Development Of A Clinical Predictive Score For Allergic Reactions To COVID-19 Vaccine (AZD1222)

Abstract

Rationale COVID-19 vaccine hypersensitivity reaction is an important issue that associated with COVID-19 vaccine hesitancy for patients’ preference and physician advice. Validated clinical decision rules are in needed to identify individuals who may develop allergic reaction following vaccination. We aimed to develop a predictive score for allergic reactions to COVID-19 vaccine. Methods A case-control study was conducted at Thammasat University Hospital between March 1, 2021 and September 30, 2021. Cases were identified as individuals who received AZD1222 and developed allergic reaction which included urticaria, angioedema, anaphylaxis, or maculopapular rash following vaccination. Controls were randomly sampled from AZD1222 recipients who reported no allergic reaction. Risk scores were derived from the coefficients of significant variables from final multivariate logistic regression model. Results One hundred and thirteen cases and 226 controls were enrolled. The model demonstrated good performance in predicting allergic reactions after vaccination with an area under the ROC curve of 0.85. Variables included in the developing of predictive model were age <60 years (2 points), female (4 points), personal history of food allergy (3 points), allergic rhinitis (2 points), atopic dermatitis (2 points), and family history of allergic disease (1 point). A cutoff of total risk score ≥7 points provided a specificity of 90.7%, sensitivity of 63.7%, and +LR of 6.86. Conclusions Our predictive score of ≥7 points showed high specificity of the test. This scoring system is useful for identification individuals who may develop allergic reaction following COVID-19 vaccination. The model is simple and easy to apply in clinical practice.