Toxic epidermal necrolysis data from the CELESTE multinational registry. Part II: Specific systemic and local risk factors for the development of infectious complications

The aim of the study was to identify the most important systemic and local risk factors for the development of infectious complications in patients with toxic epidermal necrolysis (TEN).

Material and methodology

This is a multicentric study that included all patients with TEN who were hospitalized between 2000–2015 in specialized centres in the Czech Republic and Slovakia. The total catchment area included a population of over 12.5 million inhabitants.The actual implementation of the project was carried out using data obtained from the CELESTE (Central European LyEll Syndrome: Therapeutic Evaluation) registry, wherein specific parameters related to epidemiological indicators and infectious complications in patients with TEN were evaluated as a retrospective analysis.

Results

A total of 38 patients (97%) of the group were treated with corticosteroids. The comparison of patients with different doses of corticosteroids did not exhibit a statistically significant effect of corticosteroid administration on the development of infectious complications (p = 0.421).There was no effect of the extent of the exfoliated area on the development of infectious complications in this area. The average extent of the exfoliated area was 66% TBSA (total body surface area) in patients with reported infectious complications and 71% TBSA (p = 0.675) in patients without infectious complications.In the case of the development of an infectious complication in the bloodstream (BSI), the increasing effect of the SCORTEN (SCORe of Toxic Epidermal Necrosis) value was monitored during hospitalization. Within 5 days from the beginning of the hospitalization, the average SCORTEN value was 2.7 in 6 patients with BSI and 3.0 in 32 patients without BSI (p = 0.588). In the period after the 15th day of hospitalization, 7 patients with BSI had an average SCORTEN value of 3.4, and 16 patients without BSI had an average SCORTEN value of 2.5 (p = 0.079).In the case of low respiratory tract infection (LRTI), the effects of the necessity for artificial pulmonary ventilation and the presence of tracheostomy were monitored. The statistically significant effect of mechanical ventilation on the development of LRTI occurred only during the period of 11–15 days from the beginning of the hospitalization (p = 0.016). The effect of the tracheostomy on the development of LRTI was proven to be more significant.

Conclusion

We did not find any statistically significant correlation between the nature of immunosuppressive therapy and the risk of developing infectious complications. We failed to identify statistically significant risk factors for the development of BSI. Mechanical ventilation and tracheostomy increase the likelihood of developing LRTIs in patients with TEN.