Reynolds Penny S, Garvan Cynthia S
Department of Anesthesiology, College of Medicine, University of Florida, 1600 SW Archer Road, Gainesville, FL 32610.
Mil Med. 2020 Jan 7;185(Suppl 1):88-95. doi: 10.1093/milmed/usz249.
Hemorrhage control is the top priority in far-forward care. Preclinical studies are essential for determining safety and efficacy before novel therapeutics can be tested in humans. Unfortunately, poor methodological quality jeopardizes translational potential.
We systematically reviewed 136 recent publications describing swine models of hemostasis and hemorrhage reduction to assess compliance with established standards for scientific reporting. Quality measures were summarized by descriptive statistics; randomization was assessed by using baseline group differences to test the uniform distribution assumption for observed P-values.
Most articles did not report information essential to assess study validity and reliability of experimental results. Studies claiming random allocation showed clear evidence of systematic bias. Sample sizes were small, but nearly all studies reported statistically significant effects in the direction of "benefit." Excessive hypothesis testing increased the risk of false positives.
Methodological quality was poor. Although funding agencies actively promote good scientific practice, investigators have been slow to comply. Poorly executed and reported animal research is an ethical and translational issue, wasting animals and potentially harming patients. To properly assess the therapeutic benefit of novel interventions, investigators must rely less on rote hypothesis testing, develop skills in experimental design and quantitative analysis, and comply with best-practice reporting guidelines.
出血控制是前沿医疗的首要任务。在新型疗法能够在人体进行测试之前，临床前研究对于确定其安全性和有效性至关重要。不幸的是，方法学质量欠佳会危及转化潜力。
我们系统回顾了136篇近期描述猪止血和减少出血模型的出版物，以评估其是否符合既定的科学报告标准。质量指标通过描述性统计进行总结；通过使用基线组差异来检验观察到的P值的均匀分布假设，对随机化进行评估。
大多数文章未报告评估研究有效性和实验结果可靠性所需的关键信息。声称进行随机分配的研究显示出明显的系统偏差证据。样本量较小，但几乎所有研究都报告了在“有益”方向上具有统计学显著意义的效果。过度的假设检验增加了假阳性的风险。
方法学质量较差。尽管资助机构积极推广良好的科学实践，但研究人员在遵守方面进展缓慢。执行和报告不佳的动物研究是一个伦理和转化问题，既浪费了动物，又可能伤害患者。为了正确评估新型干预措施的治疗益处，研究人员必须减少对机械假设检验的依赖，培养实验设计和定量分析技能，并遵守最佳实践报告指南。
