Long-term causes of death in patients with infective endocarditis who undergo medical therapy only or surgical treatment: a nationwide population-based study

Long-term causes of death in patients with infective endocarditis who undergo medical therapy... Abstract OBJECTIVES It is known that patients surviving infective endocarditis have a poor long-term prognosis; however, few studies have addressed the long-term causes of death in patients surviving the initial hospitalization. METHODS Using Danish administrative registries, we identified patients admitted to a hospital with 1st time infective endocarditis in the period from January 1996 to December 2014, who were alive at the time of discharge. The study population was categorized into (i) patients undergoing medical therapy only and (ii) patients undergoing surgical and medical treatment. We examined the cardiovascular and non-cardiovascular causes of death. Using the Cox analysis, we investigated the associated risk of dying from a specific prespecified cause of death (heart failure, infective endocarditis and stroke) within the surgery group when compared with the medically treated group. RESULTS We identified 5576 patients: 4220 patients belonged to the medically treated group and 1356 patients to the surgery group. At the 10-year follow-up, the mortality rate was 63.1% and 41.6% in the medically treated group and the surgery group, respectively. Cardiovascular disease was the most frequent cause of death in both groups accounting for 52.5% in the medically treated group and 55.2% in the surgery group. Patients undergoing surgery were associated with a lower risk of dying from heart failure and stroke when compared with medically treated patients [hazard ratio = 0.66 (95% confidence interval: 0.46–0.94) and hazard ratio = 0.59 (95% confidence interval: 0.37–0.96), respectively]. CONCLUSIONS No major differences were found in the main causes of death between groups. Patients in the surgical group were associated with a lower risk of dying from heart failure and stroke when compared with medically treated patients. Infective endocarditis, Cause of death, Population study INTRODUCTION Infective endocarditis (IE) is a lethal disease with an in-hospital mortality of approximately 20% [1]. IE patients have a lower survival rate when compared with age- and gender-matched patients from the background population [2–5]. Patients undergoing heart valve surgery during IE hospitalization are reported to have a lower in-hospital death rate [6, 7] and a lower 1-year mortality when compared with patients undergoing medical therapy only [7]. The decision of operating patients with IE may depend on several factors resulting in substantial differences between patients receiving medical treatment only and surgical treatment. This knowledge might help guide clinicians in targeting secondary prevention strategies in a heterogeneous patient population. To increase the knowledge in the field of IE, we need data on long-term prognosis. Key problems subsequent to the IE diagnosis are heart failure, recurrence of IE and stroke [8, 9]. However, little data exist on whether these clinical problems also are the cause of death among these patients. Heart failure and malignancy have been identified as the most frequent long-term causes of death in IE patients [2]. Conversely, most of the epidemiological studies addressing this area have been carried out as case series, single-centre studies or with regional data collection [2–4, 10]. Using the Danish Nationwide Administrative Registries, we examined long-term mortality rates and differences in causes of death in IE patients discharged while still alive and stratified in patients who received heart valve surgery and in patients who were treated with medical therapy only. METHODS Data sources In Denmark, every resident is provided with a unique identification number. This makes possible to identify Danish citizens in a range of nationwide administrative registries. We used the following registries: the Danish Population Registry, the National Patient Registry, the Danish Prescription Registry and the Danish Cause of Death Registry. The Danish Population Registry holds information on date of birth, migration status and sex [11]. The National Patient Registry was initiated in 1977 and holds information on every patient admitted to a Danish hospital. A physician registers the diagnosis code of the disease for each admission according to the International Classification of Diseases (ICD)-8 until 1994 and the ICD-10 after 1994 [12]. The National Patient Registry was used to define the study population and baseline comorbidity [13]. Furthermore, since 1996, surgical procedures have been recorded in this registry according to the NOMESCO classification. Codes on treatment and medical examinations (including pacemaker implantation) were added to the National Patient Registry in 2000. The Danish Prescription Registry was established in 1994 and holds information on every prescription redeemed from all Danish pharmacies [14]. Prescriptions 6 months prior to the index date were used to assess baseline pharmacotherapy of the study population. The Anatomic Therapeutic Chemical Classification System is used to register the type of drug dispensed from the pharmacies. The Danish Cause of Death registry has been computerized since 1970. The primary and secondary causes of death are registered by the physician, thus the registry relies on the coding done by the individual physicians. The autopsy rate in Denmark is below 10%, and the automated classification of medical entities has been used since 2002 [15]. Study population Patients discharged alive after admission to a hospital with 1st time IE in the period from 1 January 1996 to 31 December 2014 were enrolled. We used the following ICD-8 and ICD-10 diagnosis codes to assess patients: 421, DI33, DI38 and DI39.8. In Denmark, all IE treatments are carried out in-hospital. Only patients with a hospitalization of >14 days were considered as having IE. The study population was categorized into (i) patients not undergoing heart valve surgery (medical therapy only) during IE hospital admission and (ii) patients undergoing heart valve surgery during IE hospital admission (for surgical procedure codes, see the Supplementary Material, Table S1). Patients undergoing valve surgery during follow-ups were assessed and analysed according to the treatment strategy (medical or medical + surgery) at the initial IE admission. Outcome The primary outcome was all-cause mortality. Danish death certificates provide opportunities for the physician to indicate multiple causes of death and the continuum of the fatal course based on the ICD-10 classification system. To assess cardiovascular (CV) death, the primary cause and secondary cause of death as recorded in the Cause of Death Registry were used (defined as ICD-10 codes of DI). If no CV death was registered as the primary cause or secondary cause of death, the primary cause of death was assessed to examine the non-CV cause of death. We categorized non-CV causes of death according to the disease groups of the ICD-10 classification system. Because IE patients are at high risk of heart failure, relapse and recurrence of IE and stroke in the course following the primary IE [7, 16, 17], we examined heart failure, IE and stroke as specific prespecified causes of death for the 2 groups. To do so, we used the primary and secondary causes of death. Statistical analysis Patients discharged alive were included on the day of discharge and were followed up until the end of the study period, death or emigration, whichever came first. Categorical variables are presented in counts and percentages, and continuously variables are presented with a median and 25 and 75 percentiles. The mortality rate for the 2 study groups were presented in a Kaplan–Meier plot. We used the log-rank test to test for difference between curves. Cumulative incidence plots considering competing risks of other death causes were used to illustrate differences in specific prespecified contributory causes of death (heart failure, IE and stroke), and the Fine and Gray test was used to test for differences between curves. The multivariable Cox proportional hazard analyses were used to examine the associated risk of dying from heart failure, IE or stroke among patients treated with surgery when compared with patients treated with