Background. In-hospital mortality of major burn patients remains high in severe cases and varies across centers. Burn extent (TBSA), age, and inhalation injury are well-established prognostic determinants, but the independent contribution of pre-existing comorbidity burden and secondary outcomes remain underrepresented in updated single-center analyses. Aims. To identify predictors of in-hospital mortality in the adult cohort and to compare the discriminative ability of alternative prognostic models (primary aim); to analyze predictors of in-hospital sepsis and length of stay (LOS) and to describe the center's treatment patterns and procedural timing (secondary aims). Materials and methods. Single-center retrospective analysis of patients admitted to the Burn Center of Padua University Hospital between 2016 and 2025, stratified into adults (≥18 years) and pediatric patients (<18 years); prognostic analyses were conducted in adults. Primary outcome: in-hospital mortality; secondary outcomes: in-hospital sepsis and LOS. Multivariable logistic regression models were estimated, adjusted for age, TBSA class, inhalation injury, and Charlson Comorbidity Index (CCI). For mortality, the primary model was compared with the Baux Score and the Revised Baux Score using AUROC, Brier Score, DeLong's test, and internal bootstrap validation. Results. A total of 793 patients were included (630 adults, 163 pediatric). Among adults, mortality was 7.8% (49/630) and sepsis 13.0% (82/630); no deaths occurred in the pediatric cohort. - Mortality. Independent predictors: age ≥85 years (OR 18.9), TBSA 20–39% (OR 6.45) and ≥40% (OR 40.6), inhalation injury (OR 4.29), and comorbidity burden (CCI=1: OR 6.66; CCI≥2: OR 14.2). The three models showed excellent and overlapping discrimination (AUROC 0.953 primary, 0.959 Baux, 0.960 Revised Baux; DeLong's test non-significant): the model with separate predictors was not inferior to the composite scores. Psychiatric disorders were confirmed as an independent predictor (OR 2.83), whereas female sex was not. - Sepsis. Independently predicted by age and burn extent, whereas comorbidity burden lost significance after adjustment. The crude association between sepsis and mortality (OR 2.37) disappeared after adjustment for baseline severity (OR 0.40; p=0.065). - LOS. Independently predicted by age, burn extent, and inhalation injury, whereas comorbidity burden lost significance after adjustment. Sepsis remained an independent predictor of prolonged stay even after adjustment (OR 10.8). - Descriptive analyses. Associations between individual treatments and outcomes were dominated by confounding by indication; NexoBrid was associated with a shorter time to first skin coverage. Conclusions. TBSA, age, inhalation injury, and comorbidity burden (CCI) were independent predictors of in-hospital mortality, with CCI being a robust determinant that is additive to the classic ones; a model with separate predictors matches the discriminative performance of the Baux and Revised Baux scores. Sepsis is predicted by burn severity and appears associated with mortality predominantly through baseline severity itself, but it is an independent determinant of prolonged stay. These findings support the integration of comorbidity burden into the prognostic stratification of burn patients.
Premesse. La mortalità intraospedaliera del grande ustionato resta elevata nelle forme gravi e variabile tra i centri. Estensione dell'ustione (TBSA), età e lesione da inalazione sono determinanti prognostici consolidati, ma il contributo indipendente del carico comorboso preesistente e gli esiti secondari restano sottorappresentati nelle analisi monocentriche aggiornate. Obiettivi. Identificare i predittori di mortalità intraospedaliera nella coorte adulta e confrontare la capacità discriminativa di modelli prognostici alternativi (obiettivo primario); analizzare i predittori di sepsi intraospedaliera e di durata della degenza (LOS) e descrivere i pattern terapeutici e il timing procedurale del centro (obiettivi secondari). Materiali e metodi. Analisi retrospettiva monocentrica dei pazienti ricoverati presso il Centro Ustioni dell'Azienda Ospedale Università di Padova nel periodo 2016–2025, stratificati in adulti (≥18 anni) e pediatrici (<18 anni); le analisi prognostiche sono state condotte negli adulti. Outcome primario: mortalità intraospedaliera; secondari: sepsi intraospedaliera e LOS. Sono stati stimati modelli di regressione logistica multivariabile aggiustati per età, classe di TBSA, lesione da inalazione e Charlson Comorbidity Index (CCI). Per la mortalità, il modello primario è stato confrontato con Baux Score e Revised Baux Score mediante AUROC, Brier Score, test di DeLong e validazione interna bootstrap. Risultati. Sono stati inclusi 793 pazienti (630 adulti, 163 pediatrici). Negli adulti la mortalità è stata del 7,8% (49/630) e la sepsi del 13,0% (82/630); nessun decesso nella coorte pediatrica. - Mortalità. Predittori indipendenti: età ≥85 anni (OR 18,9), TBSA 20–39% (OR 6,45) e ≥40% (OR 40,6), lesione da inalazione (OR 4,29) e carico comorboso (CCI=1: OR 6,66; CCI≥2: OR 14,2). I tre modelli mostravano discriminazione eccellente e sovrapponibile (AUROC 0,953 primario, 0,959 Baux, 0,960 Revised Baux; DeLong non significativo): il modello con predittori