Characterization and clinical assessment of 571 febrile intravenous drug users at risk for infective endocarditis presenting to an urban emergency department

Abstract

Study objectives: Infective endocarditis remains a difficult disease to diagnose with certainty and carries a high risk of morbidity and mortality if not treated aggressively with antibiotics. Epidemiologic studies of infective endocarditis have been hampered because of the rarity of disease and lack of precise case definition. We evaluate demographic information, clinical predictability, bacteriologic studies, echocardiographic findings, and outcomes associated with infective endocarditis in a population of intravenous drug users. Methods: This was a 3-year prospective cohort study of febrile (temperature ≥38.0°C) intravenous drug users presenting to an academic, inner-city emergency department (ED). Enrollment began in 1998 and included patients who were 18 years or older and able to consent. Initial evaluation consisted of a medical history and physical examination by the emergency physician, as well as diagnostic testing that included a minimum of 3 sets of blood cultures before the initiation of antibiotic therapy. Clinicians indicated level of suspicion for infective endocarditis according to ED assessment. Patients were admitted to a dedicated research unit housed by the Department of Medicine staff. Transthoracic echocardiography (TTE) was ordered immediately on admission. Patients with an identifiable source of infection were treated with an appropriate antibiotic, whereas others were treated empirically for infective endocarditis. Results: One thousand fifteen patients were screened for this study, 632 were enrolled, and 571 had complete data (blood cultures and TTE). Of patients with complete data, 96 had the discharge diagnosis of infective endocarditis. The mean age of these 96 patients was 38.51±7.98 years. Fifty-five episodes (57.3%) occurred in men and 41 (42.7%) in women. Black patients and white patients accounted for 87 (90.6%) and 9 (9.4%) cases, respectively. Patients with a discharge diagnosis were not significantly different than the patients without infective endocarditis in terms of age, sex, or race. Of infective endocarditis–positive patients with complete data, 36% had a previous episode of infective endocarditis, whereas 23% of patients who were infective endocarditis–negative had a history of infective endocarditis (P=.015). Clinical assessment by medical history and physical examination alone for identifying patients with infective endocarditis had a sensitivity of 0.37 (95% confidence interval [CI] 0.26 to 0.48) and specificity of 0.87 (95% CI 0.85 to 0.89). Ninety (93.8%) of 96 patients had positive blood culture findings, and 93.3% were positive in the first 24 hours. Sensitivity and specificity of positive bacteriologic findings were 0.94 (95% CI 0.87 to 0.97) and 0.77 (95% CI 0.76 to 0.78), respectively. Eighty-seven (90.6%) of 96 patients had a significant TTE finding. Sensitivity and specificity of any positive echocardiographic finding were 0.91 (95% CI 0.83 to 0.95) and 0.46 (95% CI 0.44 to 0.47), respectively. The median length of stay for patients with a discharge diagnosis of infective endocarditis was 11 days, with a mode of 9 days. For patients with a non–infective endocarditis discharge diagnosis, the median and mode length of stay were 3 and 2 days, respectively. Forty (41.7%) of the patients with infective endocarditis had at least 1 complication during their hospitalization, with embolic events being most prevalent and occurring in 29 patients (30.0%). Furthermore, 37% of infective endocarditis patients in this study were rehospitalized, and 8.3% died during this admission or within 90 days of being discharged for their infective endocarditis. Conclusion: Infective endocarditis is a significant cause of morbidity and mortality that is difficult to diagnose in a febrile intravenous drug user according to initial ED evaluation. Even with a high level of suspicion, emergency physicians identify only a small percentage of patients who eventually have infective endocarditis. Bacteriologic and echocardiographic findings are highly sensitive in detecting infective endocarditis. In fact, blood culture and TTE combined were able to identify all but 1 patient with a discharge diagnosis of infective endocarditis in our study. These findings support the current practice of hospitalizing all intravenous drug users presenting to the ED with fever for further diagnostic evaluation. Positive blood cultures within 24 hours in patients with infective endocarditis strongly argue for future investigations to