medical therapy only. The following covariates were included in the multivariable models: age, calendar year, sex, cardiac implantable electronic device, cardiogenic shock, ischaemic heart disease, cerebrovascular disease, chronic obstructive lung disease, aortic stenosis, mitral regurgitation, renal disease, rheumatic disease, cancer, beta-blockade treatment, vitamin K-antagonist treatment, glucose lowering medication, corticosteroids and renin-angiotensin system inhibition. The proportional hazard assumption was tested, and when not fulfilled, follow-up time was split accordingly. The effect modification of sex on outcomes and age on outcomes were tested, as well as linearity of continuous variables. Results were presented with a hazard ratio (HR) and a 95% confidence interval (CI). A P-value <0.05 was considered statistically significant. All statistical analyses were performed using the SAS statistical software, version 9.4 (SAS Institute, Inc., Cary, NC, USA). For purposes of sensitivity, we matched the 2 study groups with sex- and age-matched controls from the background population to investigate differences in mortality rates and causes of death from the background population. A maximum 1-year difference in age was accepted between the study patients and matched controls. The greedy match algorithm was used for matching [18]. As a sensitivity analysis, we applied ischaemic heart disease, heart failure, cerebrovascular disease, chronic obstructive lung disease and cancer as matching criteria for the background population to investigate differences in mortality in a more comparable matched background population. In a landmark analysis, descriptive statistics was used to assess causes of death for patients surviving beyond 1 year of IE hospital discharge. As a sensitivity analysis, in-hospital mortality rate was identified by the 2 study groups. Further, baseline characteristics between groups were compared 5 years after index date. Ethics Register-based studies do not require ethical approval in Denmark. This study was approved by the Danish Data Protection Agency. RESULTS We identified 5576 patients with IE who were discharged alive; 4220 (75.7%) patients were treated with medical therapy only, and 1356 (24.3%) patients were treated with surgery during the 1st time IE hospitalization (Fig. 1). Baseline characteristics for the 2 study groups are presented in Table 1, and it can be observed that patients treated with medical therapy only were older and had overall more comorbidities. The median follow-up was 3.1 years (25 and 75 percentiles: 1.0 and 7.0 years, respectively) in the medically treated group and 5.3 years (25 and 75 percentiles 1.0 and 7.0 years, respectively) in the surgery group. Table 1: Baseline patient characteristics Medical therapy only Surgery P-value Demographics  Number 4220 1356  Male (%) 63.9 74.6 <0.0001  Age (years), median (25 and 75 percentiles) 70.0 (57.1–79.0) 60.7 (49.9–69.1) <0.0001 In-hospital  Duration of hospital admission (days), median (25 and 75 percentiles) 43.0 (32.0–53.0) 50.0 (42.0–66.0) <0.0001  Implantation of prosthetic valve (%) 91.0   Aortic prosthesisa (%) 53.2   Mitral prosthesisa (%) 27.4   Mitral repaira (%) 9.2 Comorbidity, medical history of  IHD (%) 28.6 15.1 <0.0001  Heart failure (%) 21.0 9.9 <0.0001  CVD (%) 13.7 7.3 <0.0001  COLD (%) 11.1 5.8 <0.0001  Cardiac implantable electronic device (%) 10.2 2.2 <0.0001  Aortic stenosis (%) 18.7 17.2 0.22  Mitral regurgitation (%) 7.9 7.5 0.68  Renal disease (%) 9.3 6.3 0.0005  Rheumatic disease (%) 8.1 4.7 <0.0001  Cancer (%) 15.3 8.5 <0.0001  Heart valve surgery (%) 14.6 8.4 <0.0001  Prosthetic valve (%) 12.5 6.9 <0.0001   Aortic prosthesis (%) 10.6 5.8   Mitral prosthesis (%) 1.5 1.0   Pulmonary prosthesis (%) 0.4 0.0 Prehospital medication  Beta-blocker (%) 29.9 18.7 <0.0001  VKA (%) 18.3 10.3 <0.0001  Glucose lowering medication (%) 8.8 5.8 0.0004  Corticosteroids (%) 11.4 7.2 <0.0001  Lipid lowering medication (%) 26.0 18.0 <0.0001  RAS inhibition (%) 32.8 26.6 <0.0001 Medical therapy only Surgery P-value Demographics  Number 4220 1356  Male (%) 63.9 74.6 <0.0001  Age (years), median (25 and 75 percentiles) 70.0 (57.1–79.0) 60.7 (49.9–69.1) <0.0001 In-hospital  Duration of hospital admission (days), median (25 and 75 percentiles) 43.0 (32.0–53.0) 50.0 (42.0–66.0) <0.0001  Implantation of prosthetic valve (%) 91.0   Aortic prosthesisa (%) 53.2   Mitral prosthesisa (%) 27.4   Mitral repaira (%) 9.2 Comorbidity, medical history of  IHD (%) 28.6 15.1 <0.0001  Heart failure (%) 21.0 9.9 <0.0001  CVD (%) 13.7 7.3 <0.0001  COLD (%) 11.1 5.8 <0.0001  Cardiac implantable electronic device (%) 10.2 2.2 <0.0001  Aortic stenosis (%) 18.7 17.2 0.22  Mitral regurgitation (%) 7.9 7.5 0.68  Renal disease (%) 9.3 6.3 0.0005  Rheumatic disease (%) 8.1 4.7 <0.0001  Cancer (%) 15.3 8.5 <0.0001  Heart valve surgery (%) 14.6 8.4 <0.0001  Prosthetic valve (%) 12.5 6.9 <0.0001   Aortic prosthesis (%) 10.6 5.8   Mitral prosthesis (%) 1.5 1.0   Pulmonary prosthesis (%) 0.4 0.0 Prehospital medication  Beta-blocker (%) 29.9 18.7 <0.0001  VKA (%) 18.3 10.3 <0.0001  Glucose lowering medication (%) 8.8 5.8 0.0004  Corticosteroids (%) 11.4 7.2 <0.0001  Lipid lowering medication (%) 26.0 18.0 <0.0001  RAS inhibition (%) 32.8 26.6 <0.0001 a Proportion calculated from ‘implantation of prosthetic valve’. COLD: chronic obstructive lung disease; CVD: cerebrovascular disease; IHD: ischaemic heart disease; RAS: renin-angiotensin system; VKA: vitamin K antagonists. Table 1: Baseline patient characteristics Medical therapy only Surgery P-value Demographics  Number 4220 1356  Male (%) 63.9 74.6 <0.0001  Age (years), median (25 and 75 percentiles) 70.0 (57.1–79.0) 60.7 (49.9–69.1) <0.0001 In-hospital  Duration of hospital admission (days), median (25 and 75 percentiles) 43.0 (32.0–53.0) 50.0 (42.0–66.0) <0.0001  Implantation of prosthetic valve (%) 91.0   Aortic prosthesisa (%) 53.2   Mitral prosthesisa (%) 27.4   Mitral repaira (%) 9.2 Comorbidity, medical history of  IHD (%) 28.6 15.1 <0.0001  Heart failure (%) 21.0 9.9 <0.0001  CVD (%) 13.7 7.3 <0.0001  COLD (%) 11.1 5.8 <0.0001  Cardiac implantable electronic device (%) 10.2 2.2 <0.0001  Aortic stenosis (%) 18.7 17.2 0.22  Mitral regurgitation (%) 7.9 7.5 0.68  Renal disease (%) 9.3 6.3 0.0005  Rheumatic disease (%) 8.1 4.7 <0.0001  Cancer (%) 15.3 8.5 <0.0001  Heart valve surgery (%) 14.6 8.4 <0.0001  Prosthetic valve (%) 12.5 6.9 <0.0001   Aortic prosthesis (%) 10.6 5.8   Mitral prosthesis (%) 1.5 1.0   Pulmonary prosthesis (%) 0.4 0.0 Prehospital medication  Beta-blocker (%) 29.9 18.7 <0.0001  VKA (%) 18.3 10.3 <0.0001  Glucose lowering medication (%) 8.8 5.8 0.0004  Corticosteroids (%) 11.4 7.2 <0.0001  Lipid lowering medication (%) 26.0 18.0 <0.0001  RAS inhibition (%) 32.8 26.6 <0.0001 Medical therapy only Surgery P-value Demographics  Number 4220 1356  Male (%) 63.9 74.6 <0.0001  Age (years), median (25 and 75 percentiles) 70.0 (57.1–79.0) 60.7 (49.9–69.1) <0.0001 In-hospital  Duration of hospital admission (days), median (25 and 75 percentiles) 43.0 (32.0–53.0) 50.0 (42.0–66.0) <0.0001  Implantation of prosthetic valve (%) 91.0   Aortic prosthesisa (%) 53.2   Mitral prosthesisa (%) 27.4   Mitral repaira (%) 9.2 Comorbidity, medical history of  IHD (%) 28.6 15.1 <0.0001  Heart failure (%) 21.0 9.9 <0.0001  CVD (%) 13.7 7.3 <0.0001  COLD (%) 11.1 5.8 <0.0001  Cardiac implantable electronic device (%) 10.2 2.2 <0.0001  Aortic stenosis (%) 18.7 17.2 0.22  Mitral regurgitation (%) 7.9 7.5 0.68  Renal disease (%) 9.3 6.3 0.0005  Rheumatic disease (%) 8.1 4.7 <0.0001  Cancer (%) 15.3 8.5 <0.0001  Heart valve surgery (%) 14.6 8.4 <0.0001  Prosthetic valve (%) 12.5 6.9 <0.0001   Aortic prosthesis (%) 10.6 5.8   Mitral prosthesis (%) 1.5 1.0   Pulmonary prosthesis (%) 0.4 0.0 Prehospital medication  Beta-blocker (%) 29.9 18.7 <0.0001  VKA (%) 18.3 10.3 <0.0001  Glucose lowering medication (%) 8.8 5.8 0.0004  Corticosteroids (%) 11.4 7.2 <0.0001  Lipid lowering medication (%) 26.0 18.0 <0.0001  RAS inhibition (%) 32.8 26.6 <0.0001 a Proportion calculated from ‘implantation of prosthetic valve’. COLD: chronic obstructive lung disease; CVD: cerebrovascular disease; IHD: ischaemic heart disease; RAS: renin-angiotensin system; VKA: vitamin K antagonists. Figure 1: View largeDownload slide The selection process of the study population is shown. In total, 17 patients died without a registered cause of death. CoD: