Burgess L P, Yim D W, Lepore M L
Laryngoscope. 1984 Nov;94(11 Pt 1):1472-6.
The treatment of Bell's palsy with steroids is controversial. Several series of patients have shown steroids to be beneficial to a statistically significant degree; however, their experimental design was of poor quality and made this data less clinically valuable. The negative trials showing no benefit with steroids did not examine enough patients to have a chance to show a clinically important difference. These series must be considered as partially completed studies and not as conclusive negative trials. A Type I error analysis is the accepted method of statistically analyzing the data of these trials. After reviewing the evidence, it becomes clear that further scientific research is needed.
使用类固醇治疗贝尔氏面瘫存在争议。多个患者系列研究表明，类固醇在统计学上有显著益处；然而，其实验设计质量较差，使得这些数据的临床价值降低。显示类固醇无益处的阴性试验所纳入的患者数量不足，因而没有机会显示出临床上的重要差异。这些系列研究应被视为部分完成的研究，而非确凿的阴性试验。I型错误分析是对这些试验数据进行统计学分析的公认方法。在审视证据之后，显然还需要进一步的科学研究。
