OBJECTIVE: To determine the independent contribution of prognostic factors to length of hospital stay of minor and moderate burn victims at the Hospital Universitario San José (HUSJ), Popayán, Colombia, 2000-2010. METHODS: This was a retrospective cohort study of minor and moderate burn victims admitted between 2000 and 2010, at the burn unit (HUSJ). This is a further analysis of a same cohort previously published in Burns. The following variables were recorded and analyzed: age, gender, origin, depth and extent of burn, causal agent, length of hospital stay and mortality. The main outcome under study was length of stay. Survival analysis was done to explore the association of covariates and length of hospital stay and Cox regression model to adjust the effect of covariates in the outcome. RESULTS: During the study period 2000-2010, 842 of 921 (91.5%) patients treated at the Burn Unit of HUSJ that had complete data were included. There were 520 (61.8%) males and 322 (38.2%) females with a male to female ratio of 1.6:1. Their median age was 9 years (IQR 3-28). The median of percent total body surface area burned (TBSA) was 12% (IQR 7-21) and the most common degree of burn was 2nd degree with 58% (488 patients). There were 12 deaths (censored data) and 830 patients were discharged alive. After multivariate adjustment, significant associations with length of hospital stay remained for age group, burn degree and extension of the burn. The strongest relationship found was for burn degree (2nd degree superficial vs. 3rd degree hazard ratio=2.66 CI 95% [2.13-3.33]). CONCLUSIONS: In patients admitted with mild and moderate burns at HUSJ, the main predictors of length of stay were age, burn degree and extension of the burn.
OBJECTIVE: To determine the independent contribution of prognostic factors to length of hospital stay of minor and moderate burn victims at the Hospital Universitario San José (HUSJ), Popayán, Colombia, 2000-2010. METHODS: This was a retrospective cohort study of minor and moderate burn victims admitted between 2000 and 2010, at the burn unit (HUSJ). This is a further analysis of a same cohort previously published in Burns. The following variables were recorded and analyzed: age, gender, origin, depth and extent of burn, causal agent, length of hospital stay and mortality. The main outcome under study was length of stay. Survival analysis was done to explore the association of covariates and length of hospital stay and Cox regression model to adjust the effect of covariates in the outcome. RESULTS: During the study period 2000-2010, 842 of 921 (91.5%) patients treated at the Burn Unit of HUSJ that had complete data were included. There were 520 (61.8%) males and 322 (38.2%) females with a male to female ratio of 1.6:1. Their median age was 9 years (IQR 3-28). The median of percent total body surface area burned (TBSA) was 12% (IQR 7-21) and the most common degree of burn was 2nd degree with 58% (488 patients). There were 12 deaths (censored data) and 830 patients were discharged alive. After multivariate adjustment, significant associations with length of hospital stay remained for age group, burn degree and extension of the burn. The strongest relationship found was for burn degree (2nd degree superficial vs. 3rd degree hazard ratio=2.66 CI 95% [2.13-3.33]). CONCLUSIONS: In patients admitted with mild and moderate burns at HUSJ, the main predictors of length of stay were age, burn degree and extension of the burn.