cause of death; FU: follow-up; IE: infective endocarditis. Figure 1: View largeDownload slide The selection process of the study population is shown. In total, 17 patients died without a registered cause of death. CoD: cause of death; FU: follow-up; IE: infective endocarditis. During follow-up, heart valve surgery was observed in 10.4% of patients treated with medicine only during hospitalization for IE and in 6.1% of patients in the IE surgery group. Within the 1st year of follow-up, 54.9% of patients treated with medical therapy only and 42.2% of patients who underwent surgery at IE admission had undergone heart valve surgery. Within the first 5 years of follow-up, these numbers were increased to 84.7% and 73.5% for patients treated with medical therapy only and patients who underwent surgery during IE admission, respectively. Of the patients categorized as medically treated patients, 418 (9.9%) had a pacemaker vs 2.1% among patients categorized as surgically treated during IE admission. Causes of death Of the total study population, 2544 patients died and had a registered cause of death, see Fig. 1 [17 (0.7%) patients had no registered cause of death]: 2122 (50.3%) patients in the medically treated group and 422 (31.1%) patients in the surgery group. Throughout the study period, the mortality rate was higher in IE patients treated only with medicine in-hospital when compared with IE patients treated with surgery (Fig. 2). Causes of death were similar between groups with CV disease as the most frequent cause of death followed by neoplasms for both study groups (Table 2). Table 3 illustrates the specific prespecified causes of death (heart failure, IE and stroke) for both the study groups. Table 2: Cause of death in infective endocarditis patients discharged alive in the period from 1996 to 2014 Medical therapy only Surgery Number 2122 422 Diseases of the circulatory system (%) 52.5 55.2 Neoplasms (%) 16.6 19.2 Diseases of the respiratory system (%) 5.2 3.8 Infection (%) 5.7 3.6 Endocrine, nutritional and metabolic disorders (%) 3.4 3.6 Diseases of the digestive system (%) 4.4 3.6 Diseases of the genitourinary system (%) 2.5 1.4 Injury, poisoning and external causes (%) 2.7 3.8 Unknown 2.6 3.3 Mental and behavioural disorders (%) 1.6 0.7 Diseases of the nervous system (%) 1.4 1.0 Diseases of the musculoskeletal system and connective tissue (%) 0.8 0.7 Congenital disease 0.4 0.0 Diseases of the skin (%) 0.2 0.2 Medical therapy only Surgery Number 2122 422 Diseases of the circulatory system (%) 52.5 55.2 Neoplasms (%) 16.6 19.2 Diseases of the respiratory system (%) 5.2 3.8 Infection (%) 5.7 3.6 Endocrine, nutritional and metabolic disorders (%) 3.4 3.6 Diseases of the digestive system (%) 4.4 3.6 Diseases of the genitourinary system (%) 2.5 1.4 Injury, poisoning and external causes (%) 2.7 3.8 Unknown 2.6 3.3 Mental and behavioural disorders (%) 1.6 0.7 Diseases of the nervous system (%) 1.4 1.0 Diseases of the musculoskeletal system and connective tissue (%) 0.8 0.7 Congenital disease 0.4 0.0 Diseases of the skin (%) 0.2 0.2 Disease categories are from the International Classification of Diseases-10. Table 2: Cause of death in infective endocarditis patients discharged alive in the period from 1996 to 2014 Medical therapy only Surgery Number 2122 422 Diseases of the circulatory system (%) 52.5 55.2 Neoplasms (%) 16.6 19.2 Diseases of the respiratory system (%) 5.2 3.8 Infection (%) 5.7 3.6 Endocrine, nutritional and metabolic disorders (%) 3.4 3.6 Diseases of the digestive system (%) 4.4 3.6 Diseases of the genitourinary system (%) 2.5 1.4 Injury, poisoning and external causes (%) 2.7 3.8 Unknown 2.6 3.3 Mental and behavioural disorders (%) 1.6 0.7 Diseases of the nervous system (%) 1.4 1.0 Diseases of the musculoskeletal system and connective tissue (%) 0.8 0.7 Congenital disease 0.4 0.0 Diseases of the skin (%) 0.2 0.2 Medical therapy only Surgery Number 2122 422 Diseases of the circulatory system (%) 52.5 55.2 Neoplasms (%) 16.6 19.2 Diseases of the respiratory system (%) 5.2 3.8 Infection (%) 5.7 3.6 Endocrine, nutritional and metabolic disorders (%) 3.4 3.6 Diseases of the digestive system (%) 4.4 3.6 Diseases of the genitourinary system (%) 2.5 1.4 Injury, poisoning and external causes (%) 2.7 3.8 Unknown 2.6 3.3 Mental and behavioural disorders (%) 1.6 0.7 Diseases of the nervous system (%) 1.4 1.0 Diseases of the musculoskeletal system and connective tissue (%) 0.8 0.7 Congenital disease 0.4 0.0 Diseases of the skin (%) 0.2 0.2 Disease categories are from the International Classification of Diseases-10. Table 3: Specific prespecified contributory causes of death by the 2 study groups Medical therapy only Surgery Number 2122 422 Heart failure (%) 13.2 10.4 Infective endocarditis (%) 8.9 11.1 Stroke (%) 9.3 10.4 Medical therapy only Surgery Number 2122 422 Heart failure (%) 13.2 10.4 Infective endocarditis (%) 8.9 11.1 Stroke (%) 9.3 10.4 Table 3: Specific prespecified contributory causes of death by the 2 study groups Medical therapy only Surgery Number 2122 422 Heart failure (%) 13.2 10.4 Infective endocarditis (%) 8.9 11.1 Stroke (%) 9.3 10.4 Medical therapy only Surgery Number 2122 422 Heart failure (%) 13.2 10.4 Infective endocarditis (%) 8.9 11.1 Stroke (%) 9.3 10.4 Figure 2: View largeDownload slide The mortality rates of patients who underwent medical therapy only during the hospital stay for infective endocarditis and patients who underwent heart valve surgery during the hospital stay for infective endocarditis are shown. P-value tests for difference between curves using the log-rank test. Figure 2: View largeDownload slide The mortality rates of patients who underwent medical therapy only during the hospital stay for infective endocarditis and patients who underwent heart valve surgery during the hospital stay for infective endocarditis are shown. P-value tests for difference between curves using the log-rank test. Patients undergoing surgery during follow-up In patients undergoing surgery during follow-up [437 (10.4%) patients from the medically treated group and 83 (6.1%) patients from the surgery group], the mortality rate was 3.4% and 17.6% at 1- and 5-year follow-up, respectively, in patients treated with medical therapy only during admission, whereas this rate was 6.0% and 17.4% at 1- and 5-year follow-up, respectively, in the patients undergoing surgery (P = 0.62). At 1-year follow-up, CV disease was the most common cause of death (80.0%) followed by neoplasms (13.3%) and infection (6.7%) for the medically treated group (n = 15), whereas this was CV disease (80.0%) and neurological disorder (20.0%) for the surgery group (n = 5). At 5-year follow-up, CV disease was the most common cause of death (70.4%) followed by neoplasms (19.7%) and infection (2.8%) for the medically treated group (n = 71), whereas this was CV disease (78.6%) and other (neurological disorder, external causes, unknown, 21.4%) (n = 14) for the surgery group. Landmark analysis We identified 4725 patients surviving beyond 1 year after IE hospitalization: 73.1% of the patients treated with medical therapy only during IE admission and 26.9% had received surgery. With a total of 2 and 5 years of follow-up, the mortality rate was 9.9% and 33.7% in the medically treated patients, respectively, whereas this was 4.1% and 14.7% for the surgery group, respectively. The most frequent causes of death at 5-year follow-up were CV disease (50.5%), neoplasms (16.7%) and infection (5.7%) for the medically treated group, whereas this was CV disease (60.1%), neoplasms (15.5%) and infection (3.4%) for the surgery group. Differences in specific prespecified contributory causes of death Figure 3 shows cumulative incidence plots for the specific prespecified causes of death (heart failure, IE and stroke) of the 2 study groups up to 10 years of follow-up. Patients undergoing surgery were associated with a lower risk of dying due to heart failure, HR = 0.67 (95% CI: 0.47–0.95), when compared with patients treated with medical therapy only (Fig. 4). We found no difference in the associated risk of dying from IE, HR = 0.99 (95% CI: 0.70–1.40), with the medically treated group as reference (Fig. 4). The proportional hazard assumption was not fulfilled for dying from stroke up to 10 years of follow-up. When considering stroke for the first 5 years of follow-up, patients undergoing surgery were associated with a lower risk of dying from stroke when compared