Helen E Douglas1, Andrew Ratcliffe2, Rajdeep Sandhu2, Umair Anwar2. 1. Pinderfields Hospital, Aberford Road, Wakefield WF1 4DG, England, United Kingdom. Electronic address: helendouglas@doctors.org.uk. 2. Pinderfields Hospital, Aberford Road, Wakefield WF1 4DG, England, United Kingdom.
Abstract
BACKGROUND: Many different burns mortality prediction models exist; however most agree that important factors that can be weighted include the age of the patient, the total percentage of body surface area burned and the presence or absence of smoke inhalation. METHODS: A retrospective review of all burns primarily admitted to Pinderfields Burns ICU under joint care of burns surgeons and intensivists for the past 3 years was completed. Predicted mortality was calculated using the revised Baux score (2010), the Belgian Outcome in Burn Injury score (2009) and the Boston group score by Ryan et al. (1998). Additionally 28 of the 48 patients had APACHE II scores recorded on admission and the predicted and actual mortality of this group were compared. RESULTS: The Belgian score had the highest sensitivity and negative predictive value (72%/85%); followed by the Boston score (66%/78%) and then the revised Baux score (53%/70%). APACHE II scores had higher sensitivity (81%) and NPV (92%) than any of the burns scores. DISCUSSION: In our group of burns ICU patients the Belgian model was the most sensitive and specific predictor of mortality. In our subgroup of patients with APACHE II data, this score more accurately predicted survival and mortality.
BACKGROUND: Many different burns mortality prediction models exist; however most agree that important factors that can be weighted include the age of the patient, the total percentage of body surface area burned and the presence or absence of smoke inhalation. METHODS: A retrospective review of all burns primarily admitted to Pinderfields Burns ICU under joint care of burns surgeons and intensivists for the past 3 years was completed. Predicted mortality was calculated using the revised Baux score (2010), the Belgian Outcome in Burn Injury score (2009) and the Boston group score by Ryan et al. (1998). Additionally 28 of the 48 patients had APACHE II scores recorded on admission and the predicted and actual mortality of this group were compared. RESULTS: The Belgian score had the highest sensitivity and negative predictive value (72%/85%); followed by the Boston score (66%/78%) and then the revised Baux score (53%/70%). APACHE II scores had higher sensitivity (81%) and NPV (92%) than any of the burns scores. DISCUSSION: In our group of burns ICUpatients the Belgian model was the most sensitive and specific predictor of mortality. In our subgroup of patients with APACHE II data, this score more accurately predicted survival and mortality.