Drug-induced anaphylactic deaths: Evidence-based data call for prevention measures

International audienceBackground: Drugs have been mentioned as the main cause of anaphylaxis deaths in some countries. The aim of this systematic review was to identify patterns and key issues in order to support prevention strategies and reduce avoidable deaths related to drug allergy/hypersensitivity (DA/H).Method: We searched Pubmed, MEDLINE, EMBASE, Cochrane Controlled Register of Trials, Lilacs databases for manuscripts concerning drug-induced anaphylaxis (DAF) fatalities, without publication date or language restrictions. We screened studies, extracted data, and assessed risk of bias independently in duplicate. The manual review was performed by two independent persons.Results: We selected 21 studies, most (71%) published over the last two decades. Most of publications were from USA (38%) and Europe (29%). 57% were national populational-based studies and 38% were based on registries. Most articles (52%) used the International Classification of Diseases (ICD) to access the data. DAF rate has been estimated by 0.13 to 0.53/106 population/year. General global trends of DAF indicate increasing numbers over time, the majority (89%) occurring at health care settings and in female gender (52%) with an average age of 59 years. Beta-lactam antibiotics (12%), contrast media (8%) and anesthetics (3%) were the main culprit agents. Personal history of asthma (18%) and DA/H (9%) have been mentioned as potential risk factors, but personal history of atopic disease and mastocytosis were seldom mentioned. The main manifestations described were cardiovascular (10%) and respiratory commitment (9%). Remarkable was the description of use of adrenaline mentioned in only 24% papers.Conclusion: DAF is increasing worldwide and can be iatrogenic. This first systematic review on DAF could identify a phenotype at risk for DAF: women, aged more than 59, with personal history of asthma and DA/H with need of health care assistance. Mortality data will gain accuracy due to the implementation of the ICD-11. We could highlight key aspects of population-based studies able to support preventive strategies and reduce avoidable deaths due to DA/H