A model for predicting mortality among critically ill burn victims
Introduction
The outcome of thermal trauma is closely associated with variables such as presence of inhalation injury, body surface area (BSA) burned and age [1]. However, inhalation injury is difficult to quantify [2] and, although generally accepted as a prognostic factor, some studies have failed to show a relationship between inhalation injury and outcome [3], [4]. Also, the relative prognostic significance of full-thickness BSA (FTBSA) versus total (TBSA) BSA burned remains unsettled. Finally, regression models generally used to study the association between different variables and mortality prognosis often assume a linear relationship, but an analysis that took into consideration a more complex relationship might be more appropriate.
In this study we analysed the relationship between demographic variables and outcome among people admitted to an intensive care burn unit (ICBU). Our objectives were, first, to clarify the relationship between the requirement of mechanical ventilation after trauma and mortality. We observed that some individuals might have mild inhalation injury and yet do well without mechanical ventilation, whereas others might require mechanical ventilation early after admission without having had inhalation injury. Therefore, we hypothesised that the early requirement of mechanical ventilation could be more strongly associated with mortality than inhalation injury itself. Second, we aimed to compare the relative impact of TBSA versus FTBSA burned. Our third aim was to elucidate the non-linear relationship between mortality and other variables such as age and BSA burned.
Section snippets
Participants
Data were obtained from the Hospital Universitario de Getafe Burn Registry, and every person admitted to our ICBU from April 1992 to December 2005 was prospectively included in our database. People are admitted to our ICBU with ≥20% BSA burned, inhalation injury, or less extensive burns depending on age and comorbidities. We excluded from the analysis burns due to non-thermal injury (such as chemical or electrical burns), associated non-burn trauma, admission based on non-burn-related diagnoses
Unadjusted and adjusted mortality risk factors
In this study population mean age was 46.2 ± 20.6 years, and 70.4% were men. Median TBSA and FTBSA were 20% and 9%, respectively. Mean mortality was 17.6% with a 95% confidence interval (CI) of 15.1–20.4%; mean mortality decreased from 23.8% in 1992–1995 to 12.3% in 2001–2005. The LA50% was 65%.
The derivation and validation sets were not significantly different in any of the analysed variables, except that in the derivation set a higher proportion of participants were admitted early after injury (
Effect of age
Ageing is associated with decreased ability of the respiratory and cardiovascular systems to cope with demands imposed by injury. Thus, a relationship between age and mortality after burn is anticipated and has been reported by many studies [10], [11], [12], [13]. Moreau et al. [14] developed an age-risk score after burn, and showed that age is not linearly related to mortality if the group includes children. Our finding of a linear relationship between age and mortality is probably explained
Conclusions
This study confirms and expands previous knowledge on the relationship between injury-associated factors and mortality among burn victims. Our predictive model performs in our validation set comparably or even better than other models previously published. We describe the relative contribution to mortality of TBSA and FTBSA burned, as well as the non-linear relationship between BSA burned and outcome. In line with recent studies, we did not find an independent relationship between inhalation
Conflicts of Interest
José A. Lorente-Balanza, on behalf of all the authors of the article entitled “Mortality prediction in critically ill burn patients”, states that there is not any conflict of interest to declare.
Acknowledgement
This research was partially supported by grants FIS 07/1275 and FIS 06/1664.
References (50)
- et al.
Outcome measures in burn care. Is mortality dead?
Burns
(2004) - et al.
Factors involved in burn mortality: a multivariate statistical approach based on discriminant analysis
Burns
(1992) - et al.
Evaluation predictive factors of burned patients
Ann Chir Plast Esthet
(2001) - et al.
An evaluation of risk factors for mortality after burn trauma and the identification of gender-dependent differences in outcomes
J Am Coll Surg
(2001) - et al.
Burns in the elderly
Burns
(1989) - et al.
Etiology and consequences of respiratory failure in thermally injured patients
Am J Surg
(1993) - et al.
Prognostic implications of inhalation injury in burn patients in Tokyo
Burns
(2005) - et al.
The impact of risk factors and pre-existing conditions on the mortality of burn patients and the precision of predictive admission-scoring systems
Burns
(1997) - et al.
Burns mortality and hospitalization time. A prospective statistical study of 352 patients in an Asian National Burn Centre
Burns
(1995) - et al.
Epidemiological and outcome characteristics of major burns in Tokio
Burns
(2005)