with the medically treated patients, HR = 0.59 (95% CI: 0.37–0.96). In contrast, at the follow-up period of 5–10 years, the patients in the surgery group were associated with an increased risk of dying from stroke when compared with medically treated patients, HR = 2.05 (95% CI 1.13–3.71), Fig. 4. Baseline characteristics are compared between study groups 5 years after index, and it is observed that the comorbidities of the groups become more similar during follow-up, see Supplementary Material, Table S3. Age modified the associated risk of dying from heart failure, P = 0.04 for interaction. Age did not modify the associated risk of dying from IE and stroke, P = 0.06 and P = 0.1 for interaction, respectively. Sex did not modify the associated risk of dying from heart failure, IE and stroke, P = 0.97, P = 0.68 and P = 0.27 for interaction, respectively. Figure 3: View largeDownload slide Specific prespecified contributory causes of death. The cumulative incidence of cause of death from (A) heart failure, (B) infective endocarditis and (C) stroke for the 2 study groups. P-value tests for difference between curves using Gray’s test. Figure 3: View largeDownload slide Specific prespecified contributory causes of death. The cumulative incidence of cause of death from (A) heart failure, (B) infective endocarditis and (C) stroke for the 2 study groups. P-value tests for difference between curves using Gray’s test. Figure 4: View largeDownload slide Associated risk of dying from specific prespecified contributory causes of death for patients treated with surgical intervention during IE admission when compared with patients treated with medical therapy only. *Follow-up from 0 to 5 years. **Follow-up from 5 to 10 years. CI: cumulative incidence; FU: follow-up; HR: hazard ratio. Figure 4: View largeDownload slide Associated risk of dying from specific prespecified contributory causes of death for patients treated with surgical intervention during IE admission when compared with patients treated with medical therapy only. *Follow-up from 0 to 5 years. **Follow-up from 5 to 10 years. CI: cumulative incidence; FU: follow-up; HR: hazard ratio. Sensitivity analysis: a comparison between background population and in-hospital mortality rate Each study group was compared with sex- and age-matched controls from the background population (see Supplementary Material, Table S2 for baseline characteristics). Controls for the medically treated group had a mortality rate of 3.3%, 16.3% and 31.6% after 1, 5 and 10 years of follow-up, respectively (P < 0.0001 when compared with medically treated IE patients). Controls for the surgically treated group had a mortality rate of 1.7%, 8.3% and 18.6% at 1, 5 and 10 years of follow-up, respectively (P < 0.0001 when compared with surgically treated IE patients). The most common causes of death over the 10-year study period were CV diseases (40.3%), neoplasms (23.9%), respiratory disorders (6.2%) and other (29.6%) for the controls of the medically treated group, whereas for the controls of the surgically treated group, this was CV disease (35.6%), neoplasms (33.5%), respiratory disorders (6.4%) and other (24.5%). No major differences in the causes of death were observed when we conducted a sensitivity analysis with very strict matching criteria for the matched background population as shown in the Supplementary Material, Table S4a and b. When including patients who died during IE hospitalization, we identified a mortality rate of 23.4% among patients treated with medical therapy only and 12.6% among patients in the surgical group, see Supplementary Material, Table S5. DISCUSSION This study investigated differences in long-term mortality and causes of death in surviving IE patients treated with medical therapy only and patients treated with surgery during IE admission. The study had 3 major findings. First, over a 10-year follow-up, 63.1% died in the medically treated group, whereas 41.6% died in the surgery group. Second, the most common cause of death was CV disease in both groups. For non-CV causes of death, neoplasms were the most common in both groups. Third, differences were found between groups when looking at specific prespecified causes of death (heart failure, IE and stroke) where patients from the surgery group had a lower associated risk of dying from heart failure and stroke when compared with patients in the medically treated group. Several studies have shown differences in outcomes in patients undergoing medical therapy only and surgical treatment during IE admission [5–7, 19, 20]. Reasons for this difference are due to differences in patient characteristics and the severity of the IE case and less likely the intervention itself. Patients with valve destruction leading to heart failure, uncontrolled infection or large vegetations with high risk of emboli should undergo surgery if the patient has a low burden of comorbidity. On the other hand, patients treated with medical therapy only constitute a patient group with IE without the need for surgical treatment or a patient group with a need for surgical treatment where surgery is avoided due to a high burden of comorbidities. Hence, different approaches may be needed in the follow-up of IE survivors to secure patient care at a high level. Our study found no differences in the relative frequencies of the major causes of death between patients undergoing medical therapy only and patients undergoing surgery. This illustrates the need to address diseases known to be frequent causes of death in the background population in IE patients as well [21, 22]. Our study adds valuable knowledge to the clinical awareness of IE survivors. Furthermore, our results justify the close monitoring of patients surviving IE, which has also been suggested by other authors [4]. These findings are of considerable clinical interest as little attention has been given to IE survivors. To ensure the best prevention strategy, it is of importance to acknowledge the diseases that are the most frequent causes of death among these patients. The differences in patient characteristics are reflected in the substantial differences in the mortality rate found in our study (61.3% vs 41.6% over a 10-year study period for the medically treated group and the surgery group, respectively). In line with the results of our study, several other studies have found an excess mortality in IE survivors when compared with sex- and age-matched controls [2, 4]; however, up until now, it has been unknown as to which disease category is the most frequent cause of death. A Dutch study investigating long-term follow-up in 191 patients with IE found a mortality rate of 41% after 10-year follow-up and found the major causes of death to be congestive heart failure and malignancy [2]. A German study identified a mortality rate of 42.8% after 12 years of follow-up in IE patients who underwent valve surgery where long-term cause of death primarily was of extracardiac origin [10]. The mortality rate identified in the German study is in line with our results (41.6%); however, our data show that the most common cause of death was a CV disease rather than an extracardiac origin. An explanation for this difference may be found in the different populations studied. Our study is strengthened by investigating a nationwide cohort with a large sample size where very few patients emigrated during follow-up. We found that patients undergoing surgery for IE were associated with a lower risk of dying from heart failure and stroke when compared with patients treated with medical therapy only. A more complex clinical picture with renal failure, sepsis and surgical contraindications in the medically treated group may be the reasons for these findings. A large population-based study from Taiwan showed that IE survivors carried a significantly higher long-term risk of major adverse cardiac events when compared with matched controls from the background population [23]. These findings along with the results of our study show a need for increased clinical awareness and patient care in IE survivors. Limitations Our study has several limitations. First, our results are found in the Danish Cause of Death Registry where the cause of death is assessed by a single physician, and the autopsy rate is less than 10% [15]. Validation of such a registry is difficult; however, a validation study of the Danish Cause of Death registry identified