Clinicoetiological Study of Stevens-Johnson Syndrome and Toxic Epidermal Necrolysis Spectrum and the Correlation of SCORTEN with Prognosis

Abstract

Background: Stevens-Johnson syndrome (SJS) and Toxic Epidermal Necrolysis (TEN) are severe cutaneous adverse reactions of major concern because of its high mortality. The prognosis of SJS and TEN is widely assessed with SCORTEN (SCORe of TEN). Although, it is a largely useful scale, the predictive ability is still variable. Aims and Objectives: This study was conducted to assess the clinicoetiological profile and outcome of SJS and TEN and to evaluate the validity of SCORTEN in assessing the prognosis in South Indian population. Methods: This prospective observational study was conducted in the Department of Dermatology, Venereology and Leprosy in a Tertiary care hospital from January 2016 to June 2017. Detailed history, examination findings, treatment and SCORTEN scores were recorded. SCORTEN's accuracy in predicting the mortality was assessed on day 1, 3 and 5 of admission. Results: The incidence of SJS/TEN among other drug reactions was 29.5%. The most common age group affected was 30-49years (41.1%), with male preponderance (76.5%). The age range of patients was 6 and 67 years. TEN (64.7%) was the predominant spectrum followed by SJS and SJS-TEN overlap in 17.6% each. Anticonvulsants (47%) were the commonest causative drug, followed by analgesics (35%) and antibiotics (11%). The validity of SCORTEN was the same on days 1, 3, and 5. There was good agreement between the actual and predicted mortality on all three days. A mortality of 17.6% (3 cases) was recorded in this study. Three patients (17.6%) died in our study. All survivors had a score of 4 or less. The predicted mortalities were 0.417, 1.836, and 2.574 and the observed mortalities were 0, 2, and 1 in SJS, SJS-TEN overlap, and TEN respectively. Analysis of SCORTEN on a single day, either day 1, 3, or 5 was found to be as useful as the serial analysis. Conclusion: SCORTEN gave a significant estimation of mortality in SJS-TEN overlap patients, whereas it overestimated mortality in TEN patients. An increase in individual scores for the elevation of blood urea nitrogen (BUN) in existing SCORTEN and the inclusion of new parameters like raised liver enzymes, thrombocytopenia, and pulmonary infiltrates aided in proposing a modified SCORTEN for the South Indian population. Further studies on a larger scale, are needed to validate the modified SCORTEN proposed by us.