Unnebrink K, Pritsch M
Abteilung Medizinische Biometrie, Universität Heidelberg.
Med Klin (Munich). 2000 May;95(1 Spec No):47-53.
Treatment evaluation is one of the most important tasks in medical research. Detailed standards have been developed during the last decades. The efficacy/effectiveness of treatments can only be assessed in comparison to control groups. To guarantee the internal validity of these comparisons, the groups have to be comparable at the beginning of the study. This can be achieved by randomized allocation of patients to treatment. Furthermore, as far as possible patient and physician should be blinded to treatment in order to avoid subjective influences on treatment results. Groups should still be comparable when analysing the trial, thus an analysis according to the principle of intention-to-treat ("as randomized") should be performed. These indispensable principles for design, conduct and analysis of clinical trials are widely accepted and contribute to reliable and credible results.
治疗评估是医学研究中最重要的任务之一。在过去几十年里已经制定了详细的标准。治疗的疗效/有效性只能通过与对照组比较来评估。为了保证这些比较的内部有效性，研究开始时各组必须具有可比性。这可以通过将患者随机分配到治疗组来实现。此外，患者和医生应尽可能对治疗不知情，以避免对治疗结果产生主观影响。在分析试验时，各组仍应具有可比性，因此应按照意向性分析原则（“按随机分组情况”）进行分析。这些临床试验设计、实施和分析中不可或缺的原则已被广泛接受，并有助于获得可靠和可信的结果。
