Emergency Medical Services (EMS) and the California EMS Information System (CEMSIS) working paper [R12]

The data shows that there are substantial differences in response, scene, and transport times between collisions that occur in urban and rural zip codes. However, there are several limitations influencing interpretation of the results. Data on EMS response, scene, and transport times are missing for much of the state (see Fig A1- 3 in Appendix). Even for zip codes that have records of EMS events, the dataset is likely incomplete (i.e., not all EMS responses to motor vehicle collisions in a zip code are recorded). Therefore, the data recorded in CEMSIS may not be representative of EMS events throughout California. Additionally, zip code level location data is insufficient for adequate study of the effects of the built environment and road network on prehospital time. The analyses that can be conducted using zip code level data are limited by missing data. Fields such as patient condition are not populated, thus analyses for such factors cannot be completed. For example, without data on patient condition, the effect of response time on health outcomes cannot be determined. CEMSIS, like NEMSIS, is a convenience sample, in which data are submitted voluntarily by local EMS agencies. According to the NEMSIS User Manual, the dataset suffers from selection bias, which occurs when a perceptible difference between two groups is caused by different criteria being included. The dataset also suffers from information bias which is a difference between two groups that is due to differences in the data available for comparison (NEMSIS, n.d.). The problems that the national database experiences are likely also present in CEMSIS data.

本数据集显示，城市与乡村邮政编码区域内发生的机动车碰撞事故，其应急医疗服务（Emergency Medical Services，EMS）响应时长、现场处置时长与转运时长存在显著差异。然而，若干局限性会对本研究结果的解读造成制约。
该州大部分地区的EMS响应、现场处置及转运时长数据存在缺失（详见附录图A1-3）。即便部分邮政编码区域留存有应急医疗事件的相关记录，该数据集仍可能存在不完备性——即某一邮政编码区域内针对机动车碰撞事故的全部EMS响应事件并非均被纳入收录范围。因此，加州应急医疗服务信息系统（California Emergency Medical Services Information System，CEMSIS）所收录的数据，未必能代表加州全境的EMS事件整体情况。
此外，以邮政编码为单元的位置数据不足以支撑对建成环境与道路网络影响院前急救时长的充分研究。基于邮政编码层级数据开展的分析亦因数据缺失而受到限制。诸如患者病情等字段未被填充，因此无法针对此类因素开展相关分析。例如，若缺乏患者病情相关数据，则无法明确响应时长对健康结局的影响效应。
与国家应急医疗服务信息系统（National Emergency Medical Services Information System，NEMSIS）类似，CEMSIS属于便利抽样样本，其数据由各地EMS机构自愿提交。据NEMSIS用户手册所述，该数据集存在选择偏倚——即当两组群体间的可观测差异源于不同的纳入标准时，便会产生此类偏倚。该数据集同时存在信息偏倚，即两组群体间的差异源于可用于对比的数据集本身存在差异（NEMSIS，未注明日期）。国家应急医疗服务信息系统所存在的上述问题，大概率也同样存在于CEMSIS数据集之中。