Unformatted text preview: e control or to at least show difference. (To many patients and clinicians, this is the information of greatest relevance, namely, the comparative effectiveness of a new drug to a standard therapy.) But showing difference or superiority is not enough because ‘‘many kinds of study defects decrease the likelihood of showing a difference between treatments’’ (Temple 2002:222) and make data on difference less reliable. Study defects arise from external factors like poor patient compliance, poor diagnostic criteria, and the use of concomitant medication that can obscure effect. Other defects can include inconsistencies in the application of the definition of disease, the use of insensitive or inappropriate measures of drug effectiveness, and the chance of spontaneous recovery in a study population. These factors can be ‘‘fatal to a trial designed to show a difference’’ (Temple 2002:222; also see Pocock 2002:244 – 245). They can decrease difference or increase the chances of finding no difference, such that, in the end, in Templ...
View Full Document
This note was uploaded on 10/10/2010 for the course ENG 000121 taught by Professor Mcgrand during the Spring '10 term at Cornell.
- Spring '10