a sensitivity of the myocardial infarction diagnosis at 92.8% [24]. Second, clinical characteristics during IE admission were not available. Important characteristics such as microbiological aetiology, echocardiographic findings and differentiation of left- or right-sided IE or prosthetic valve endocarditis would have given a more complete picture of the study population as it is known that Staphylococcus aureus IE is associated with worse outcomes [25]. Furthermore, data on antimicrobial therapy, antimicrobial resistance and duration of antimicrobial therapy were not accessible through the administrative registries. Third, in view of the long-term follow-up, time-dependent confounding factors may have influenced the results. CONCLUSION In conclusion, no major differences were found in the relative frequencies of CV and non-CV cause of death between patients treated with medical therapy only and patients treated with surgery. Patients undergoing surgical treatment for IE were associated with a lower risk of dying from heart failure and stroke when compared with patients treated medically only. No difference was found between surgically and medically treated patients in the associated risk of dying from IE subsequent to index hospitalization. SUPPLEMENTARY MATERIAL Supplementary material is available at EJCTS online. Conflict of interest: none declared. REFERENCES 1 Cresti A , Chiavarelli M , Scalese M , Nencioni C , Valentini S , Guerrini F et al. Epidemiological and mortality trends in infective endocarditis, a 17-year population-based prospective study . Cardiovasc Diagn Ther 2017 ; 7 : 27 – 35 . Google Scholar CrossRef Search ADS PubMed 2 Mokhles MM , Ciampichetti I , Domburg R , van Cheng JM , Bogers AJJC , Witsenburg M. Infective endocarditis in a tertiary referral hospital: long-term follow up . J Heart Valve Dis 2012 ; 21 : 118 – 24 . Google Scholar PubMed 3 Mokhles MM , Ciampichetti I , Head SJ , Takkenberg JJ , Bogers AJ. Survival of surgically treated infective endocarditis: a comparison with the general Dutch population . Ann Thorac Surg 2011 ; 91 : 1407 – 12 . Google Scholar CrossRef Search ADS PubMed 4 Thuny F , Giorgi R , Habachi R , Ansaldi S , Le Dolley Y , Casalta J-P et al. Excess mortality and morbidity in patients surviving infective endocarditis . Am Heart J 2012 ; 164 : 94 – 101 . Google Scholar CrossRef Search ADS PubMed 5 Ternhag A , Cederström A , Törner A , Westling K. A nationwide cohort study of mortality risk and long-term prognosis in infective endocarditis in Sweden . PLoS One 2013 ; 8 : e67519 . Google Scholar CrossRef Search ADS PubMed 6 Delahaye F , Alla F , Béguinot I , Bruneval P , Doco-Lecompte T , Lacassin F et al. In-hospital mortality of infective endocarditis: prognostic factors and evolution over an 8-year period . Scand J Infect Dis 2007 ; 39 : 849 – 57 . Google Scholar CrossRef Search ADS PubMed 7 Kiefer T , Park L , Tribouilloy C , Cortes C , Casillo R , Chu V et al. Association between valvular surgery and mortality among patients with infective endocarditis complicated by heart failure . JAMA 2011 ; 306 : 2239 – 47 . Google Scholar CrossRef Search ADS PubMed 8 Thuny F , Avierinos J-F , Tribouilloy C , Giorgi R , Casalta J-P , Milandre L et al. Impact of cerebrovascular complications on mortality and neurologic outcome during infective endocarditis: a prospective multicentre study . Eur Heart J 2007 ; 28 : 1155 – 61 . Google Scholar CrossRef Search ADS PubMed 9 Tornos P , Iung B , Permanyer-Miralda G , Baron G , Delahaye F , Gohlke-Bärwolf C et al. Infective endocarditis in Europe: lessons from the Euro heart survey . Heart 2005 ; 91 : 571. Google Scholar CrossRef Search ADS PubMed 10 Spiliopoulos K , Giamouzis G , Haschemi A , Karangelis D , Antonopoulos N , Fink G et al. Surgical management of infective endocarditis: early and long-term mortality analysis. single-center experience and brief literature review . Hellenic J Cardiol 2014 ; 55 : 462 – 74 . Google Scholar PubMed 11 Schmidt M , Pedersen L , Sørensen HT. The Danish Civil Registration System as a tool in epidemiology . Eur J Epidemiol 2014 ; 29 : 541 – 9 . Google Scholar CrossRef Search ADS PubMed 12 Schmidt M , Schmidt SAJ , Sandegaard JL , Ehrenstein V , Pedersen L , Sørensen HT. The Danish National Patient Registry: a review of content, data quality, and research potential . Clin Epidemiol 2015 ; 7 : 449 – 90 . Google Scholar CrossRef Search ADS PubMed 13 Thygesen SK , Christiansen CF , Christensen S , Lash TL , Sørensen HT. The predictive value of ICD-10 diagnostic coding used to assess Charlson comorbidity index conditions in the population-based Danish National Registry of Patients . BMC Med Res Methodol 2011 ; 11 : 83 . Google Scholar CrossRef Search ADS PubMed 14 Kildemoes HW , Sørensen HT , Hallas J. The Danish National Prescription Registry . Scand J Public Health 2011 ; 39 : 38 – 41 . Google Scholar CrossRef Search ADS PubMed 15 Helweg-Larsen K. The Danish Register of causes of death . Scand J Public Health 2011 ; 39 : 26 – 9 . Google Scholar CrossRef Search ADS PubMed 16 Merkler AE , Chu SY , Lerario MP , Navi BB , Kamel H. Temporal relationship between infective endocarditis and stroke . Neurology 2015 ; 85 : 512 – 16 . Google Scholar CrossRef Search ADS PubMed 17 Alagna L , Park LP , Nicholson BP , Keiger AJ , Strahilevitz J , Morris A et al. Repeat endocarditis: analysis of risk factors based on the International Collaboration on Endocarditis—Prospective Cohort Study . Clin Microbiol Infect 2014 ; 20 : 566 – 75 . Google Scholar CrossRef Search ADS PubMed 18 Locally Written SAS Macros—Division of Biomedical Statistics and Informatics—Mayo Clinic Research. http://www.mayo.edu/research/departments-divisions/department-health-sciences-research/division-biomedical-statistics-informatics/software/locally-written-sas-macros (30 March 2018, date last accessed). 19 Bannay A , Hoen B , Duval X , Obadia J-F , Selton-Suty C , Le Moing V et al. The impact of valve surgery on short- and long-term mortality in left-sided infective endocarditis: do differences in methodological approaches explain previous conflicting results? Eur Heart J 2011 ; 32 : 2003 – 15 . Google Scholar CrossRef Search ADS PubMed 20 Heiro M , Helenius H , Hurme S , Savunen T , Metsärinne K , Engblom E et al. Long-term outcome of infective endocarditis: a study on patients surviving over one year after the initial episode treated in a Finnish teaching hospital during 25 years . BMC Infect Dis 2008 ; 8 : 49. Google Scholar CrossRef Search ADS PubMed 21 GBD 2013 Mortality and Causes of Death Collaborators . Global, regional, and national age–sex specific all-cause and cause-specific mortality for 240 causes of death, 1990–2013: a systematic analysis for the Global Burden of Disease Study 2013 . Lancet 2015 ; 385 : 117 – 71 . CrossRef Search ADS PubMed 22 Feigin VL , Forouzanfar MH , Krishnamurthi R , Mensah GA , Connor M , Bennett DA et al. Global and regional burden of stroke during 1990-2010: findings from the Global Burden of Disease Study 2010 . Lancet Lond Engl 2014 ; 383 : 245 – 54 . Google Scholar CrossRef Search ADS 23 Shih C-J , Chu H , Chao P-W , Lee Y-J , Kuo S-C , Li S-Y et al. Long-term clinical outcome of major adverse cardiac events in survivors of infective endocarditis: a nationwide population-based study . Circulation 2014 ; 130 : 1684 – 91 . Google Scholar CrossRef Search ADS PubMed 24 Madsen M , Davidsen M , Rasmussen S , Abildstrom SZ , Osler M. The validity of the diagnosis of acute myocardial infarction in routine statistics: a comparison of mortality and hospital discharge data with the Danish MONICA registry . J Clin Epidemiol 2003 ; 56 : 124 – 30 . Google Scholar CrossRef Search ADS PubMed 25 Chu VH , Cabell CH , Benjamin DK , Kuniholm EF , Fowler VG , Engemann J et al. Early predictors of in-hospital death in infective endocarditis . Circulation 2004 ; 109 : 1745 – 9 . Google Scholar CrossRef Search ADS PubMed © The Author(s) 2018. Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png European Journal of Cardio-Thoracic Surgery Oxford University Press

Long-term causes of death in patients with infective endocarditis who undergo medical therapy only or surgical treatment: a nationwide population-based study

Loading next page...
 
/lp/ou_press/long-term-causes-of-death-in-patients-with-infective-endocarditis-who-CLcXinmL7q
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved.
ISSN
1010-7940
eISSN
1873-734X
D.O.I.
