Insil Jang1, Sukjung Choo2, Kyunghee Kim3. 1. Department of Nursing, University of Ulsan, Ulsan, South Korea. 2. Department of Thoracic and Cardiovascular Surgery, Asan Medical Center, University of Ulsan College of Medicine, Seoul, South Korea. 3. College of Nursing, Chung-Ang University, Seoul, South Korea. Electronic correspondence: kyung@cau.ac.kr.
Abstract
BACKGROUND AND AIM OF THE STUDY: Mechanical valve replacement is associated with positive outcomes, but patients must undergo life-long anticoagulation therapy with warfarin, which carries a significant risk of bleeding complications. Therefore, a systematic and continuous assessment and supervision of warfarin treatment is essential in such patients, and approaches that can predict the risk of bleeding in advance are required. The study aim was to develop a classification tool to predict bleeding events in South Korean patients with mechanical valve replacement undergoing oral warfarin therapy. METHODS: The retrospective cohort study included 2,453 patients followed up for at least one year after valve replacement surgery, between January 2003 and December 2012. Discriminant analysis was used to assess potential bleeding risk factors out of 31 patient- related and disease-related descriptors. The prediction capability of the descriptors was evaluated based on accuracy, sensitivity,specificity, positive predictive value, and negative predictive value. RESULTS: A total of 13 descriptors including general, clinical-related and medication-related risk factors was selected as suitable predictors for bleeding risk. A novel classification tool was developed using these risk factors, and evaluated for accuracy (91.5%), sensitivity (80.2%), and specificity (95.2%). CONCLUSIONS: The classification tool developed in the present study can be reliably used in a clinical context to predict bleeding events in patients with mechanical valve replacement undergoing oral warfarin therapy.
BACKGROUND AND AIM OF THE STUDY: Mechanical valve replacement is associated with positive outcomes, but patients must undergo life-long anticoagulation therapy with warfarin, which carries a significant risk of bleeding complications. Therefore, a systematic and continuous assessment and supervision of warfarin treatment is essential in such patients, and approaches that can predict the risk of bleeding in advance are required. The study aim was to develop a classification tool to predict bleeding events in South Korean patients with mechanical valve replacement undergoing oral warfarin therapy. METHODS: The retrospective cohort study included 2,453 patients followed up for at least one year after valve replacement surgery, between January 2003 and December 2012. Discriminant analysis was used to assess potential bleeding risk factors out of 31 patient- related and disease-related descriptors. The prediction capability of the descriptors was evaluated based on accuracy, sensitivity,specificity, positive predictive value, and negative predictive value. RESULTS: A total of 13 descriptors including general, clinical-related and medication-related risk factors was selected as suitable predictors for bleeding risk. A novel classification tool was developed using these risk factors, and evaluated for accuracy (91.5%), sensitivity (80.2%), and specificity (95.2%). CONCLUSIONS: The classification tool developed in the present study can be reliably used in a clinical context to predict bleeding events in patients with mechanical valve replacement undergoing oral warfarin therapy.