Prediction Model For Predicting Positive Burn Wound Culture Based On Measured Index Of Body Fluid Exudation Period

Research purposes: In the acute infection stage of burn patients,if there is obvious wound infection,the wound will not heal,it can also lead to deeper burns,and even infection spreads the whole body,causing sepsis and life-threatening.If the risk of positive wound culture can be predicted according to the relevant in dicators measured in the early stage of the patient's burn(the fluid exudation period),then the targeted treatment plan,such as the application of preventive antibiotics,to prevent wound sepsis after burn,It is important to reduce the length of hospital stay and hospitalization expenses.Methods: This study collected patients who were diagnosed as “several burns” and met the inclusion criteria during the period of exudation of burns and plastics in the First People's Hospital of Changde City from July 2014 to December 2018.For example,the patient's gender,age group,burn cause,major burndepth,burn area(% TBSA total),shock,inhalation injury,type of antibiotic use,and positive wound culture were counted at the time of admission.EmppowerStats software was used.And R software(version 3.4.3)completed the input,collation and statistical analysis of the data.Using univariate analysis,preliminary analysis of relevant indicators related to whether the wound culture was positive,and then through multi-factor regression analysis to determine the indicators that are still positively correlated with the wound culture,to establish a predictive model,and through the split sample The way to perform internal verification.Results: Univariate analysis: age group,cause of burns,major burn depth,burn area(% TBSA),whether or not shock,whether inhalation injury was associated with positive wound culture,gender and prevention of antibiotic use were not associated with positive wound culture.Multivariate regression analysis: After controlling the confounding factors,whether the wound culture was positive was related to the burn area(% TBSA),whether it was shock or burn depth,and it was not related to age,burn cause,or whether it was inhaled.Validation after model establishment: In the modeled population,the area under the ROC curve of the clinical predictionmodel was 0.90(95% confidence interval(CI)),0.86-0.95).In the validation population,the area under the ROC curve of the clinical prediction model was 0.89(95% CI,0.80-0.97).Conclusions: Age,cause of burns,major burn depth,burn area(% TBSA),whether shock,whether inhalation injury is associated with positive wound culture.Whether the wound culture is positive or not is significantly related to the burn area(% TBS A),whether it is shock or burn depth,and is not affected by other variables.It can be used as a predictor for predicting wound culture.The established predictive model can predict the probability of positive burn wound culture through the relevant indicators measured during the body fluid exudation period.