10.1093/ejcts/ezy156
Publisher site
See Article on Publisher Site

Abstract

Abstract OBJECTIVES It is known that patients surviving infective endocarditis have a poor long-term prognosis; however, few studies have addressed the long-term causes of death in patients surviving the initial hospitalization. METHODS Using Danish administrative registries, we identified patients admitted to a hospital with 1st time infective endocarditis in the period from January 1996 to December 2014, who were alive at the time of discharge. The study population was categorized into (i) patients undergoing medical therapy only and (ii) patients undergoing surgical and medical treatment. We examined the cardiovascular and non-cardiovascular causes of death. Using the Cox analysis, we investigated the associated risk of dying from a specific prespecified cause of death (heart failure, infective endocarditis and stroke) within the surgery group when compared with the medically treated group. RESULTS We identified 5576 patients: 4220 patients belonged to the medically treated group and 1356 patients to the surgery group. At the 10-year follow-up, the mortality rate was 63.1% and 41.6% in the medically treated group and the surgery group, respectively. Cardiovascular disease was the most frequent cause of death in both groups accounting for 52.5% in the medically treated group and 55.2% in the surgery group. Patients undergoing surgery were associated with a lower risk of dying from heart failure and stroke when compared with medically treated patients [hazard ratio = 0.66 (95% confidence interval: 0.46–0.94) and hazard ratio = 0.59 (95% confidence interval: 0.37–0.96), respectively]. CONCLUSIONS No major differences were found in the main causes of death between groups. Patients in the surgical group were associated with a lower risk of dying from heart failure and stroke when compared with medically treated patients. Infective endocarditis, Cause of death, Population study INTRODUCTION Infective endocarditis (IE) is a lethal disease with an in-hospital mortality of approximately 20% [1]. IE patients have a lower survival rate when compared with age- and gender-matched patients from the background population [2–5]. Patients undergoing heart valve surgery during IE hospitalization are reported to have a lower in-hospital death rate [6, 7] and a lower 1-year mortality when compared with patients undergoing medical therapy only [7]. The decision of operating patients with IE may depend on several factors resulting in substantial differences between patients receiving medical treatment only and surgical treatment. This knowledge might help guide clinicians in targeting secondary prevention strategies in a heterogeneous patient population. To increase the knowledge in the field of IE, we need data on long-term prognosis. Key problems subsequent to the IE diagnosis are heart failure, recurrence of IE and stroke [8, 9]. However, little data exist on whether these clinical problems also are the cause of death among these patients. Heart failure and malignancy have been identified as the most frequent long-term causes of death in IE patients [2]. Conversely, most of the epidemiological studies addressing this area have been carried out as case series, single-centre studies or with regional data collection [2–4, 10]. Using the Danish Nationwide Administrative Registries, we examined long-term mortality rates and differences in causes of death in IE patients discharged while still alive and stratified in patients who received heart valve surgery and in patients who were treated with medical therapy only. METHODS Data sources In Denmark, every resident is provided with a unique identification number. This makes possible to identify Danish citizens in a range of nationwide administrative registries. We used the following registries: the Danish Population Registry, the National Patient Registry, the Danish Prescription Registry and the Danish Cause of Death Registry. The Danish Population Registry holds information on date of birth, migration status and sex [11]. The National Patient Registry was initiated in 1977 and holds information on every patient admitted to a Danish hospital. A physician registers the diagnosis code of the disease for each admission according to the International Classification of Diseases (ICD)-8 until 1994 and the ICD-10 after 1994 [12]. The National Patient Registry was used to define the study population and baseline comorbidity [13]. Furthermore, since 1996, surgical procedures have been recorded in this registry according to the NOMESCO classification. Codes on treatment and medical examinations (including pacemaker implantation) were added to the National Patient Registry in 2000. The Danish Prescription Registry was established in 1994 and holds information on every prescription redeemed from all Danish pharmacies [14]. Prescriptions 6 months prior to the index date were used to assess baseline pharmacotherapy of the study population. The Anatomic Therapeutic Chemical Classification System is used to register the type of drug dispensed from the pharmacies. The Danish Cause of Death registry has been computerized since 1970. The primary and secondary causes of death are registered by the physician, thus the registry relies on the coding done by the individual physicians. The autopsy rate in Denmark is below 10%, and the automated classification of medical entities has been used since 2002 [15]. Study population Patients discharged alive after admission to a hospital with 1st time IE in the period from 1 January 1996 to 31 December 2014 were enrolled. We used the following ICD-8 and ICD-10 diagnosis codes to assess patients: 421, DI33, DI38 and DI39.8. In Denmark, all IE treatments are carried out in-hospital. Only patients with a hospitalization of >14 days were considered as having IE. The study population was categorized into (i) patients not undergoing heart valve surgery (medical therapy only) during IE hospital admission and (ii) patients undergoing heart valve surgery during IE hospital admission (for surgical procedure codes, see the Supplementary Material, Table S1). Patients undergoing valve surgery during follow-ups were assessed and analysed according to the treatment strategy (medical or medical + surgery) at the initial IE admission. Outcome The primary outcome was all-cause mortality. Danish death certificates provide opportunities for the physician to indicate multiple causes of death and the continuum of the fatal course based on the ICD-10 classification system. To assess cardiovascular (CV) death, the primary cause and secondary cause of death as recorded in the Cause of Death Registry were used (defined as ICD-10 codes of DI). If no CV death was registered as the primary cause or secondary cause of death, the primary cause of death was assessed to examine the non-CV cause of death. We categorized non-CV causes of death according to the disease groups of the ICD-10 classification system. Because IE patients are at high risk of heart failure, relapse and recurrence of IE and stroke in the course following the primary IE [7, 16, 17], we examined heart failure, IE and stroke as specific prespecified causes of death for the 2 groups. To do so, we used the primary and secondary causes of death. Statistical analysis Patients discharged alive were included on the day of discharge and were followed up until the end of the study period, death or emigration, whichever came first. Categorical variables are presented in counts and percentages, and continuously variables are presented with a median and 25 and 75 percentiles. The mortality rate for the 2 study groups were presented in a Kaplan–Meier plot. We used the log-rank test to test for difference between curves. Cumulative incidence plots considering competing risks of other death causes were used to illustrate differences in specific prespecified contributory causes of death (heart failure, IE and stroke), and the Fine and Gray test was used to test for differences between curves. The multivariable Cox proportional hazard analyses were used to examine the associated risk of dying from heart failure, IE or stroke among patients treated with surgery when compared with patients treated with medical therapy only. The following covariates were included in the multivariable models: age, calendar year, sex, cardiac implantable electronic device, cardiogenic shock, ischaemic heart disease, cerebrovascular disease, chronic obstructive lung disease, aortic stenosis, mitral regurgitation, renal disease, rheumatic disease, cancer, beta-blockade