Background

Substantial discrepancies among anaphylaxis severity scores may delay epinephrine administration.

Objective

The study aims to develop a transparent severity grading system of food-induced acute allergic reactions with a decision model for epinephrine use.

Methods

The natural course of 315 acute food-induced allergic reactions in children hospitalized at the Allergology department between May 2016 and July 2019 owing to follow-up treatment and allergy diagnostics was evaluated. The severity of episodes was classified according to the 5 most accepted grading systems. The interrater reliability of classification between anaphylaxis severity scores was assessed. All symptoms were grouped into a heat map according to their real-life incidence and clinical relevance. Based on the heat map analysis, a severity grading system of food-induced acute allergic reactions in children with the epinephrine administration decision model was created.

Results

Data from 259 food-induced anaphylaxis episodes in 157 children were included in the analysis. Comparing the grading systems, we observed a 24.7% to 70.2% disagreement between severity scores. The heat map illustrated a strong association between 29 symptoms and their categorization. A new severity grading system was developed and a 2-stage decision model was proposed: “epinephrine yes” (any rapidly progressing symptoms, even mild ones or from 1 organ system; any symptoms from more than 1 organ system; or every grade of anaphylaxis), and “epinephrine available and prepared to use” (nonprogressing mild systemic allergic reaction from 1 system area only; no anaphylaxis).

Conclusion

A new severity grading system of food-induced acute allergic reactions in children could serve as a clinical tool for health care professionals to avoid epinephrine administration delay.

中文翻译：

---

食物引起的急性过敏反应的严重程度分级系统，以避免延迟服用肾上腺素

背景

过敏反应严重程度评分之间的巨大差异可能会延迟肾上腺素的给药。

客观的

该研究旨在开发一个透明的食物引起的急性过敏反应严重程度分级系统，并提供肾上腺素使用的决策模型。

方法

评估了 2016 年 5 月至 2019 年 7 月期间因随访治疗和过敏诊断而在过敏科住院的 315 例急性食物过敏反应的自然过程。发作的严重程度根据 5 个最被接受的分级系统进行分类。评估了过敏反应严重程度评分之间分类的交互者可靠性。所有症状根据其现实生活中的发病率和临床相关性被分组到一个热图中。基于热图分析，建立了具有肾上腺素给药决策模型的儿童食物诱发的急性过敏反应严重程度分级系统。

结果

分析包括 157 名儿童的 259 次食物引起的过敏反应发作的数据。比较分级系统，我们观察到严重性评分之间的差异为 24.7% 至 70.2%。热图显示了 29 种症状与其分类之间的强烈关联。开发了一个新的严重程度分级系统并提出了一个 2 阶段决策模型：“肾上腺素是”（任何快速进展的症状，即使是轻度症状或来自 1 个器官系统的症状；来自 1 个以上器官系统的任何症状；或每个级别的过敏反应) 和“肾上腺素可用并准备使用”（仅来自 1 个系统区域的非进展性轻度全身过敏反应；无过敏反应）。

结论

一种新的儿童食物引起的急性过敏反应严重程度分级系统可以作为医疗保健专业人员的临床工具，以避免延迟服用肾上腺素。