临床试验方法学：临床试验概述.ppt

* * * * * * * 两组的可比性、举例：进来之前抽签，一部分中餐，一部分火锅，检查血脂。血脂的平均水平叫做计量资料，是否异常是计数资料 * * * * * * * Suggested Remarks ·??? Open-label: some studies are open label in which both the study team/investigator and the subject know if the subject is receiving the test or control treatment; this design may lead to biased results: –??? the study team may be biased in their recording, interpretation, and analysis of study data –??? the subjects may be biased in their reporting of qualitative data or in their administration of the treatment –??? the study team and/or patients may underestimate safety aspects in the placebo arm or overestimate efficacy results in the test treatment arm ·??? However, many studies use blinding/masking to minimize the potential for management bias –??? note: the term blinding is used for most trials; however, the term masking is used in ophthalmology trials –?? trials may be blinded/masked in different ways according to the aims of the trial, including: ·???? single-blind: the investigator/study team knows which treatment the subject is receiving, but the subject does not; this design may lead to bias in the interpretation and analysis of study data but may be used where it is important for the study team to know which treatment the subject is receiving (eg, to rapidly detect and treat adverse events) ·???? double-blind: neither the investigator/study team nor the subject knows which treatment the subject receives; this ensures that the administration of the treatment, the recording, interpretation, and analysis of study data, and the subject reported data are unbiased ·???? double-dummy: a form of double-blind trial, this trial design is used in cases where it would be unethical not to treat the subject, but where the 2 treatments being compared cannot be made indistinguishable from each other (for example, one is a solution and the other a tablet) –?? the subject receives either the new treatment or the existing treatment, and a placebo; in this exam