treatment, vitamin K-antagonist treatment, glucose lowering medication, corticosteroids and renin-angiotensin system inhibition. The proportional hazard assumption was tested, and when not fulfilled, follow-up time was split accordingly. The effect modification of sex on outcomes and age on outcomes were tested, as well as linearity of continuous variables. Results were presented with a hazard ratio (HR) and a 95% confidence interval (CI). A P-value <0.05 was considered statistically significant. All statistical analyses were performed using the SAS statistical software, version 9.4 (SAS Institute, Inc., Cary, NC, USA). For purposes of sensitivity, we matched the 2 study groups with sex- and age-matched controls from the background population to investigate differences in mortality rates and causes of death from the background population. A maximum 1-year difference in age was accepted between the study patients and matched controls. The greedy match algorithm was used for matching [18]. As a sensitivity analysis, we applied ischaemic heart disease, heart failure, cerebrovascular disease, chronic obstructive lung disease and cancer as matching criteria for the background population to investigate differences in mortality in a more comparable matched background population. In a landmark analysis, descriptive statistics was used to assess causes of death for patients surviving beyond 1 year of IE hospital discharge. As a sensitivity analysis, in-hospital mortality rate was identified by the 2 study groups. Further, baseline characteristics between groups were compared 5 years after index date. Ethics Register-based studies do not require ethical approval in Denmark. This study was approved by the Danish Data Protection Agency. RESULTS We identified 5576 patients with IE who were discharged alive; 4220 (75.7%) patients were treated with medical therapy only, and 1356 (24.3%) patients were treated with surgery during the 1st time IE hospitalization (Fig. 1). Baseline characteristics for the 2 study groups are presented in Table 1, and it can be observed that patients treated with medical therapy only were older and had overall more comorbidities. The median follow-up was 3.1 years (25 and 75 percentiles: 1.0 and 7.0 years, respectively) in the medically treated group and 5.3 years (25 and 75 percentiles 1.0 and 7.0 years, respectively) in the surgery group. Table 1: Baseline patient characteristics Medical therapy only Surgery P-value Demographics  Number 4220 1356  Male (%) 63.9 74.6 <0.0001  Age (years), median (25 and 75 percentiles) 70.0 (57.1–79.0) 60.7 (49.9–69.1) <0.0001 In-hospital  Duration of hospital admission (days), median (25 and 75 percentiles) 43.0 (32.0–53.0) 50.0 (42.0–66.0) <0.0001  Implantation of prosthetic valve (%) 91.0   Aortic prosthesisa (%) 53.2   Mitral prosthesisa (%) 27.4   Mitral repaira (%) 9.2 Comorbidity, medical history of  IHD (%) 28.6 15.1 <0.0001  Heart failure (%) 21.0 9.9 <0.0001  CVD (%) 13.7 7.3 <0.0001  COLD (%) 11.1 5.8 <0.0001  Cardiac implantable electronic device (%) 10.2 2.2 <0.0001  Aortic stenosis (%) 18.7 17.2 0.22  Mitral regurgitation (%) 7.9 7.5 0.68  Renal disease (%) 9.3 6.3 0.0005  Rheumatic disease (%) 8.1 4.7 <0.0001  Cancer (%) 15.3 8.5 <0.0001  Heart valve surgery (%) 14.6 8.4 <0.0001  Prosthetic valve (%) 12.5 6.9 <0.0001   Aortic prosthesis (%) 10.6 5.8   Mitral prosthesis (%) 1.5 1.0   Pulmonary prosthesis (%) 0.4 0.0 Prehospital medication  Beta-blocker (%) 29.9 18.7 <0.0001  VKA (%) 18.3 10.3 <0.0001  Glucose lowering medication (%) 8.8 5.8 0.0004  Corticosteroids (%) 11.4 7.2 <0.0001  Lipid lowering medication (%) 26.0 18.0 <0.0001  RAS inhibition (%) 32.8 26.6 <0.0001 Medical therapy only Surgery P-value Demographics  Number 4220 1356  Male (%) 63.9 74.6 <0.0001  Age (years), median (25 and 75 percentiles) 70.0 (57.1–79.0) 60.7 (49.9–69.1) <0.0001 In-hospital  Duration of hospital admission (days), median (25 and 75 percentiles) 43.0 (32.0–53.0) 50.0 (42.0–66.0) <0.0001  Implantation of prosthetic valve (%) 91.0   Aortic prosthesisa (%) 53.2   Mitral prosthesisa (%) 27.4   Mitral repaira (%) 9.2 Comorbidity, medical history of  IHD (%) 28.6 15.1 <0.0001  Heart failure (%) 21.0 9.9 <0.0001  CVD (%) 13.7 7.3 <0.0001  COLD (%) 11.1 5.8 <0.0001  Cardiac implantable electronic device (%) 10.2 2.2 <0.0001  Aortic stenosis (%) 18.7 17.2 0.22  Mitral regurgitation (%) 7.9 7.5 0.68  Renal disease (%) 9.3 6.3 0.0005  Rheumatic disease (%) 8.1 4.7 <0.0001  Cancer (%) 15.3 8.5 <0.0001  Heart valve surgery (%) 14.6 8.4 <0.0001  Prosthetic valve (%) 12.5 6.9 <0.0001   Aortic prosthesis (%) 10.6 5.8   Mitral prosthesis (%) 1.5 1.0   Pulmonary prosthesis (%) 0.4 0.0 Prehospital medication  Beta-blocker (%) 29.9 18.7 <0.0001  VKA (%) 18.3 10.3 <0.0001  Glucose lowering medication (%) 8.8 5.8 0.0004  Corticosteroids (%) 11.4 7.2 <0.0001  Lipid lowering medication (%) 26.0 18.0 <0.0001  RAS inhibition (%) 32.8 26.6 <0.0001 a Proportion calculated from ‘implantation of prosthetic valve’. COLD: chronic obstructive lung disease; CVD: cerebrovascular disease; IHD: ischaemic heart disease; RAS: renin-angiotensin system; VKA: vitamin K antagonists. Table 1: Baseline patient characteristics Medical therapy only Surgery P-value Demographics  Number 4220 1356  Male (%) 63.9 74.6 <0.0001  Age (years), median (25 and 75 percentiles) 70.0 (57.1–79.0) 60.7 (49.9–69.1) <0.0001 In-hospital  Duration of hospital admission (days), median (25 and 75 percentiles) 43.0 (32.0–53.0) 50.0 (42.0–66.0) <0.0001  Implantation of prosthetic valve (%) 91.0   Aortic prosthesisa (%) 53.2   Mitral prosthesisa (%) 27.4   Mitral repaira (%) 9.2 Comorbidity, medical history of  IHD (%) 28.6 15.1 <0.0001  Heart failure (%) 21.0 9.9 <0.0001  CVD (%) 13.7 7.3 <0.0001  COLD (%) 11.1 5.8 <0.0001  Cardiac implantable electronic device (%) 10.2 2.2 <0.0001  Aortic stenosis (%) 18.7 17.2 0.22  Mitral regurgitation (%) 7.9 7.5 0.68  Renal disease (%) 9.3 6.3 0.0005  Rheumatic disease (%) 8.1 4.7 <0.0001  Cancer (%) 15.3 8.5 <0.0001  Heart valve surgery (%) 14.6 8.4 <0.0001  Prosthetic valve (%) 12.5 6.9 <0.0001   Aortic prosthesis (%) 10.6 5.8   Mitral prosthesis (%) 1.5 1.0   Pulmonary prosthesis (%) 0.4 0.0 Prehospital medication  Beta-blocker (%) 29.9 18.7 <0.0001  VKA (%) 18.3 10.3 <0.0001  Glucose lowering medication (%) 8.8 5.8 0.0004  Corticosteroids (%) 11.4 7.2 <0.0001  Lipid lowering medication (%) 26.0 18.0 <0.0001  RAS inhibition (%) 32.8 26.6 <0.0001 Medical therapy only Surgery P-value Demographics  Number 4220 1356  Male (%) 63.9 74.6 <0.0001  Age (years), median (25 and 75 percentiles) 70.0 (57.1–79.0) 60.7 (49.9–69.1) <0.0001 In-hospital  Duration of hospital admission (days), median (25 and 75 percentiles) 43.0 (32.0–53.0) 50.0 (42.0–66.0) <0.0001  Implantation of prosthetic valve (%) 91.0   Aortic prosthesisa (%) 53.2   Mitral prosthesisa (%) 27.4   Mitral repaira (%) 9.2 Comorbidity, medical history of  IHD (%) 28.6 15.1 <0.0001  Heart failure (%) 21.0 9.9 <0.0001  CVD (%) 13.7 7.3 <0.0001  COLD (%) 11.1 5.8 <0.0001  Cardiac implantable electronic device (%) 10.2 2.2 <0.0001  Aortic stenosis (%) 18.7 17.2 0.22  Mitral regurgitation (%) 7.9 7.5 0.68  Renal disease (%) 9.3 6.3 0.0005  Rheumatic disease (%) 8.1 4.7 <0.0001  Cancer (%) 15.3 8.5 <0.0001  Heart valve surgery (%) 14.6 8.4 <0.0001  Prosthetic valve (%) 12.5 6.9 <0.0001   Aortic prosthesis (%) 10.6 5.8   Mitral prosthesis (%) 1.5 1.0   Pulmonary prosthesis (%) 0.4 0.0 Prehospital medication  Beta-blocker (%) 29.9 18.7 <0.0001  VKA (%) 18.3 10.3 <0.0001  Glucose lowering medication (%) 8.8 5.8 0.0004  Corticosteroids (%) 11.4 7.2 <0.0001  Lipid lowering medication (%) 26.0 18.0 <0.0001  RAS inhibition (%) 32.8 26.6 <0.0001 a Proportion calculated from ‘implantation of prosthetic valve’. COLD: chronic obstructive lung disease; CVD: cerebrovascular disease; IHD: ischaemic heart disease; RAS: renin-angiotensin system; VKA: vitamin K antagonists. Figure 1: View largeDownload slide The selection process of the study population is shown. In total, 17 patients died without a registered cause of death. CoD: cause of death; FU: follow-up; IE: infective endocarditis. Figure 1: View largeDownload slide The selection process of the study population is shown. In total, 17 patients died without a registered cause of death. CoD: cause of death; FU: follow-up; IE: infective endocarditis. During follow-up, heart valve surgery was observed in 10.4% of patients treated with medicine only during hospitalization for IE and in 6.1% of patients in the IE surgery group. Within the 1st year of follow-up, 54.9% of patients treated with medical therapy