In a recent study published in the journal JAMA, researchers investigated the link between commonly prescribed oral antibiotics and the subsequent development of serious cutaneous adverse drug reactions (cADRs). They used long-term (20 years) records from a large dataset of elderly participants (age > 66, n = 3,257,181) who suffered antibiotic-related cARDs (n = 21,758) compared to controls that did not (n = 87,025).

Study findings revealed that commonly consumed oral antibiotics, particularly sulfonamides (aOR = 2.9), cephalosporins (aOR = 2.6), nitrofurantoins (aOR = 2.2), penicillins (aOR = 1.4), and fluoroquinolones (aOR = 1.3) resulted in significantly higher cADRs, hospitalizations, and mortality rates compared to macrolides. These findings highlight the need for medical practitioners to select lower-risk antibiotic prescriptions wherever clinically possible.

Study: Oral Antibiotics and Risk of Serious Cutaneous Adverse Drug Reactions. Image Credit: areeya_ann / Shutterstock

Background

Cutaneous adverse drug reactions (cARDs), commonly named ‘toxidermia,’ are a group of rare but life-threatening hypersensitivity reactions affecting the skin, mucous membranes, and adnexa. Common examples of serious cARDs include toxic epidermal necrolysis (TEN) and Stevens-Johnson syndrome (SJS), the former of which has a known mortality rate of 20-40%.

Previous research has identified numerous drugs, particularly antibiotics, as risk factors in the development of cARDs. Unfortunately, no studies have investigated the differential contributions of different antibiotic classes of cARD risk. This knowledge would be essential for medical practitioners, allowing them to preferentially select low-risk antibiotics when patients’ clinical symptoms allow for multiple treatment options.

Optimal antibiotic selection is instrumental in older patient populations, given their disproportionately high antibiotic consumption, comorbidity risk, and polypharmacy.

About the study

The present study follows a population-based study design to explore the associations between different classes of oral antibiotics and their relative subsequent cARDs risk. Data for the study was obtained from the administrative health databases of Ontario, Canada, between April 2002 and March 2022. Specifically, the Ontario Drug Benefit database was perused for antibiotic data, the Canadian Institute for Health Information (CIHI) National Ambulatory Care Reporting System for hospitalization data, and the Registered Persons Database for participants’ demographic information.

The study methodology follows the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) and Reporting of Studies Conducted Using Observational Routinely Collected Health Data Statement for Pharmacoepidemiology (RECORD-PE) guidelines. Participants were included in the study if they were at least 65 at the time of study enrolment, had completed medical and demographic records, and (for the cARD/case cohort) were hospitalized for serious cARDs with confirmed antibiotic prescriptions in the 60 days before admission.

Each case (cARD hospitalization) was age- and sex-matched with up to four controls (patients who received antibiotic interventions but did not develop cARDs). Patients undergoing more than one simultaneous antibiotic course were excluded from the study to prevent cross-reactivity-associated complications. The outcome under focus was a serious antibiotic-associated cARD requiring hospitalization, wherein the class of prescribed antibiotics comprised the exposure.

Previous research has established the low risk of macrolide-based antibiotics in triggering cARDs, and they were hence used as a reference group for comparisons between antibiotic classes. Statistical analyses included multivariable logistic regressions. All models were adjusted for age, sex, demographics, socioeconomics, and presence of chronic diseases.

Study findings

Over the 20-year study period, 3,257,181 participants (median age = 75), collectively representing 21,758 cARD-associated hospital admissions, were identified and included in the analysis dataset. These individuals were age, sex, and demographics matched to 87,025 controls. The average latency between antibiotic prescription and hospitalization was revealed to be 14 days.

Penicillin was observed to be the most commonly prescribed antibiotic across both cohorts (n = 42,323; 28.9%). Cephalosporins, fluoroquinolones (16.5%), macrolides (14.8%), nitrofurantoin (8.6%), and sulfonamides (6.2%) followed. Antibiotic prescriptions of classes excluding those already mentioned were collectively classified as ‘others’ and comprised 6.9% of prescriptions.

Cumulatively, 34,114,254 antibiotics courses were associated with 72,449 serious cARDs (2.12 cARDs per 1,000 prescriptions). Following corrections, multivariable analyses revealed sulfonamides as the drugs most associated with cARD risk (OR = 2.9; 3.22 cARD per 1,000 prescriptions), followed by cephalosporins (OR = 2.6; 4.92), ‘others’ (OR = 2.3), nitrofurantoin (OR = 2.2), penicillins (OR = 1.4), and fluoroquinolones (OR = 1.3).

“During the study period, 2852 (13.1%) of the 21 758 case patients were hospitalized for a cADR within 60 days of outpatient antibiotics; of these, only 50 (1.8%)were explicitly identified as having SJS/TEN. In-hospital mortality was low for all serious cADRs (n = 150; 5.3%) but higher for the SJS/TEN subtype (n = 10; 20.0%)”

Conclusions

The present study verifies the association between antibiotic consumption and subsequent cARD risk. Notably, it highlights the differential risk contributions of different antibiotic classes. Sulfonamides and cephalosporins were identified as the most harmful antibiotics, though all classes were observed to increase cARD-associated hospitalizations to varying extents.

This study emphasizes the need for medical practitioners to take note of the potential cARD contributions of different antibiotics and preferentially select the lowest-risk antibiotic class wherever clinically possible.