Abstract
Symptomatic venous thromboembolism (SVTE) is a potentially significant complication which may occur following injury or surgery. Recent NICE guidelines, and clinical targets have all focused on decreasing in hospital death from acquired SVTE. Despite these guidelines there are no large studies investigating the risk factors for or incidence of SVTE in acute trauma admission.
Data from a prospective series of 9167 consecutive patients with a diagnosis of fractured neck of femur (NOF) at a single institution was used to construct a risk score for SVTE. Twenty three factors were screened with pairwise analysis. The cohort had an event rate of 1.4%. A multiple logistic regression model was used to construct a risk score and correct for confounding variables from nine significant factors identified by the pairwise analysis. Four factors; length of stay; chest infection; cardiac failure and transfusion were used to produce the final risk score. The score was statistically significant (p< 0.0001) and highly predictive (ROC analysis, AUC=0.76) of SVTE.
The score was separately validated in two cohorts from different Level 1 trauma centres. In one prospective consecutive cohort of 1000 NOF patients all components of the Nottingham SVTE score were found to be individually statistically significant (p< 0.0045). The score was further validated in a separate cohort of 3200 patients undergoing elective hip surgery. The score was found to be statistically significantly predictive of SVTE as a whole, and three of the four components were individually predictive in this patient cohort.
Balancing risks and benefits for thromboprophylaxis is key to reducing the risk of thromboembolic events, minimising bleeding and other complications associated with the therapy. Our study of 13,367 prospective patients is the largest of its type and we have successfully constructed and validated a scoring system that can be used to inform patient treatment decisions.
