. Many trials rely on a central adjudication group to make outcome assessments for all subjects in a study. In most studies, the treatments are compared with regard to multiple outcomes. From sample size considerations to central pathology review, from eligibility reviews to interim monitoring plans, all methodological considerations relate in one way or other to minimizing the potential for bias and reducing random error.
This chapter mentions multi-arm randomized trials in multiple hypothesis testing and non-inferiority trials with more than two treatments. Randomization is as central to the conduct of noninferiority as it is to superiority trials. Well-defined endpoints, clearly prioritized and stated in advance, are important components of powerful trials yielding useful results. As with most aspects of the analysis of a randomized clinical trial, the standard to which the worst performance of the treatment vs. the control is to be compared should be specified in advance. Non-inferiority trials can make useful scientific contributions when ethical considerations disallow a placebo or other inactive control. However, unlike the scenario of a superiority trial with a placebo, their assay sensitivity is not directly ensured by randomized comparison and so there are numerous cautions in their use. The measures of study quality mentioned in key comparisons between superiority and noninferiority trials are examined.