| dc.description.abstract | BACKGROUND: Cutaneous adverse reac<ons range from mild manifesta<ons to severe
condi<ons such as Toxic Epidermal Necrolysis (TEN), with drugs causing up to 94% of the cases.
The es<mated mortality rate for TEN is approximately 30%, with sepsis associated with
mul<organ failure or respiratory compromise being the main causes of death. Despite this, the
SCORTEN scale, validated for predic<ng mortality in these pa<ents, does not incorporate organ
failure as a parameter of its evalua<on. The hypothesis stated that the presence of organ
dysfunc<on is associated with mortality and unfavorable outcomes in pa<ents with toxic
epidermal necrolysis.
MATERIAL AND METHODS: An observa<onal, retrospec<ve study was conducted in pa<ents
over 18 years diagnosed with TEN and treated in Hospital Universitario de Getafe. Clinical data
was collected through the clinical history in HCIS and/or requested to the archive of the Hospital.
RESULTS: Of the 84 pa<ents included, 16 died (19%), of which 2 were early deaths, and the
remaining 14 were late deaths. The median age of the pa<ents was 50 (34.25-65). The median
SCORTEN was 3, and the median affected skin surface (%) was 75% [Interquar<le Range (IQR)
50-90] with a mean length of stay of 13.85 days [95% Confidence Interval (CI) 6.74-26.49]
hospitalized in the Burn ICU. Significant differences were iden<fied between the age, skin surface
area affected, APACHE II, SAPS II, SCORTEN II, and SOFA from days 0 to 4 of the survivors and
those who exhibited late mortality; in addi<on to a significantly higher presence of
cardiovascular, respiratory, and renal failure in the laeer one. A higher incidence of infec<ons
was also observed in pa<ents with organ failure. The ra<o (fluid input/diuresis) on day +2 in the
ICU and arterial pH on day +1 were significantly different between pa<ents who survived and
those who experienced late mortality, showing the highest Area Under the Receiver Opera<ng
Characteris<c (ROC) Curve (AURC) among all variables analyzed. Addi<onally, arterial pH on day
1 greater than 7.353 was iden<fied as the variable with the highest sensi<vity and specificity for
discrimina<ng between survivors and pa<ents with late mortality, supported by an Odds Ra<o
(OR) of 9.646 (95% CI 2.578-34.746).
CONCLUSION: Organ failure frequently complicates the course of TEN, exhibi<ng an
independent correla<on with mortality measurable by the SOFA scale. Mortality predic<on
improves when considering variables like age, affected body surface area, and variables related
to resuscita<on (ra<o fluid input/diuresis and arterial pH during resuscita<on). Adjusted
predic<ve models, such as those considering age, affected skin surface area, APACHE II, and
13
arterial pH on day 1, have been found to discriminate mortality more effec<vely than the
SCORTEN method. | es |
