Frequency of self-reported drug allergy

Publication date: Available online 2 August 2017
Source:Annals of Allergy, Asthma & Immunology
Author(s): Bernardo Sousa-Pinto, João Almeida Fonseca, Eva Rebelo Gomes
BackgroundPatients reporting drug allergy are treated with second-line therapies, with possible negative clinical and health consequences.ObjectiveTo assess the prevalence of self-reported drug allergy.MethodsWe performed a systematic review of observational studies assessing the prevalence of self-reported drug allergy. We searched 4 electronic databases. From selected studies, we extracted data on self-reported drug allergy prevalence, study design, participants' demographic characteristics, reported clinical manifestations, and suspected culprit drugs. We performed a random-effects meta-analysis followed by a meta-regression.ResultsFifty-three studies were included in the systematic review, assessing a total of 126,306 participants, of whom 8.3% (range across studies 0.7–38.5%) self-reported drug allergy. Cutaneous manifestations were reported by 68.2% of participants, and anaphylactic or systemic reactions were reported by 10.8%. Antibiotics, nonsteroidal anti-inflammatory drugs, and anesthetics were the most frequently reported culprit drug classes. The frequency of self-reported drug allergy was higher in female (11.4%) than in male (7.2%) patients, adults (10.0%) than in children (5.1%), and in studies in the medical setting (15.9% in inpatients, 11.4% in outpatients) than in the general population (5.9%). The meta-analysis rendered a pooled prevalence of 7.9% (95% confidence interval 6.4–9.6), and the meta-regression identified study region, participants' age group, and study setting as factors associated with significant heterogeneity. Confirmation tests (including skin, in vitro, and drug provocation tests) were performed in only 3 studies.ConclusionThe prevalence of self-reported drug allergy is highly variable and is higher in female patients, adults, and inpatients. To overcome this variability, further studies using confirmation tests are needed.



http://ift.tt/2v2nQk0