develop novel approaches to rapidly evaluate at-risk patients in a short-stay setting. Study objectives: Infective endocarditis remains a difficult disease to diagnose with certainty and carries a high risk of morbidity and mortality if not treated aggressively with antibiotics. Epidemiologic studies of infective endocarditis have been hampered because of the rarity of disease and lack of precise case definition. We evaluate demographic information, clinical predictability, bacteriologic studies, echocardiographic findings, and outcomes associated with infective endocarditis in a population of intravenous drug users. Methods: This was a 3-year prospective cohort study of febrile (temperature ≥38.0°C) intravenous drug users presenting to an academic, inner-city emergency department (ED). Enrollment began in 1998 and included patients who were 18 years or older and able to consent. Initial evaluation consisted of a medical history and physical examination by the emergency physician, as well as diagnostic testing that included a minimum of 3 sets of blood cultures before the initiation of antibiotic therapy. Clinicians indicated level of suspicion for infective endocarditis according to ED assessment. Patients were admitted to a dedicated research unit housed by the Department of Medicine staff. Transthoracic echocardiography (TTE) was ordered immediately on admission. Patients with an identifiable source of infection were treated with an appropriate antibiotic, whereas others were treated empirically for infective endocarditis. Results: One thousand fifteen patients were screened for this study, 632 were enrolled, and 571 had complete data (blood cultures and TTE). Of patients with complete data, 96 had the discharge diagnosis of infective endocarditis. The mean age of these 96 patients was 38.51±7.98 years. Fifty-five episodes (57.3%) occurred in men and 41 (42.7%) in women. Black patients and white patients accounted for 87 (90.6%) and 9 (9.4%) cases, respectively. Patients with a discharge diagnosis were not significantly different than the patients without infective endocarditis in terms of age, sex, or race. Of infective endocarditis–positive patients with complete data, 36% had a previous episode of infective endocarditis, whereas 23% of patients who were infective endocarditis–negative had a history of infective endocarditis (P=.015). Clinical assessment by medical history and physical examination alone for identifying patients with infective endocarditis had a sensitivity of 0.37 (95% confidence interval [CI] 0.26 to 0.48) and specificity of 0.87 (95% CI 0.85 to 0.89). Ninety (93.8%) of 96 patients had positive blood culture findings, and 93.3% were positive in the first 24 hours. Sensitivity and specificity of positive bacteriologic findings were 0.94 (95% CI 0.87 to 0.97) and 0.77 (95% CI 0.76 to 0.78), respectively. Eighty-seven (90.6%) of 96 patients had a significant TTE finding. Sensitivity and specificity of any positive echocardiographic finding were 0.91 (95% CI 0.83 to 0.95) and 0.46 (95% CI 0.44 to 0.47), respectively. The median length of stay for patients with a discharge diagnosis of infective endocarditis was 11 days, with a mode of 9 days. For patients with a non–infective endocarditis discharge diagnosis, the median and mode length of stay were 3 and 2 days, respectively. Forty (41.7%) of the patients with infective endocarditis had at least 1 complication during their hospitalization, with embolic events being most prevalent and occurring in 29 patients (30.0%). Furthermore, 37% of infective endocarditis patients in this study were rehospitalized, and 8.3% died during this admission or within 90 days of being discharged for their infective endocarditis. Conclusion: Infective endocarditis is a significant cause of morbidity and mortality that is difficult to diagnose in a febrile intravenous drug user according to initial ED evaluation. Even with a high level of suspicion, emergency physicians identify only a small percentage of patients who eventually have infective endocarditis. Bacteriologic and echocardiographic findings are highly sensitive in detecting infective endocarditis. In fact, blood culture and TTE combined were able to identify all but 1 patient with a discharge diagnosis of infective endocarditis in our study. These findings support the current practice of hospitalizing all intravenous drug users presenting to the ED with fever for further diagnostic evaluation. Positive blood cultures within 24 hours in patients with infective endocarditis strongly argue for future investigations to develop novel approaches to rapidly evaluate at-risk patients in a short-stay setting.