Introduction: Scholarly research on road accidents over the past 50 years has generated substantial literature. We propose a robust search strategy to retrieve and analyze this literature. Method: Analyses was focused on estimating the size of this literature and examining its intellectual anatomy and temporal trends using bibliometric indicators of its articles. Results: The size of the literature is estimated to have exceeded N = 25,000 items as of 2020. At the highest level of aggregation, patterns of term co-occurrence in road accident articles point to the presence of six major divisions: (i) law, legislation & road trauma statistics; (ii) vehicular safety technology; (iii) statistical modelling; (iv) driving simulator experiments of driving behavior; (v) driver style and personality (social psychology); and (vi) vehicle crashworthiness and occupant protection division. Analyses identify the emergence of various research clusters and their progress over time along with their respective influential entities. For example, driver injury severity ” and crash frequency show distinct characteristics of trending topics, with research activities in those areas notably intensified since 2015 Also, two developing clusters labelled autonomous vehicle and automated vehicle show distinct signs of becoming emerging streams of road accident literature. Conclusions: By objectively documenting temporal patterns in the development of the field, these analyses could offer new levels of insight into the intellectual composition of this field, its future directions, and knowledge gaps. Practical Applications: The proposed search strategy can be modified to generate specific subsets of this literature and assist future conventional reviews. The findings of temporal analyses could also be instrumental in informing and enriching literature review sections of original research articles. Analyses of authorships can facilitate collaborations, particularly across various divisions of accident research field.

简介：过去 50 年关于道路事故的学术研究产生了大量文献。我们提出了一种强大的搜索策略来检索和分析这些文献。方法：分析的重点是估计该文献的规模，并使用其文章的文献计量指标检查其知识结构和时间趋势。结果：文献规模估计已超过N = 截至 2020 年的 25,000 项。在最高聚合水平上，道路事故文章中的术语共现模式表明存在六个主要部门：(i) 法律、立法和道路创伤统计；(ii) 车辆安全技术；(iii) 统计建模；(iv) 驾驶行为的驾驶模拟器实验；(v) 驾驶员风格和个性（社会心理学）；(vi) 车辆耐撞性和乘员保护部门。分析确定了各种研究集群的出现及其随着时间的推移以及各自有影响力的实体的进展。例如， “驾驶员受伤严重程度”和碰撞频率显示出趋势主题的明显特征，自 2015 年以来，这些领域的研究活动显着加强。此外，标记为自动驾驶汽车和自动驾驶汽车的两个发展中的集群显示出成为道路事故文献新兴流的明显迹象。结论：通过客观地记录该领域发展的时间模式，这些分析可以为该领域的知识构成、其未来方向和知识差距提供新的洞察力。实际应用：可以修改建议的搜索策略以生成该文献的特定子集并协助未来的常规评论。时间分析的结果也可能有助于为原始研究文章的文献综述部分提供信息和丰富。对作者身份的分析可以促进合作，特别是在事故研究领域的各个部门之间。