Talmor Daniel, Jones Alan E, Rubinson Lewis, Howell Michael D, Shapiro Nathan I
Trauma Anesthesia and Critical Care, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, USA.
Crit Care Med. 2007 May;35(5):1251-6. doi: 10.1097/01.CCM.0000262385.95721.CC.
In the event of pandemic influenza, the number of critically ill victims will likely overwhelm critical care capacity. To date, no standardized method for allocating scarce resources when the number of patients in need far exceeds capacity exists. We sought to derive and validate such a triage scheme.
: Retrospective analysis of prospectively collected data.
Emergency departments of two urban tertiary care hospitals.
Three separate cohorts of emergency department patients with suspected infection, comprising a total of 5,133 patients.
None.
A triage decision rule for use in an epidemic was developed using only those vital signs and patient characteristics that were readily available at initial presentation to the emergency department. The triage schema was derived from a cohort at center 1, validated on a second cohort from center 1, and then validated on a third cohort of patients from center 2. The primary outcome for the analysis was in-hospital mortality. Secondary outcomes were intensive care unit admission and use of mechanical ventilation.
Multiple logistic regression demonstrated the following as independent predictors of death: a) age of >65 yrs, b) altered mental status, c) respiratory rate of >30 breaths/min, d) low oxygen saturation, and e) shock index of >1 (heart rate > blood pressure). This model had an area under the receiver operating characteristic curve of 0.80 in the derivation set and 0.74 and 0.76 in the validation sets. When converted to a simple rule assigning 1 point per covariate, the discrimination of the model remained essentially unchanged. The model was equally effective at predicting need for intensive care unit admission and mechanical ventilation.
If, as expected, patient demand far exceeds the capability to provide critical care services in an epidemic, a fair and just system to allocate limited resources will be essential. The triage rule we have developed can serve as an initial guide for such a process.
在大流行性流感爆发时，重症患者的数量可能会超出重症监护能力。迄今为止，当有需求的患者数量远远超过医疗能力时，尚无分配稀缺资源的标准化方法。我们试图推导并验证这样一种分诊方案。
对前瞻性收集的数据进行回顾性分析。
两家城市三级护理医院的急诊科。
三组独立的疑似感染急诊科患者队列，共5133例患者。
无。
仅使用患者初次到急诊科时即可获得的生命体征和患者特征，制定了用于疫情期间的分诊决策规则。分诊方案源自中心1的一个队列，在中心1的第二个队列上进行验证，然后在中心2的第三个患者队列上进行验证。分析的主要结局是住院死亡率。次要结局是重症监护病房入住率和机械通气的使用情况。
多因素逻辑回归显示以下因素为死亡的独立预测因素：a）年龄>65岁，b）精神状态改变，c）呼吸频率>30次/分钟，d）低氧饱和度，e）休克指数>1（心率>血压）。该模型在推导集中的受试者工作特征曲线下面积为0.80，在验证集中分别为0.74和0.76。当转换为每个协变量赋予1分的简单规则时，模型的辨别力基本保持不变。该模型在预测重症监护病房入住需求和机械通气需求方面同样有效。
如果如预期那样，在疫情期间患者需求远远超过提供重症监护服务的能力，那么一个公平公正的有限资源分配系统将至关重要。我们制定的分诊规则可作为这一过程的初步指导。
