Stevens-Johnson Syndrome (SJS) and Toxic Epidermal Necrolysis (TEN) exist on a spectrum of autoimmune conditions which cause epidermal detachment and keratinocyte necrosis. Due to the rare incidence of these conditions, a dramatic heterogeneity in treatment algorithms exists. To better appreciate pharmacologic immunosuppressive therapies\' impact on survival, the authors queried a multi-institutional data network. Data for this study was extracted from TriNetX Research Network, a platform that contains ICD-9/ICD-10 coding data from a consortium of international healthcare organizations. Seventy-one institutions were queried to identify adult patients diagnosed with SJS, TEN or SJS-TEN Overlap. Cohorts were created based on the therapy received: systemic steroids (SS), diphenhydramine (DH), cyclosporine (CS), intravenous immunoglobulin (IVIG), tumor necrosis factor alpha inhibitors (TNFαi), or a combination of treatments. Cohorts were then propensity matched with patients who received supportive care. Patients who only received one of the above treatments showed no significant reduction in 90-day mortality. Patients who received CS or IVIG as part of their multitherapy showed a significantly increased risk of death when compared to supportive care (CS: RR = 1.583, 95% CI [1.119, 2.240]; IVIG: RR = 2.132, 95% CI [1.485, 3.059]). Despite their frequent utilization, this study\'s analysis suggests that none of these therapies confer significant 90-day mortality survival over supportive care alone. These results highlight the heterogeneity of therapies and emphasize the need for critical prospective appraisal of their outcomes in SJS and TEN.

BACKGROUND: Epidermal necrolysis (EN) involving ≥10% of the body surface area (BSA) is often complicated by bacterial infections.
OBJECTIVE: We sought to describe the epidemiology of bloodstream infections (BSIs) in EN involving a BSA ≥10% and the diagnostic performances of skin cultures for predicting the pathogen(s) isolated from BSIs.
METHODS: This retrospective single-center observational study was conducted between 2009 and 2017. All patients referred at the acute phase for EN involving a BSA ≥10% were included. All clinical and bacteriologically relevant data were collected (blood and skin cultures results, number, and severity and time of BSI). Sensitivity, specificity, and predictive values of skin cultures and impact of the bacterial inoculum were investigated.
RESULTS: Of 98 patients, 46 (46.9%) had ≥1 BSI episode during the hospital stay (BSIs were caused by Staphylococcus aureus [n = 17, 36.9%] and Pseudomonas aeruginosa [n = 17, 36.9%]). Skin cultures were concordant with blood cultures in 32 cases (71.1%). The positive and negative predictive values were 57.7% and 89.4% for S aureus and 50.0% and 80.9% for P aeruginosa, respectively. BSI increased with cutaneous inoculum of S aureus.
CONCLUSIONS: This was a retrospective single-center design with a low total number of BSIs.
CONCLUSIONS: Skin cultures for S aureus and P aeruginosa may help predict the pathogens involved in BSIs.