Methods & Patients

Patients

From January 1st, 2016 to December 31st, 2017 a cohort of 48.629 routine patients or sera of routine patients were referred to the Floridsdorf Allergy Center with 91.438 diagnoses (many had more than one referral diagnosis). Of these, 3.875 (8.0%) carried a diagnosis compatible with a history of a DHR. In 1.532 individuals (1085 female / 447 male; 40.6 years ± 22.1) the suspected drug belonged to penicillin and/or cephalosporine antibiotics and a serological test was made. As the laboratory of the allergy clinic also serves as a tertiary referral centre for external serological tests, clinical data was not available for 76 patients that were eliminated leaving 1456 patients for the intention-to-treat analysis (Figure 1). A significant proportion of 523 patients had to be excluded from the per-protocol population because they did not show up for their scheduled skin tests and another single patient stepped down from skin testing on the day of the test (low compliance). This resulted in 932 individuals (669 female / 263 male; 42.5 years old ± 22.1) available for the per-protocol analysis (Table 1 and Figure 1).
Supplementary Figure 2 depicts the standard algorithm, which was a modified approach according to the guideline of the German speaking countries (21) adapted for the needs of our allergy outpatient clinic without a possibility for performing DPTs. The attending physician could deviate from the algorithm according to individual patient-specific factors. The primary outcome (DHR ‘confirmed’, or ‘possible’, or ‘unresolved’) depended on the interpretation of the summary of all available tests by the attending physician.

Materials & Methods

Specific IgE, total IgE and serum tryptase were measured on an ImmunoCAP 250 laboratory robot with commercially available tests from ThermoFisher (Uppsala, Sweden): Penicilloyl G (c1) & V (c2), Amoxicilloyl (c6), Ampicillin (c5), Minor determinate mixture (MDM) (U233), Cefaclor (c7). Due to production limits of the manufacturer and the high demand at our centre, not all test reagents were available during the whole study period (especially c5, c6, c7 and U233).
Skin prick (SPT), intradermal (IDT) and patch tests (PT) were performed with nationally licensed drugs for intravenous use in nationally recommended concentrations and read accordingly (21, 22): Penicillin G, Amoxicillin/Clavulanic Acid, Cefazolin and Ceftriaxon: “Penicillin G-Natrium Sandoz”, “Curam®”, “Cefazolin Sandoz”, “Ceftriaxon Sandoz”, all from Sandoz, Kundl, Austria; Ampicillin/Sulbactam: “Unasyn®”, Pfizer, Borgo San Michele, Italy; Cefuroxim: “Cefuroxim MIP”, Cephasaar, Sankt Ingbert, Germany). The commonly used penicillin derivatives MDM & PPL for skin tests marketed by Diater, Madrid, Spain are not licensed in Austria and cannot be used in routine settings outside academic hospitals. PTs were performed using Curatest® (Lohman und Rauscher, Vienna, Austria) and read after 24 hours together with the late reading of the IDT.
Generally, skin tests were performed in the following order:
1st) all SPTs at once, when negative after 20 min followed by
2nd) all IDTs at once, when negative after 20 min followed by
3rd) all PTs at once followed by
4th) late readings of all tests after 24 hours.

Ethics and Statistics

The study was approved by the ethics committee of the Medical University of Vienna, Austria, during Christian Ostermayer’s medical diploma thesis according to the Helsinki Declaration of Human Rights (ECS 1103/2018). Χ²-tests were calculated using MedCalc Statistical Software version 19.2 (MedCalc Software Ltd, Ostend, Belgium; https://www.medcalc.org; 2020).