BACKGROUND: This study aimed to develop an objective model for predicting mortality after burn injury in Taiwan. METHODS: From 1997 to 2010, 23,147 patients with acute burn injury in 44 hospitals were retrospectively reviewed. Variables examined were age, sex, depth and extent of burn, inhalation injury, flushing time, hospital admission and referral status, intensive care unit admission, and mortality. Logistic regression analyses were used to evaluate risk factors. Model performance and calibration was evaluated by measures of discrimination and goodness-of-fit statistic, respectively. A nomogram of four major risk factors was used to calculate the probability of mortality. RESULTS: Only 22,665 patients (mean [SD] age, 31.05 [22.67] years; mean second-degree and third-degree burn sizes, 8.67% [10.64%] and 3.25% [10.91%], respectively) survived until discharge, for a mortality rate of 2.08%. CONCLUSION: Burn depth is an important predictive factor for mortality. An objective model can help estimate the probability of death in acute burn injury. LEVEL OF EVIDENCE: Prognostic study, level II.
BACKGROUND: This study aimed to develop an objective model for predicting mortality after burn injury in Taiwan. METHODS: From 1997 to 2010, 23,147 patients with acute burn injury in 44 hospitals were retrospectively reviewed. Variables examined were age, sex, depth and extent of burn, inhalation injury, flushing time, hospital admission and referral status, intensive care unit admission, and mortality. Logistic regression analyses were used to evaluate risk factors. Model performance and calibration was evaluated by measures of discrimination and goodness-of-fit statistic, respectively. A nomogram of four major risk factors was used to calculate the probability of mortality. RESULTS: Only 22,665 patients (mean [SD] age, 31.05 [22.67] years; mean second-degree and third-degree burn sizes, 8.67% [10.64%] and 3.25% [10.91%], respectively) survived until discharge, for a mortality rate of 2.08%. CONCLUSION: Burn depth is an important predictive factor for mortality. An objective model can help estimate the probability of death in acute burn injury. LEVEL OF EVIDENCE: Prognostic study, level II.
Authors: Katherine J Choi; Christopher H Pham; Zachary J Collier; Melissa Mert; Ryan K Ota; Ruibei Li; Haig A Yenikomshian; Mandeep Singh; T Justin Gillenwater; Catherine M Kuza Journal: J Burn Care Res Date: 2020-07-03 Impact factor: 1.845