Martijn Sijbom

120 Chapter 5 Quality analysis of the allergy registration Quality analysis consisted of an assessment on completeness and correctness of the antibiotic allergy registration in the primary care EMR based on a previously published checklist by Salden et al. (S1 Table) (1). The checklist was modified for one item: the maximal time between start of symptoms and first intake of antibiotic was extended to up to 6 hours for immediate type allergies (See Box 1, Immediate type versus delayed type antibiotic allergy). Assessment was conducted with information available in the registration. A complete registration was defined as a registration that contained a description of symptoms and time to onset of symptoms and duration of symptoms. Antibiotic allergy registrations were then classified as an ‘immediate type reaction’ (possible/probable), ‘delayed type reaction’ (possible/probable), ‘non-allergic side effect’ or ‘insufficient data available for diagnosis’. A correct antibiotic allergy registration was defined as a registration in which the conclusion was concordant with the diagnosis according to the modified checklist. Box 1. Immediate type versus delayed type antibiotic allergy Immediate type allergies are IgE mediated reactions. The symptoms are the result of immediate release of histamine and other cytokines upon exposure to an allergen. The most frequently reported symptoms are urticaria, angio-oedema, exanthema, dyspnoea and hypotension, and occur within a few hours. This is opposed to delayed type reactions, which generally develop a few days after exposure, as they are cell-mediated. A mild exanthema is the most frequent delayed type reaction. To represent daily practice, analysis of antibiotic allergy registrations was limited to the five antibiotic groups most frequently prescribed in primary care in The Netherlands: penicillins, tetracyclines, nitrofuran derivatives (i.e. nitrofurantoin), macrolides and fluoroquinolones (10). A sample of 300 antibiotic allergy registrations was obtained for quality analysis. The size of the random sample was calculated using a random sample formula (12). We used a confidence level of 90% and a margin of error of 5%, including the entire ELAN data warehouse population for each type of registration. These 300 patients were selected through randomisation by SPSS (version 25, SPSS Inc., Chicago, IL). If a patient had multiple antibiotic allergy registrations, one registration was randomly selected and used for further analysis.

RkJQdWJsaXNoZXIy MTk4NDMw