Short and long-term mortality of patients presenting with bleeding events to the Emergency Department

Abstract

BackgroundDeath of patients presenting with bleeding events to the Emergency Department still represent a major problem. We sought to analyze clinical characteristics associated with worse outcomes including short- and long-term death, beyond antithombotic treatment strategy. MethodsPatients presenting with any bleeding events during 2016–2017years were enrolled. Clinical parameters, site of bleeding, major bleeding, ongoing anti-thrombotic treatment strategy and death were collected. Hard 5:1 propensity score matching was performed to adjust dead patients in baseline characteristics. Endpoints were one-month and one-year death. ResultsOut of 166,000 visits to the Emergency Department, 3.050 patients (1.8%) were enrolled and eventually 429 were analyzed after propensity. Overall, anticoagulants or antiplatelets were given to 234(54%). Major bleeding account for 111(26%) patients, without differences between those taking anticoagulants or antiplatelets versus others. Death at one-month and one-year was 26(6%) and 72(17%), respectively. Independent predictors of one-month death were major bleeding (Odds Ratio, OR 26, p<0.001), female gender (OR 7, p<0.001) and white blood cells (OR 1.2, p=0.01); of one-year were major bleeding (OR 7, p<0.001), age (OR 1.1, p<0.001) and female gender (OR 2.3, p=0.043). Of note, death rate of gastrointestinal and intracranial bleeding where higher than others (p<0.001). Overall mortality was approximately 40% on one-month; 60% in older patients and 80% in female gender with CHA2D2VASC-score≥2. Receiver operator characteristics analysis showed larger areas for major bleeding and age (0.75 and 0.72, respectively) over others; p<0.05 on C-statistic. ConclusionsIn patients with bleeding events, death rate was driven by major bleeding on short-term and older age on long-term. Among dead patients mortality was approximately 40% on one-month; 60% in older patients, and 80% in female gender.