Predictors of in-hospital mortality, sepsis and prolonged length of stay in burn patients: a ten-year monocentric experience and comparison with the Baux and revised Baux scores
Year: 2026
Type: article
Study design: Observational Study
Abstract: Burn mortality is usually predicted with the Baux and revised Baux scores, which do not include comorbidity; the independent role of psychiatric disease is also poorly defined. We reviewed 793 consecutive burn patients (630 adults, 163 children) admitted to a regional referral centre between 2016 and 2025. In adults, logistic models for in-hospital mortality, sepsis, and prolonged length of stay (at least 30 days) used age group, burn size by total body surface area, inhalation injury, and the Charlson Comorbidity Index as covariates, with internal bootstrap validation. Adult in-hospital mortality was 7.8% (49 of 630); no child died. Independent predictors of mortality were age 85 years or older (odds ratio 18.9), burn size 40% or more (odds ratio 40.6), inhalation injury (odds ratio 4.29), and a comorbidity index of 2 or more (odds ratio 14.2). Psychiatric disorders, not captured by the comorbidity index, were also independently associated with mortality (odds ratio 2.83). The multivariable model, a Baux-based model, and a revised Baux-based model (both including the comorbidity index) had almost identical discrimination (optimism-corrected 0.941, 0.955, 0.957; pairwise p>0.05). For sepsis and prolonged stay, burn size and age dominated, while comorbidity lost significance; the crude association between sepsis and mortality disappeared after adjustment, whereas sepsis remained strongly associated with prolonged stay (adjusted odds ratio 10.8). A multivariable model discriminates as well as the Baux scores while allowing comorbidity and psychiatric history to be quantified.
Source: Journal of Burn Care & Research
DOI: 10.1093/jbcr/irag187
Institution: University of Padua
Language: English
Subfield: Surgery
Field: Medicine
Domain: Health Sciences
Open Access status: hybrid
Estimated APC paid: $3,698
OpenAlex ID: W7227905016
