Airway, Breathing or Circulation Failure in Fatal Food Anaphylaxis: A Nationally Representative Case Series

The World Allergy Organisation (WAO) characterises severe anaphylaxis as ‘a potentially life-threatening compromise in airway, breathing and/or the circulation’ due to anaphylaxis [1]. Although food is the most common trigger for fatal anaphylaxis in children in the United Kingdom [2], fortunately fatal food-related anaphylaxis in children occurs very infrequently [3]. In terms of acute treatment of severe anaphylaxis, current international guidelines [1] identify the administration of a patient's own adrenaline auto-injector as the first step of management, followed by a repeated dose if there is no response within 5 min. The second step includes additional interventions, personalised to the patient's history, upon the arrival of health-care professionals [1]. Furthermore, the evidence base that underpins these treatments' effectiveness, including that of intramuscular adrenaline is limited [4]. Previous work has identified that there are different patterns in systems compromised (airway, breathing or circulation) in anaphylaxis caused by different triggers [5]. Fatal food-related anaphylaxis has been more strongly associated with airway and breathing compromise when compared to fatal medication or venom-associated anaphylaxis [5, 6]. The WAO recommends that further elucidation of underlying mechanisms of anaphylaxis is required to better characterise anaphylaxis phenotypes and endotypes [3]. To this end, we have performed a review of cases of fatal food anaphylaxis registered in England's National Child Mortality Database (NCMD) between 1 April 2019 and 31 March 2023. The NCMD is funded by the National Health Service in England (NHS England). It collects statutory data on all children in England who die before their 18th birthday. Data collection started on 1 April 2019. This review is based on data for children who have died due to anaphylaxis, for which the child death review was completed by the child death overview panel and recorded on the NCMD system. Case identification followed the same methodology as used in the NCMD publication Child deaths due to asthma or anaphylaxis [7]. The review aimed to identify if airway, breathing or circulatory failure or a combination of failures was present leading to death in each of these cases. Data review was performed independently by 2 authors (J.C. and T.R.). Reviewers reviewed all available documents including NCMD notification form, asthma and anaphylaxis supplementary reporting form, post-mortem reports, pre-hospital reports, ambulance log calls and any other documentation available on the database. Reviewers were asked to identify if anaphylaxis was present based on the WAO definition [3]. They were then asked to identify if food was the trigger and then whether airway, breathing or circulatory compromise was present in each case. They used the criteria to define airway, breathing or circulatory compromise that have been developed by an expert panel including members from allergy and clinical immunology, emergency medicine, anaesthesia and primary care [8]. Systems compromised were defined based on the presence of one of the following symptoms or signs: Airway: Stridor, altered voice, oropharyngeal swelling, laryngeal swelling at laryngoscopy, difficulty breathing due to throat tightness, significant laryngeal swelling at autopsy. Breathing: Wheeze, any bronchodilator use other than adrenaline, patient's familiar asthma symptoms, shortness of breath with a history of asthma, hypoxia or cyanosis while upright/with perfusing blood pressure, non-airway increased work of breathing, shortness of breath while in an upright posture, chest tightness with a history of asthma, high peak airway pressures in a ventilated patient, obstructive (upsloping or low amplitude) CO2 or obstructive airflow pattern (mechanical ventilation), difficult to ventilate. Circulation: Feeling faint, presyncope, hypotension, syncope, postural dizziness without significant breathing problems, cyanosis or severe tachycardia (prior to adrenaline) without increased work of breathing or bronchospasm symptoms. The reviewers had the option to decide that there was insufficient data available to identify the type of compromise present. Where cases were excluded from analysis or there was disagreement between authors, a third author (N.S.) reviewed all available data and adjudicated. Of a total of 24 cases reviewed, 17 were included. The reasons for exclusion were: insufficient data (five cases); drug-induced anaphylaxis (one case), death due to another cause (one case). Agreement between the initial two reviewers in terms of inclusion and systems compromised was present in 21 of 24 cases (86%). Demographic details of the included cases are displayed in Table 1. In 8 of the 17 cases reviewed (47%) airway and breathing compromise were identified, while breathing compromise alone was also identified in eight cases (47%). In a single case, cardiovascular compromise, without documented airway or breathing compromise was identified (6%). In this case the child had eaten food within 30 min of the symptom onset, developed vomiting and then rapidly collapsed in the presence of family members without any reported airway or breathing symptoms and went into cardiac arrest suddenly. No signs of airway swelling or wheeze were documented by the pre-hospital team who resuscitated the child. The child had a significantly elevated Mast-Cell Tryptase level. This case series of fatal food-related anaphylaxis suggests that the systems compromised prior to the onset of cardiac arrest are most commonly breathing, with or without airway compromise. This is consistent with previous work linking asthma and acute wheezing with fatal or non-fatal food anaphylaxis [5, 6]. From a primary prevention point of view, this should inform advice given to families, carers and healthcare professionals regarding anaphylaxis recognition in children and young people with food allergy. Breathing difficulty appears to be a relevant sign of potentially life-threatening symptoms. In the context of an allergic reaction, and where there is no other obvious cause, breathing difficulty should trigger anaphylaxis management. As noted in the NCMD publication Child deaths due to asthma or anaphylaxis [7] in all cases of fatal food-related anaphylaxis included in this review children had a premorbid diagnosis of asthma and therefore advice given to families should focus on the treatment of both conditions concurrently when acute airway or breathing symptoms develop. This case series suggests that in paediatric food-related anaphylaxis, in addition to adrenaline treatment, further personalised treatment for those with anaphylaxis refractory to initial treatment should focus on early interventions to improve oxygenation rather than a focus on circulatory problems. Of note, these findings are unlikely to be generalisable to either the adult population or to non-food-related anaphylaxis where death is commonly associated with circulatory collapse [9]. A strength of the study is that to limit misclassification bias, we used a consensus definition of the systems affected, and the cases were independently assessed by two reviewers with a high degree of agreement. While a strength of the analysis is that all cases identified as anaphylaxis in England over the study period will have been reported to the database, it is possible, although unlikely, that a further group of patients with unrecognised fatal food-related anaphylaxis has different characteristics from this group. Of note, only fatal cases have been reviewed, and severe non-fatal cases of anaphylaxis were not included, and it is possible that the surviving cohort has different characteristics to this group. In summary we have found that in a case series of paediatric fatal food-related anaphylaxis, compromise of airway and breathing is most common, followed by breathing compromise alone, whilst circulatory compromise without prior breathing or airway compromise appears to be rare. All authors conceived the presented idea. J.C. and T.R. performed a systemic clinical review of each case. N.S. performed a systemic clinical review of selected cases where J.C. and T.R. disagreed. All authors discussed the results and contributed to the final manuscript. The authors declare no conflicts of interest. The data that support the findings of this study are available from the corresponding author upon reasonable request.