separati non risultava inferiore agli score compositi. I disturbi psichiatrici si confermavano predittore indipendente (OR 2,83), il sesso femminile no. - Sepsi. Predetta in modo indipendente da età ed estensione dell'ustione, mentre il carico comorboso perdeva significatività dopo aggiustamento. L'associazione grezza tra sepsi e mortalità (OR 2,37) si annullava dopo aggiustamento per la gravità basale (OR 0,40; p=0,065). - LOS. Predetta in modo indipendente da età, estensione dell'ustione e lesione da inalazione, mentre il carico comorboso perdeva significatività dopo aggiustamento. La sepsi rimaneva predittore indipendente di degenza prolungata anche dopo aggiustamento (OR 10,8). - Analisi descrittive. Le associazioni tra singoli trattamenti ed esiti erano dominate da confondimento per indicazione; NexoBrid si associava a un tempo più breve alla prima copertura cutanea. Conclusioni. TBSA, età, lesione da inalazione e carico comorboso (CCI) sono risultati predittori indipendenti di mortalità intraospedaliera, con il CCI robusto e additivo rispetto ai determinanti classici; un modello con predittori separati eguaglia la performance discriminativa di Baux e Revised Baux. La sepsi è predetta dalla gravità dell'ustione e appare associata alla mortalità prevalentemente attraverso la gravità basale, ma è un determinante indipendente di degenza prolungata. I risultati supportano l'integrazione del carico comorboso nella stratificazione prognostica del paziente ustionato.
Mortalità intraospedaliera e decorso clinico nel paziente grande ustionato: analisi retrospettiva di una coorte monocentrica (2016-2025)
CARLON, SAMUEL
2025/2026
Abstract
Background. In-hospital mortality of major burn patients remains high in severe cases and varies across centers. Burn extent (TBSA), age, and inhalation injury are well-established prognostic determinants, but the independent contribution of pre-existing comorbidity burden and secondary outcomes remain underrepresented in updated single-center analyses. Aims. To identify predictors of in-hospital mortality in the adult cohort and to compare the discriminative ability of alternative prognostic models (primary aim); to analyze predictors of in-hospital sepsis and length of stay (LOS) and to describe the center's treatment patterns and procedural timing (secondary aims). Materials and methods. Single-center retrospective analysis of patients admitted to the Burn Center of Padua University Hospital between 2016 and 2025, stratified into adults (≥18 years) and pediatric patients (<18 years); prognostic analyses were conducted in adults. Primary outcome: in-hospital mortality; secondary outcomes: in-hospital sepsis and LOS. Multivariable logistic regression models were estimated, adjusted for age, TBSA class, inhalation injury, and Charlson Comorbidity Index (CCI). For mortality, the primary model was compared with the Baux Score and the Revised Baux Score using AUROC, Brier Score, DeLong's test, and internal bootstrap validation. Results. A total of 793 patients were included (630 adults, 163 pediatric). Among adults, mortality was 7.8% (49/630) and sepsis 13.0% (82/630); no deaths occurred in the pediatric cohort. - Mortality. Independent predictors: age ≥85 years (OR 18.9), TBSA 20–39% (OR 6.45) and ≥40% (OR 40.6), inhalation injury (OR 4.29), and comorbidity burden (CCI=1: OR 6.66; CCI≥2: OR 14.2). The three models showed excellent and overlapping discrimination (AUROC 0.953 primary, 0.959 Baux, 0.960 Revised Baux; DeLong's test non-significant): the model with separate predictors was not inferior to the composite scores. Psychiatric disorders were confirmed as an independent predictor (OR 2.83), whereas female sex was not. - Sepsis. Independently predicted by age and burn extent, whereas comorbidity burden lost significance after adjustment. The crude association between sepsis and mortality (OR 2.37) disappeared after adjustment for baseline severity (OR 0.40; p=0.065). - LOS. Independently predicted by age, burn extent, and inhalation injury, whereas comorbidity burden lost significance after adjustment. Sepsis remained an independent predictor of prolonged stay even after adjustment (OR 10.8). - Descriptive analyses. Associations between individual treatments and outcomes were dominated by confounding by indication; NexoBrid was associated with a shorter time to first skin coverage. Conclusions. TBSA, age, inhalation injury, and comorbidity burden (CCI) were independent predictors of in-hospital mortality, with CCI being a robust determinant that is additive to the classic ones; a model with separate predictors matches the discriminative performance of the Baux and Revised Baux scores. Sepsis is predicted by burn severity and appears associated with mortality predominantly through baseline severity itself, but it is an independent determinant of prolonged stay. These findings support the integration of comorbidity burden into the prognostic stratification of burn patients.| File | Dimensione | Formato | |
|---|---|---|---|
|
Carlon_Samuel.pdf
Accesso riservato
Dimensione
3.37 MB
Formato
Adobe PDF
|
3.37 MB | Adobe PDF |
The text of this website © Università degli studi di Padova. Full Text are published under a non-exclusive license. Metadata are under a CC0 License
https://hdl.handle.net/20.500.12608/109069