Paul A O, Kay M V, Huppertz T, Mair F, Dierking Y, Hornburger P, Mutschler W, Kanz K-G
Chirurgische Klinik und Poliklinik, Campus Innenstadt, Klinikum der Universität München, München.
Unfallchirurg. 2009 Jan;112(1):23-30, 32. doi: 10.1007/s00113-008-1517-6.
Successful management of a mass casualty incident requires integrated operating procedures. A common division of victims into descriptive needs-based groups and the corresponding decision processes is the key to ensuring a successful operational response. The mSTaRT ("modified simple triage and rapid treatment") algorithm should enable emergency medical technicians to conduct triage, perform appropriate medical interventions, and coordinate transportation to adequate care facilities. The aim of this study was to design a concept to validate the mSTaRT algorithm.
Standardized evaluation sheets were distributed to emergency medical services (EMS) staff to prospectively classify trauma patients according to the mSTaRT algorithm: red (immediate: critically injured patients who can be helped by immediate transport), yellow (urgent: severely injured patients whose transport can be delayed), or green (delayed: patients with minor injuries who need help less urgently). The patients were then reevaluated in the emergency department, and the results were compared. The main points of the comparison were consistency of triage category and rates of overtriage and undertriage.
The study included 151 trauma patients. Of these, 62.3% were triaged correctly, 10.6% were overtriaged (2.6% critical overtriage), and 27.1% were undertriaged (4.0% critical undertriage). In the critically injured (immediate) category, the positive likelihood ratio (LR+) was 17.3 (95% CI 3.8-795), and the negative likelihood ratio (LR-) was 0.51 (95% CI 0.22-0.83). The probability of identifying a critically injured (immediate) patient was 17.3 times higher than the probability of identifying a severely (urgent) or minor (delayed) injured patient as immediate. Therefore, the rate of overtriage was very low. But every second patient who should have been classified as immediate was undertriaged by the EMS personnel. This undertriage was due to patients' suffering from head trauma, a well-known problem in the clinical context but a new problem in the triage context.
The results of our pilot study show that by using mSTaRT, patients designated as yellow (urgent) and green (delayed) will be accurately distinguished from red (immediate) patients; therefore, only a small number of patients will be overtriaged as red. However, some patients with severe head injury may not be initially assigned to the red category as required, resulting in undertriage. Consequently, modification of the mSTaRT procedures should be considered. A further identifier in the algorithm or checkpoint in the process should act as a safety net for catching severe head injury. A larger data set is required to further validate the mSTaRT algorithm. This will be acquired by means of a multicenter study.
成功管理大规模伤亡事件需要综合的操作程序。将受害者按基于需求的描述性分组以及相应的决策过程进行常见划分，是确保成功操作响应的关键。改良的简单分诊与快速治疗（mSTaRT）算法应能使急救医疗技术人员进行分诊、实施适当的医疗干预，并协调转运至合适的护理机构。本研究的目的是设计一个概念来验证mSTaRT算法。
向急救医疗服务（EMS）人员分发标准化评估表，以便根据mSTaRT算法对创伤患者进行前瞻性分类：红色（立即：可通过立即转运得到救治的重伤患者）、黄色（紧急：重伤患者，其转运可延迟）或绿色（延迟：轻伤患者，不太急需帮助）。然后在急诊科对患者进行重新评估，并比较结果。比较的要点是分诊类别一致性以及过度分诊和分诊不足的发生率。
该研究纳入了151名创伤患者。其中，62.3%被正确分诊，10.6%被过度分诊（2.6%为严重过度分诊），27.1%被分诊不足（4.0%为严重分诊不足）。在重伤（立即）类别中，阳性似然比（LR+）为17.3（95%可信区间3.8 - 795），阴性似然比（LR-）为0.51（95%可信区间0.22 - 0.83）。识别重伤（立即）患者的概率比将重伤（紧急）或轻伤（延迟）患者误识别为立即患者的概率高17.3倍。因此，过度分诊率非常低。但每两名本应被归类为立即的患者中就有一名被EMS人员分诊不足。这种分诊不足是由于患者患有头部创伤，这在临床环境中是一个众所周知的问题，但在分诊环境中是一个新问题。
我们的初步研究结果表明，通过使用mSTaRT，被指定为黄色（紧急）和绿色（延迟）的患者将与红色（立即）患者准确区分开；因此，只有少数患者会被过度分诊为红色。然而，一些严重头部受伤的患者可能最初未按要求被归入红色类别，导致分诊不足。因此，应考虑对mSTaRT程序进行修改。该算法中的另一个标识符或过程中的检查点应作为捕捉严重头部损伤的安全保障。需要更大的数据集来进一步验证mSTaRT算法。这将通过多中心研究来获取。
