Joint Modeling Of Longitudinal International Normalized Ratio And Activated Partial Thromboplastin Time Test Of Deep Venous Thrombosis Patients Receiving Treatment In Jimma University Medical Center:Ethiopia

Abstract

Background: Deep Venous Thrombosis is a condition in which one or multiple blood clots form in the deep veins of the legs. People at high risk of deep venous throm- bosis could take treatment of Anticoagulation to reduce their chance of developing serious conditions of stroke and heart attacks. The aim of this study is to address joint evolution and association of international normalized ratio and activated partial thromboplastin time over time and their associated risk factors. Methods: The study was conducted at Jimma University Medical Center. The latest data were collected from Jimma University Medical Center from rst, September 2018 to rst, January 2021. A linear mixed-e ects model was tted for the international normalized ratio and activated partial thromboplastin time outcomes. Then, a joint mixed-e ects model was tted for the two endpoints. Results: At baseline, the mean and standard deviation of both international normal- ized ratio and activated partial thromboplastin time were 1.378 and 0.6535 and 33.701 and 10.5738 respectively. In a separate analysis, the covariate smoking, alcohol user, prolonged immobilization, family history of DVT, and age are signi cant factors that a ect measure of deep venous thrombosis. In joint analysis case, smoking, alcohol user, family history of DVT, and age are important signi cant factors of deep venous thrombosis patients. Conclusion: The joint model ts the data well as compared to the separate model due to its smaller standard error. The results of the joint model suggested a very strong association between the evolutions of international normalized ratio and activated par- tial thromboplastin time. The evolution of association slowly increasing over time. Recommendation: Based on the results of this thesis joint model is recommended for researchers to any types of multivariate response variables for evaluating correlation and rate of changes over time.