Construction And Validation Of A Risk Prediction Model For Lower Extremity Varicose Veins In Operating Room Nurses

Objective: To construct a risk prediction model of varicose veins of lower extremities in operating room nurses and validate the model.The main objectives and significance of this project are: To help operating room nurses detect their own risks,and improve the awareness of varicose vein prevention of lower extremities with data guidance;Prompt asymptomatic patients to clarify diagnosis in time and treat them as soon as possible;Provide a reference for flexible scheduling for operating room nursing managers.Methods: A total of 483 operating room nurses from 9 tertiary hospitals and 7 secondary hospitals in Lanzhou were randomly assigned(random number was 20211219)to the modeling set and validation set according to the ratio of 7:3,univariate analysis and multivariate logistic regression were used to screen the independent influencing factors of varicose veins of the lower extremities in operating room nurses,the importance of the influencing factors was ranked by random forest algorithm,and the nomogram model was constructed and verified by R software and shinyapps.io The website is federated to build a dynamic nomogram page.Results:1.The prevalence of varicose veins of the lower extremities of the nurses in the modeling set was 49.55%,and the prevalence of varicose veins of the lower extremities of the nurses in the operating room was verified to be 50%.2.Through univariate analysis and multivariate logistics regression analysis,the results showed that nursing age,educational background,standing time≥ 4h/day,nature of work,and the number of days per week to wear varicose veins < 3 days were the main influencing factors for lower extremity varicose veins in operating room nurses.3.The results of the random forest algorithm ranking the importance of the influencing factors are as follows: age care > education> nature of work> number of days per week to wear varicose vein socks> standing time.4.The dynamic nomogram website built by RStudio software and shinyapps.io website is: https://xiazhijingmanquzhang.shinyapps.io/dangqiong/.5.The AUC value of the area under the internal verification ROC curve is 0.696,the calibration curve in the calibration diagram is close to the diagonal,the AUC value of the area under the external verification ROC curve is 0.683,and the calibration curve in the calibration diagram deviates from the diagonal compared with the internal verification,but the overall distribution is diagonal.Conclusion:1.In this study,non-invasive indicators were screened through univariate and multivariate logistic regression analysis,a varicose vein risk prediction model of lower extremities of operating room nurses was constructed,and a dynamic nomogram web page was made,which was simple and easy to operate,and could quickly evaluate its own risk score and screen high-risk groups,which could be used as the basis for the implementation of individualized prevention programs and as a reference for nursing scheduling in the operating room.2.The model verification index indicates that the model has only a good degree of differentiation,and the author will continue to improve the model in the future to meet the expected standards.