Prediction of significant bleeding during vitamin K antagonist treatment for venous thromboembolism in outpatients

Abstract

Bleeding is the most concerning complication associated with anticoagulant therapy but poorly characterized and important for risk/benefit assessment. We developed a risk stratification score to predict vitamin K antagonist (VKA)-associated bleeding in venous thromboembolism (VTE) using the UK Clinical Practice Research Datalink. Significant bleeding events in outpatients consisted of major bleeding and clinically relevant non-major bleeding requiring hospitalisation (CRNMB-H) within 90 days of VKA initiation. A scoring scheme for predicting bleeding was developed from subhazard ratios, validated using cross-validation and expressed by the Cstatistic. The study cohort consisted of 10,010 patients with first VTE receiving initial VKA treatment, mean age 62Á2 years. Between 2008 and 2016, 167 significant bleeding events were recorded (1Á7%), i.e. incidence rate was 7Á4/100 person-years. Independent predictors for community-acquired significant bleeding included active cancer, trauma/surgical procedure, male gender, dementia, liver disease, anaemia, history of bleeding, cerebrovascular, renal and chronic pulmonary disease, VTE presenting as pulmonary embolism and age over 75. The overall C-statistic was 0Á68 (95% CI, 0Á60-0Á76), 0Á75 (0Á60-0Á88) for major bleeding and 0Á65 (0Á55-0Á75) for CRNMB-H, and higher than in other risk schemes applied to our study population. The developed risk score may identify patients having a significant bleeding risk, in particular major bleeding events, in outpatients.

Estimating Bleeding Risk in Patients with Cancer-Associated Thrombosis: Evaluation of Existing Risk Scores and Development of a New Risk Score

“…This led to the identification of one additional relevant risk score. 30 ►Table 1 provides an overview of the eligible risk scores. Of these risk scores, unfortunately, the EINSTEIN score could not be externally validated in the Hokusai VTE Cancer study dataset due to the unavailability of age as a continuous variable and race for individual patients.…”

A Novel Risk Prediction Score for Clinically Significant Bleeding in Patients Anticoagulated for Venous Thromboembolism with Active Cancer

“…The incidence rate for a significant bleed was 5.12 (4.36-5.97) per 100 person-years in the low-risk group (0 to 1 score point), 18 A Novel Risk Prediction Score for Bleeding in Patients Anticoagulated for VTE Cohen et al…”

“…Algorithms for the definition of significant bleeding events have previously been developed. 18 Furthermore, a subset of identified potential significant bleeding events was manually reviewed and adjudicated by three physicians (A.T. C., S.C., and C.M.) utilizing patient summaries consisting of symptoms, signs, laboratory tests, therapies, discharge diagnoses, procedures and causes of death derived from GP records, HES, and ONS mortality data.…”

“…10 Most VTE-focused studies had very few Ca-VTE patients, which likely affected performance, due to reduced study power. 4,[10][11][12][13][14][15][16][17][18] Other limitations of these scores include failure to differentiate between cancer types, or between patients with a history of cancer and those with active cancer. Many include outcome events such as clinically relevant nonmajor bleeding (CRNMB) which does not require hospitalization (CRNMB-NH).…”

“…The range of possible risk factors for bleeding is broad and includes: age greater than 75 years, [20][21][22] prior history of major bleeding, 22 anemia, 20,21 chronic kidney disease, 22,23 liver disease, 18 thrombocytopenia, 20,21,24 cancer site, 21,[23][24][25] advanced or metastatic disease, 2,3,20,21 and concomitant anticancer therapies. 21 The ability to predict the risk of clinically significant bleeding during anticoagulation therapy would aid clinicians in mitigating the risks and in making decisions around short-and long-term duration (3-6 months), as well as the type and intensity of anticoagulation therapy.…”