Introduction: Drug reaction with eosinophilia and systemic symptoms (DRESS) is a rare severe cutaneous adverse reaction (AR), which can be life threatening J Am Acad Dermatol. 2013;68(5):693.e1–14. The reporting of DRESS in Australian electronic medical records (EMRs) has not been studied. Methods: A dual ascertainment strategy was employed in this multi-centre, retrospective study. Part 1 involved data collection from a 2.5-year cohort of consecutive inpatient admissions. Patients with an AR label of DRESS, eosinophilia or severe rash were identified, along with the causative agents. Part 2 involved evaluating the AR label documentation for patients from an independently derived list of confirmed DRESS cases from an immunology department register. Case note review was undertaken for all possible cases of DRESS that were identified, and RegiSCAR scores were calculated. Results: Of the 135,080 inpatients from Part 1, there were 17 patients (prevalence 12.59 per 100,000) with at least possible DRESS (RegiSCAR >2). The prevalence of patients with a RegiSCAR score consistent with probable or definite DRESS was 6.66 per 100,000 individuals. Among the 135,080 patients, 12 individuals had a DRESS AR label on the EMR (prevalence 8.88 per 100,000). In Part 2 of the study, 16 confirmed DRESS cases were identified from an Immunology department register over a period of 8 years. The mean age of this group was 49.3 at diagnosis (SD 21.58) and 50% were female (n = 8). Of the 16 cases, 14 (87.5%) were correctly labelled as “DRESS syndrome” in the EMR. In total, there were 31 patients with at least possible DRESS identified, with two cases (6.45%) of false negative AR documentation, and one case (3.23%) of false positive AR documentation. The most common drug culprits were vancomycin (n = 11, 29.7%), penicillin-based antibiotics (n = 9, 24.3%), and carbamazepine (n = 3, 8.1%). Conclusion: The prevalence of EMR-documented DRESS syndrome in South Australia is higher than seen in other studies. Most DRESS was caused by antibiotics. The majority of patients were documented correctly in the EMR as far as can be determined.
Jiang et al. (Mon,) studied this question.