烧伤严重程度分度方法的探讨

Categorization of burn severity

目的探寻烧伤严重程度分度的新方法。方法选取笔者单位1958年12月-2004年12月收治的有烧伤面积记录的单纯烧伤患者，用同病死率（病死率取0．5％）法确定烧伤患者的年龄分组。根据组内细分的各面积单元组病死率的统计学差异及临床情况，把各年龄组内患者烧伤程度划分为4度：轻、中、重、特重度。统计组内各烧伤程度对应的烧伤总面积范围、例数、病死率范围和Ⅲ度面积范围，将烧伤总面积范围和Ⅲ度面积范围作为各年龄组烧伤程度的面积分度标准。以患者病死概率为应变量，年龄、烧伤总面积、各种深度的烧伤面积为协变量建立Logistic回归模型。用该模型预测笔者单位2005年收治的患者病死概率，核对患者相关指标与前述分度标准是否符合；判断合并有吸人性损伤、严重合并伤或伤前重大疾病患者的烧伤严重程度。结果患者年龄最终分组为≤2岁、〉2岁且≤55岁、〉55岁。各组内烧伤程度的面积分度标准详见正文表2，Logistic回归模型为P（病死概率）=1÷[1＋e^-（-5.666-0.014×年龄＋0.041×烧伤总面积＋0.027×深Ⅱ度面积＋0.060×Ⅲ度面积）]，该方程预测的病死概率及相关指标与该分度标准符合率较高。有中度以上吸入性损伤患者可直接定为重度或特重度烧伤，仅有轻度吸人性损伤者依面积标准界定烧伤严重程度。结论Logistic回归模型预测的病死概率可作为判断烧伤严重程度的良好指标，前述年龄分组较合理，对应的烧伤程度分度标准较为准确、实用。

Objective To seek a new method for the categorization of burn severity. Methods Burn patients hospitalized in our center from December of 1958 to December of 2004 were enrolled in the study, and they were divided into different age groups according to same mortality, then the patients in each group were subdivided into 4 groups according to the burn severity ： i. e, mild burns, moderate burns, severe burns, serious severe burns. The total burn area, the number of eases, the mortality, and the area of Ⅲ degree burns were statistically analyzed in each subgroup, and the scope in total burn area and area of Ⅲ degree burns were taken as standards to define the degree of burns. The logistic regression equation was established with probability of death as the variable, and age, total burn area, burn area of different degrees as concomitant variables to form a logistic regression formula. It was used to predict the probability of death of patients hospitalized in 2005, so as to cheek whether the corresponding indices of these patients were eonsisrant with above standard of categorization into degrees, and to judge burn severity of the patients who had concomitant inhalation injury, severe associated injury, or those with serious disease before burns. Results The patients were divided into three groups： less than 2 years of age（ including 2 years of age） , 2 to 55 years of age（ including 55 years of age） , and older than 55 years of age groups. The classification standard of burn area was shown in table 2 of the article. The probability of death and corresponding indices predicted by the logistic regression equation were highly coincident with our standard. Patients with moderate inhalation injury could be regarded as patients with severe or most severe burns, while severity of those with mild inhalation injury could be determined by burn area alone. Conclusion The logistic regression equation is a good method to predict the severity of burn patients, with reasonable age specificity grouping, and accurate and practical scoring of division for corresponding burn severity.