The Birmingham Black Country Venous Thromboembolism (BBC-VTE) registry of outcomes of venous thromboembolism patients admitted to hospital

Abstract Background Venous thromboembolism (VTE) is a condition which causes significant morbidity and mortality in hospitalised patients as well as in the community. These are related to adverse outcomes associated with the first incidence of VTE, such as recurrence, major haemorrhage and mortality. These outcomes are difficult to measure and compare among the different regions of the world not least because of varying prevalence of risk factors, paucity of comparable studies and a differing approach to treatment of VTE (1–3). Purpose Our VTE registry aims to estimate the above mentioned outcomes as applicable to a developed world population with a multi-ethnic background. Furthermore, our registry has provided a data set to develop and validate a machine learning algorithm in order to predict outcomes in patients admitted to hospital with VTE. This will have profound implications for clinicians who will be able to make evidence-based decisions about which patients are low risk and so can be discharged early as well as those who are high risk and may need more intensive follow up. Methods BBC-VTE registry is a retrospective, multi-centre, observational registry. We identified all patients (N=1554) who were admitted with a final radiologically confirmed diagnosis of pulmonary embolism and/or lower limb deep vein thrombosis at three regional hospitals in the UK during the years 2012–2014. Each patient's electronic record was accessed by clinicians to confirm radiological diagnosis of VTE and also collect data on demographics, physical examination findings and laboratory analysis on admission, past medical history, and treatment plan. Outcomes were also recorded including recurrence of VTE, subsequent major bleeding and all-cause mortality. A simple multivariate analysis (logistic regression) was used to determine risk factors associated with all-cause mortality. Odds ratios (OR) and 95% CI are reported. Results The main factors determining higher all-cause mortality were age, a history of diabetes, admission laboratory analysis (c-reactive protein and neutrophil count), and previous malignancy (OR >1) (see Fig. 1). Conversely, hypercholesterolaemia, discharge oral anticoagulation, immobilization, and post-PE syndrome were all significantly correlated with a reduced risk of mortality (OR <1). Conclusions BBC-VTE provides unique data on VTE mortality risks in a multi-ethnic cohort. The strengths of our registry are that we are only including radiologically verified VTE patients unlike many of the epidemiological studies to date which rely on clinical coding. We have also shown that it is possible to risk-stratify patients admitted with VTE using simple clinical variables which has implications for their discharge decisions. Funding Acknowledgement Type of funding sources: None. Figure 1