Discussion
Anaphylaxis management and severity labeling is still a subject of debate in 2024, even within a specialized allergy department. The lack of a unanimous international consensus and the multiplicity of published classifications complicate the management of patients in an emergency setting. To our knowledge, this study is the first one comparing different classifications of anaphylaxis and severity levels by using the WHO classification proposed for the ICD-117 as the reference one. The results of this study demonstrate a certain degree of variability in anaphylaxis classification systems and therefore their implications for treatment management.
Given the multiple definitions of anaphylaxis, several groups proposed severity scores, often based on expert opinions, but only few of them have been validated14-17. In our study, each one of the 4 assessed classification provided different degrees of severity, when compared with the ICD-11 one. It is clear that the grading value differ, depending on the evaluated classification: indeed, Muraro’s classification10 has 3 levels of severity, Blazowski’s one11 has 4, while those from CoFAR8and Sampson9 go from 1 to 5. Therefore, in clinical practice, when a patient presents, for example, a reaction classified as a grade 2 anaphylaxis, this could have a different meaning, following the used classification, including different possible symptoms and organs involved, and thus a different therapeutical approach. In his paper, Blazowski already highlighted the discrepancy between several classifications, some of them not necessarily used to classify food-induced anaphylaxis, and he underlined that a new severity grading system was needed, especially to harmonize the definition of severity and to avoid any delay in the administration of adrenaline11. Our work underlines the importance of using the same classification in a work unit and of communicating which one is employed, to better approach patients. Also, it would be helpful to internationally use the same grading system, to better inform and exchange clear information on patients between physicians. We decided to refer to the ICD-11 classification as a gold standard since it’s the one used on a daily base in our unit, and it’s easy to use; it is also the classification used by the WHO post-coding system, which is already used and will be even more implemented over the coming years, being available in all countries for all healthcare professionals and will therefore allow to identify even cases coded by non-allergists, especially primary care physicians.
Our study is based on oral food challenge results, during which vomiting is a common symptom, often combined with other symptomatic elements warranting the injection of adrenaline18. The most frequently reported symptom during OFC in our anaphylactic patients was abdominal pain. This result is noteworthy because abdominal pain is not always considered as one of the main symptoms of anaphylaxis. Also, in some countries as in the UK and Australia19, abdominal symptoms are not considered in the classification of food allergy reactions. However, the results of our study show that abdominal pain is often the first symptom found in patients further developing anaphylactic reactions. Therefore, the question arises as to whether this reported symptom should be considered as a red flag for a possible anaphylactic reaction. In our gold standard classification (ICD-11), it accounts for one organ involvement (the gastro-intestinal track). 98 patients (41.7%) developed anaphylaxis during the oral provocation test, while none of them had a clinical history of severe reaction. Main foods associated with a worsening of the reaction, compared with the one reported in patients’ clinical history, were peanuts (28.8% of patients), eggs (12.0%), and pistachios (8.0%). On the other hand, considering patients with a history of anaphylaxis, 45 (19.5%) presented an anaphylactic reaction during the OFC, while 16 (7%) did not. These data underline the fact that the evolution of a food allergy remains often unpredictable, and OFC should be performed only by well trained professionals, in a safe hospital environment.
Regarding the management of patients, our study revealed a delay, or even an absence, in the use of adrenaline in cases of anaphylaxis, regardless of the classification used by the physician. A study by Eller et al.20 emphasizes the existence of different severity grading systems for anaphylaxis and the use of adrenaline, leading to variations in its administration, if based on the grading system. There are no universally validated diagnostic tools to determine which symptoms warrant adrenaline treatment compared to those that do not21, which may explain the underuse of adrenaline found in our study. A recent study highlights the underuse of adrenaline even in pediatric emergency services in France22. Such issue poses a problem in many healthcare institutions, even within a specialized allergy service, which may be explained by two major factors: a lack of training in anaphylaxis management and hesitancy among healthcare professionals to use adrenaline (due to a fear regarding the bathmotropic effects of adrenaline) as well as simply a lack of recognition of anaphylactic reactions23. Additionally, the multitude of existing classifications and their perceived complexity may also contribute to a less clear indication for such a vital treatment. What physicians should remember is that adrenaline is the key treatment of severe allergic reaction, and that its injection during a positive OFC is safe, even when not strictly needed24. At the same time, most adverse reactions to the intramuscular injection are not serious and are transient, while the risk for a patient experiencing anaphylaxis may raise up to a possible fatality. These factors show that the risk-benefit ratio is strongly in favor of the use of adrenaline to promptly treat any anaphylactic reaction. At last, as for treatment of the reaction, we recorded a statistical difference in the use of corticosteroids, during an anaphylactic reaction. Even though such a therapeutical approach is not recommended5, we could speculate that the fear of not providing enough treatments wrongly pushed the physicians in including glucocorticoids in the list of the administered ones.
Our study presents some limitations: it is a retrospective study, and some data may therefore be missing from the medical records of certain patients. Also, the choice to use the ICD-11 classification as the reference one, well-justified for its ease of use in daily practice and its validation by the WHO as stated above, could potentially introduce biases. Nevertheless, we also present results that show a certain strength, considering the number of included patients, the double-blind verification of the classifications by two specialized allergists, and the use of four recent classifications for comparison purposes.
While the emergency of different classifications of anaphylaxis and of its severity is a real asset for clinical management, their multiplicity create confusion among healthcare professionals. Therefore, it would be important to consider consolidating these different classifications into one that is both appropriate and intuitive, favoring sensitivity with a good compromise regarding specificity. Our work highlights the need to refine theses scoring systems, to accurately capture anaphylactic reactions and ensure appropriate management, without neither over- nor under-treat patients. In conclusion, there is a real need to adopt a universal, intuitive, and easy-to-use classification, such as the ICD-11 one, while destigmatizing at the same time the use of adrenaline.