中文摘要：

目的 探讨周围神经临床修复效果的科学评价的指标.方法以主题词方式检索Pubmed数据库（1990-2010）的英文文献,以关于周围神经修复材料临床应用的文献为目标文献,类型包括综述和临床研究,排除动物实验、重复研究及研究内容与目的 不符的文献,同时检索在美国临床试验信息网站注册的周围神经缺损修复材料临床研究的文献.结果 计算机初检得到1 578篇,通过对资料进行分析,排除不符合纳入标准的文献,最终31篇全文阅读分析.在美国临床试验信息网站检索到周围神经修复材料临床研究注册项目2个.通过对材料进行提取分析可知：①通过设置入选条件和排除条件可使遴选病例达到最大程度的齐同.②指固有神经疗效观察指标临床最为常用的是MackinnonDellon感觉评定标准,混合神经功能评定临床常用的是BMRC法,肌电图是次要评价方法.③目前的科学技术水平还达不到对植入物安全性的终生评价,仅能通过临床实践长期观察.④随机对照临床试验如采用随机化分组,均衡性好,可比性强,但非随机对照临床试验同样具有重要价值.结论 临床试验设计在限制的时空条件下不可能做到所有因素齐同,但可通过设置入选条件和排除条件使遴选病例达到最大程度的齐同.神经修复效果可采用合理、国际公认的神经功能评定方法、严格的符合要求的随访和统计方案来进行科学评价.

英文摘要：

Objective To discuss how to assess scientifically on the outcome of clinical application with peripheral nerve graft materials. Methods All Pubmed database from 1990 to 2010 were retrieved,and searched the English literatures about the application with peripheral nerve graft materials. The literatures consisted of original clinical research and review excluding animal experiments, repetitive research and irrelevant literatures. The clinical trials data of U.S. was also our target. The information about the safety and effectiveness of peripheral nerve graft materials and related statistical problems were discussed. Results Totally 1578 literatures were identified. Following reading titles and abstracts, we excluded some irrelevant articles. Finally 31 literatures and 2 issue of clinical research from clinical trial data of U.S. were included. After analysis on the literatures, we gained the following results： a remarkable degree of homogeneity among patients can be formed by setting the inclusion and exclusion criteria. For the assessment of proper digital nerve repair, Macknnon- Dellon evaluation is commonly applied, but for the composite nerve, BMRC evaluation is the main method and electromyography can be used as a secondary choice. The safety of peripheral nerve graft materials cannot be evaluated throughout one＇s life according to the current level of science and technology. It should be evaluated by long-term clinical observation. Randomized clinical trials with random grouping was a gold standard for clinical trials with a good balance and strong comparability. However, non-randomized controlled trials also have an important value. Conclusion It is impossible to make all affected factors homogeneity in a limited timespace conditions of clinical trial. However, we can try our best to keep factors homogeneity to maximum degree by setting the inclusion and exclusion criteria. The scientific assessment of outcome of peripheral nerve repair can be carried out with reasonable and internationally recognized ne