ABSTRACT
Background: Stevens-Johnson syndrome (SJS) and Toxic Epidermal Necrolysis (TEN) are life-threatening dermatological emergencies. SCORTEN (SCORe of toxic epidermal necrolysis) is a validated score to predict mortality; however, there is a paucity of data to determine its usefulness in the Indian population. Objective: To evaluate the accuracy of SCORTEN as a prognostic marker in SJS-TEN. Methods: A prospective observational study was conducted at a tertiary care hospital for two years. SCORTEN was calculated on days one and three of admission. The actual death rates were compared to the predicted rates as estimated by the SCORTEN by standardised mortality ratio analysis (SMR). Results: Of 40 cases included in the study, the mean age was 36.2 ± 14 years (range 11-65) with the male: female ratio being 1.67:1. Antibiotics (37.5%) were the most common group followed by anticonvulsants (22.5%). Comorbidities were observed in 60% of cases, with epilepsy (17.5%) and HIV (human immunodeficiency virus) infection (12.5%) being common. On univariate analysis, heart rate > 120/min, epidermal detachment > 10% BSA, and Se HCO3 (bicarbonate) <20 mmol/L were associated significantly with the death of the subjects (P < 0.05). The observed mortalities were 4.34%, 0, 0 and 80% for SCORTEN 0-1 (3.2%), 2 (12.1%), 3 (35.8%) and 4 (58.3%) respectively when compared to expected mortality. SMR of SJS was 0.69 and of TEN was 1.49. Conclusion: SCORTEN gave an overestimation of mortality in patients with lower scores and an underestimation of mortality in patients with higher scores in our study. Minor refinements based on the study population may increase the predictive accuracy of the original scale.
