Introduction

About 10% of the population report an adverse reaction to “penicillin” (1). While the term “penicillin” in common language is often misused as a representant for the large group of β-lactam antibiotics including cephalosporins (common structure see supplementary Figure 1) or more general as a synonym for “antibiotic”, the word “allergy” in this context commonly represents various sorts of drug hypersensitivity reactions (DHR). “Drug allergy” entries appear as red flags in electronic health records of up to 35% of patients upon which “penicillin” is the most frequently mentioned suspected drug (2). However, only 4% of these patients will show positive reactions when tested with penicillin either by skin, blood or provocation tests (3). There are several reasons to explain this high rate of obviously false histories:
  1. Although a DHR may have been caused by a drug, the pathophysiological mechanism may not be an immunological one. Most adverse reactions are caused by the specific pharmacological mechanism of the antibiotic and are classified as non-allergic type A reactions. Antibiotics not only kill the pathogenic but also – as an obligatory side effect – the essential symbiotic bacteria leading to e.g. gastrointestinal malfunction. Type A reactions comprise more than 80% of all DHR and are no contraindication for a future reintroduction of the suspected drug even without testing (4, 5).
  2. A true allergic reaction (also referred as B-type DHR(4)) is usually regarded as a contraindication for future use. It may be of the immediate, anaphylactic type I (IgE-mediated), of the delayed type IV allergy (T-cell mediated), and rarely of the type II (IgG mediated) or type III (IgG/IgM mediated). Still, also immunological memory may diminish and even completely disappear over the years (6).
  3. Differential diagnoses such as urticaria/angioedema driven by infection for immediate type reactions (7) and viral exanthema for delayed reactions are much more frequent than DHRs (8). Infections can frequently result in a prescription of an antibiotic, which in turn may easily be misinterpreted as being the cause of the DHR.
This situation bears considerable risks not only to the patient, who may be prescribed unnecessary alternative, sometimes less effective antibiotics, but also to society because the overuse of alternative antibiotics can propagate antibiotic resistance (9, 10). Hence, current guidelines on the management on DHRs demand an allergy workup for a “de-labelling” of false histories of DHR in the case of important drugs and β-lactam antibiotics are generally regarded as belonging to such a kind (4, 11).
About how to reach this goal, there seems to be some disagreement on both sides of the Atlantic (12). While Europeans and US experts agree on the importance of skin testing, there are different views on the usefulness of in vitro tests (13). Skin tests offer the advantage of giving an immediate result and are cheap for the healthcare system (14). While their sensitivity is limited their specificity is high (15).In vitro tests are the safest test for patients. Unfortunately, there are not many marketed, standardized tests except for drug-specific IgE (sIgE) to β-lactam antibiotics including the cephalosporin cefaclor. Measurement of sIgE has a low sensitivity but at a high specificity (16).
In 2018, a basic DHR test without drug provocation tests (DPT) was calculated causing costs of US$ 220 for the US healthcare system. Including DPTs increased the price to at least US$ 359 (17). Nevertheless, in recent times some US experts have gone even one step further and started propagating DPTs for mild reactions even without a prior skin or in vitro tests (10, 18-20).
With this study we wanted to take a step back and ask, how many cases of suspected DHRs to β-lactams could be solved by applying simple, broadly available methods causing only limited costs without the resource of performing DPTs. We report the results of a retrospective chart review of a cohort of 932 patients with a history of DHRs to β-lactam from the years 2016 to 2017 from a single centre outpatient clinic. The patients underwent the following algorithm: 1) DHR-specific history, 2) drug-specific IgE test (depending upon availability), 3) a series of skin prick, intradermal and patch tests.