Relationship of Reaction History to Positive Penicillin Skin Tests

Abstract

Previous data suggest that up to one-third of patients classified as allergic based on positive penicillin skin tests have a vague reaction history. Direct oral challenge (DOC) has been recommended for patients with a low-risk reaction history. A variety of published models stratify reaction risk to guide the use of DOC. To reassess the proportion of penicillin skin test-positive patients with vague or low-risk reaction histories and evaluate the relationship between the reaction risk history and the likelihood of positive skin test results. We identified patients who underwent penicillin allergy evaluation over a 5-year period. We recorded drug reaction history, demographic variables, skin testing, and challenge results. Matched controls whose skin tests were negative were identified for skin test-positive patients. Drug reaction histories were assigned a risk category based on 2 previously published risk stratification models. We used logistic regression to investigate whether reaction history risk was associated with positive skin test results. Penicillin skin testing was performed in 3382 patients; 207 (6.1%) were positive. Positive skin tests were more frequent in outpatients (P < .001), younger patients (P < .001), and female patients (P < .001). Percentages of each risk category in each model were similar in cases versus matched controls. The likelihood for positive skin tests increased with a high-risk reaction history in one stratification model. Our data confirm that a substantial proportion of patients who self-report penicillin allergy and have positive skin test results have a low-risk history and imply that penicillin skin testing is associated with a poor positive predictive value.