Construction And Validation Of A Nomogram Predictive Model For Nosocomial Infetions After Hepatectomy

Objectives:We investigated the current situation of nosocomial infections after hepatectomy further analyzed the risk factors of nosocomial infections after hepatectomy,then constructed and verified a nomogram model for predicting the risk of nosocomial infections after hepatectomy,aiming to improve the attention of clinical medical personnel to nosocomial infections,enhance the understanding and awareness of nosocomial infections after hepatectomy,and promote the clinical translation and application of research results,help clinical staff to screen the high-risk groups of nosocomial infections at an early stage,formulate targeted nosocomial infections prevention measures,and finally achieve the purpose of reducing the incidence of nosocomial infections.Methods:Patients who underwent hepatectomy treatment in Shandong Province from January 1,2021,to December 31,2022,were used as the study subjects,following strict inclusion and exclusion criteria.We used the self-designed "Targeted Surveillance Questionnaire for Nosocomial Infection in Patients after Hepatectom" to collect relevant data,including basic patient information,perioperative conditions,and the occurrence of nosocomial infections.Divide the collected data into a modeling droup and a validation group using a 7:3 grouping method.By using univariate analysis,LASSO regression,ligistic regression,and other methods to gradually screen predictive factors,a nomogram prediction model is established.Verification group was used to verify the constructed model from three aspects:differentiation,consistency and clinical effectiveness.Results:1.A total of 678 patients with an average age of 55.57±11.194 years after hepatectomy were included,including 422 male patients(62.24%)and 256 female patients(37.76%).The mean age of patients in the modeling group(n=467)was 57.52±10.651 years old,and that in the validation group(n=211)was 55.21±10.923 years old.The two groups of data met the requirements of complete randomization.2.A total of 52 patients out of 678 hepatectomized patients in this study had hospitalacquired infections,with a hospital-acquired infection rate of 7.67%,and the incidence rate of hospital infection caces were 11.80%.The commonest sites of nosocomial infections after hepatectomy were surgical site infection(46.25%)and pulmonary infection(43.75%),with bloodstream infection(7.50%),and urinary tract infection(2.50%)occurring less frequently.3.In this study,38 strains of pathogenic bacteria were cultured,including 16 strains of Gram-positive bacteria(composition ratio 42.11%),21 strains of Gram-negative bacteria(composition ratio 55.26%)and 1 strain of fungus(composition ratio 2.63%).Fecal Enterococcus accounted for the largest proportion of Gram positive bacteria,and Klebsiella pneumoniae and Escherichia colicultivated fungi was tropical Candida.4.Multivariate regression analysis showed that bilioenterostomy,bile leakage,postoperative albumin infusion time,abdominal drainage time and temperature ≥38℃ on the third day after surgery were independent risk factors for postoperative nosocomial infection.5.Both internal and external verification of the line diagram model showed good prediction effect:the calibration curve was relaticlose to the actual curve,indicating that the calibration degree of the model was good;The AUC of the modeling group and validation group were 0.843 and 0.823,both greater than 0.75,indicating that the model had good distinguishing ability.The decision curve and clinical impact curve indicate that the medel had good clinical application value.6.In this study,static line segment nomogram prediction modle was transformed into dynamic web page nomogram prediction modle,which saved evaluation steps and time and was convenient for clinical application.Conclusions:The incidence of nosocomial infections after hepatectomy is high.Bilioenterostomy,bile leakage,postoperative albumin infusion time,abdominal drainage time,and temperature on the third postoperative day were independently associated with the occurrence of nosocomial infections after hepatectomy.The nomogram prediction model had good predictive performance and can be used as an effective tool to predict the risk of hospital infections after liver resection.It can identify high-risk groups,help medical staff formulate plans and take corresponding measures to reduce the occurrence of hospital infection.