| ObjectiveWith the rapid development of modern medicine,such as the widespread use of immunosuppressants,the development of large-scale organ transplantation,and the increase of invasive operations,the number of people with secondary immunodeficiency is increasing,and the number of infections is also increasing.This study mainly discusses the risk factors related to the population with secondary immunodeficiency and the composition of common pathogen spectrum(bacteria,fungi,viruses,etc.).At the same time,through the establishment of nomograph prediction model,it can predict the pathogen spectrum and the risk of infection with this pathogen spectrum of the population under different immunodeficiency states,so as to provide a simple,convenient and personalized nomograph for clinicians,So as to guide clinicians to start anti infection treatment in time and provide reference basis.Research methodExtract data(authentication id:46096653)from the public database minmic through structured query language(SQL),extract the results of pathogenic microorganisms in blood culture or bone marrow culture of patients who meet the inclusion criteria of secondary immunodeficiency as the research object,as the building module,and establish the nomogram prediction model;The data conforming to secondary immunodeficiency collected from the relevant departments of our hospital is used as the validation set data;In the modeling data,single factor analysis and multi factor Logistic regression analysis were used to screen the independent risk factors leading to the corresponding pathogen spectrum in patients with secondary immunodeficiency,and the nomogram model of the corresponding pathogen spectrum risk in patients with secondary immunodeficiency was constructed according to the independent risk factors;At the same time,internal verification is carried out on the training set data.Through the area AUC under the receiver operating characteristic curve(ROC)curve,Hosmer lemeshow test and decision curve analysis(DCA)are used to evaluate the effectiveness of the constructed model.Research results1.A total of 10483 patients with secondary immunodeficiency were included,including 5102 males and 5381 females,with an average age of 64.88 ± 15.96 years;2.10483 cases of pathogenic bacteria were detected,of which Gram-negative bacteria were the most common,mainly Escherichia coli,with a composition ratio of28.17%;There were 4204 cases of Gram-positive bacteria,mainly Staphylococcus aureus,with a constituent ratio of 21.39%;There were 48 cases of fungi,mainly Candida albicans,and the constituent ratio was 0.11%.3.Through univariate analysis and multivariate regression analysis,it was found that the independent risk factors leading to secondary immunodeficiency infection were age,sex,diabetes and the proportion of neutrophils;The nomograph risk prediction model based on these risk factors has an area of 0.6286 under the ROC curve,and the discrimination is general;Hosmer lemeshow test(H-L test)was used to calculate the calibration degree of the model.The results showed that: c2=11.65,P=0.1677(P > 0.05),indicating that the model fit well;The decision curve shows that the model can benefit from clinical intervention in a large range.Conclusion1.According to the distribution of pathogen spectrum,E.coli is the main Gbacteria infected by people with secondary immunodeficiency;Among G + bacteria,Staphylococcus aureus is the main one;Among the fungi,Candida albicans is the main one;2.Neutrophil count,age,sex and diabetes are independent risk factors for predicting the corresponding pathogen spectrum in patients with secondary immunodeficiency. |
PDF Full Text Request