only and 42.2% of patients who underwent surgery at IE admission had undergone heart valve surgery. Within the first 5 years of follow-up, these numbers were increased to 84.7% and 73.5% for patients treated with medical therapy only and patients who underwent surgery during IE admission, respectively. Of the patients categorized as medically treated patients, 418 (9.9%) had a pacemaker vs 2.1% among patients categorized as surgically treated during IE admission. Causes of death Of the total study population, 2544 patients died and had a registered cause of death, see Fig. 1 [17 (0.7%) patients had no registered cause of death]: 2122 (50.3%) patients in the medically treated group and 422 (31.1%) patients in the surgery group. Throughout the study period, the mortality rate was higher in IE patients treated only with medicine in-hospital when compared with IE patients treated with surgery (Fig. 2). Causes of death were similar between groups with CV disease as the most frequent cause of death followed by neoplasms for both study groups (Table 2). Table 3 illustrates the specific prespecified causes of death (heart failure, IE and stroke) for both the study groups. Table 2: Cause of death in infective endocarditis patients discharged alive in the period from 1996 to 2014 Medical therapy only Surgery Number 2122 422 Diseases of the circulatory system (%) 52.5 55.2 Neoplasms (%) 16.6 19.2 Diseases of the respiratory system (%) 5.2 3.8 Infection (%) 5.7 3.6 Endocrine, nutritional and metabolic disorders (%) 3.4 3.6 Diseases of the digestive system (%) 4.4 3.6 Diseases of the genitourinary system (%) 2.5 1.4 Injury, poisoning and external causes (%) 2.7 3.8 Unknown 2.6 3.3 Mental and behavioural disorders (%) 1.6 0.7 Diseases of the nervous system (%) 1.4 1.0 Diseases of the musculoskeletal system and connective tissue (%) 0.8 0.7 Congenital disease 0.4 0.0 Diseases of the skin (%) 0.2 0.2 Medical therapy only Surgery Number 2122 422 Diseases of the circulatory system (%) 52.5 55.2 Neoplasms (%) 16.6 19.2 Diseases of the respiratory system (%) 5.2 3.8 Infection (%) 5.7 3.6 Endocrine, nutritional and metabolic disorders (%) 3.4 3.6 Diseases of the digestive system (%) 4.4 3.6 Diseases of the genitourinary system (%) 2.5 1.4 Injury, poisoning and external causes (%) 2.7 3.8 Unknown 2.6 3.3 Mental and behavioural disorders (%) 1.6 0.7 Diseases of the nervous system (%) 1.4 1.0 Diseases of the musculoskeletal system and connective tissue (%) 0.8 0.7 Congenital disease 0.4 0.0 Diseases of the skin (%) 0.2 0.2 Disease categories are from the International Classification of Diseases-10. Table 2: Cause of death in infective endocarditis patients discharged alive in the period from 1996 to 2014 Medical therapy only Surgery Number 2122 422 Diseases of the circulatory system (%) 52.5 55.2 Neoplasms (%) 16.6 19.2 Diseases of the respiratory system (%) 5.2 3.8 Infection (%) 5.7 3.6 Endocrine, nutritional and metabolic disorders (%) 3.4 3.6 Diseases of the digestive system (%) 4.4 3.6 Diseases of the genitourinary system (%) 2.5 1.4 Injury, poisoning and external causes (%) 2.7 3.8 Unknown 2.6 3.3 Mental and behavioural disorders (%) 1.6 0.7 Diseases of the nervous system (%) 1.4 1.0 Diseases of the musculoskeletal system and connective tissue (%) 0.8 0.7 Congenital disease 0.4 0.0 Diseases of the skin (%) 0.2 0.2 Medical therapy only Surgery Number 2122 422 Diseases of the circulatory system (%) 52.5 55.2 Neoplasms (%) 16.6 19.2 Diseases of the respiratory system (%) 5.2 3.8 Infection (%) 5.7 3.6 Endocrine, nutritional and metabolic disorders (%) 3.4 3.6 Diseases of the digestive system (%) 4.4 3.6 Diseases of the genitourinary system (%) 2.5 1.4 Injury, poisoning and external causes (%) 2.7 3.8 Unknown 2.6 3.3 Mental and behavioural disorders (%) 1.6 0.7 Diseases of the nervous system (%) 1.4 1.0 Diseases of the musculoskeletal system and connective tissue (%) 0.8 0.7 Congenital disease 0.4 0.0 Diseases of the skin (%) 0.2 0.2 Disease categories are from the International Classification of Diseases-10. Table 3: Specific prespecified contributory causes of death by the 2 study groups Medical therapy only Surgery Number 2122 422 Heart failure (%) 13.2 10.4 Infective endocarditis (%) 8.9 11.1 Stroke (%) 9.3 10.4 Medical therapy only Surgery Number 2122 422 Heart failure (%) 13.2 10.4 Infective endocarditis (%) 8.9 11.1 Stroke (%) 9.3 10.4 Table 3: Specific prespecified contributory causes of death by the 2 study groups Medical therapy only Surgery Number 2122 422 Heart failure (%) 13.2 10.4 Infective endocarditis (%) 8.9 11.1 Stroke (%) 9.3 10.4 Medical therapy only Surgery Number 2122 422 Heart failure (%) 13.2 10.4 Infective endocarditis (%) 8.9 11.1 Stroke (%) 9.3 10.4 Figure 2: View largeDownload slide The mortality rates of patients who underwent medical therapy only during the hospital stay for infective endocarditis and patients who underwent heart valve surgery during the hospital stay for infective endocarditis are shown. P-value tests for difference between curves using the log-rank test. Figure 2: View largeDownload slide The mortality rates of patients who underwent medical therapy only during the hospital stay for infective endocarditis and patients who underwent heart valve surgery during the hospital stay for infective endocarditis are shown. P-value tests for difference between curves using the log-rank test. Patients undergoing surgery during follow-up In patients undergoing surgery during follow-up [437 (10.4%) patients from the medically treated group and 83 (6.1%) patients from the surgery group], the mortality rate was 3.4% and 17.6% at 1- and 5-year follow-up, respectively, in patients treated with medical therapy only during admission, whereas this rate was 6.0% and 17.4% at 1- and 5-year follow-up, respectively, in the patients undergoing surgery (P = 0.62). At 1-year follow-up, CV disease was the most common cause of death (80.0%) followed by neoplasms (13.3%) and infection (6.7%) for the medically treated group (n = 15), whereas this was CV disease (80.0%) and neurological disorder (20.0%) for the surgery group (n = 5). At 5-year follow-up, CV disease was the most common cause of death (70.4%) followed by neoplasms (19.7%) and infection (2.8%) for the medically treated group (n = 71), whereas this was CV disease (78.6%) and other (neurological disorder, external causes, unknown, 21.4%) (n = 14) for the surgery group. Landmark analysis We identified 4725 patients surviving beyond 1 year after IE hospitalization: 73.1% of the patients treated with medical therapy only during IE admission and 26.9% had received surgery. With a total of 2 and 5 years of follow-up, the mortality rate was 9.9% and 33.7% in the medically treated patients, respectively, whereas this was 4.1% and 14.7% for the surgery group, respectively. The most frequent causes of death at 5-year follow-up were CV disease (50.5%), neoplasms (16.7%) and infection (5.7%) for the medically treated group, whereas this was CV disease (60.1%), neoplasms (15.5%) and infection (3.4%) for the surgery group. Differences in specific prespecified contributory causes of death Figure 3 shows cumulative incidence plots for the specific prespecified causes of death (heart failure, IE and stroke) of the 2 study groups up to 10 years of follow-up. Patients undergoing surgery were associated with a lower risk of dying due to heart failure, HR = 0.67 (95% CI: 0.47–0.95), when compared with patients treated with medical therapy only (Fig. 4). We found no difference in the associated risk of dying from IE, HR = 0.99 (95% CI: 0.70–1.40), with the medically treated group as reference (Fig. 4). The proportional hazard assumption was not fulfilled for dying from stroke up to 10 years of follow-up. When considering stroke for the first 5 years of follow-up, patients undergoing surgery were associated with a lower risk of dying from stroke when compared with the medically treated patients, HR = 0.59 (95% CI: 0.37–0.96). In contrast, at the follow-up period of 5–10 years, the patients in the surgery group were associated with an increased risk of dying from stroke when compared with medically treated patients, HR = 2.05 (95% CI 1.13–3.71), Fig. 4. Baseline characteristics are compared between study groups 5 years after index, and it is observed that the comorbidities of the groups become more similar during follow-up, see Supplementary Material, Table S3. Age modified the associated risk of dying from heart failure, P = 0.04 for interaction. Age did not modify the associated risk of dying from IE and stroke, P = 0.06 and P = 0.1 for interaction, respectively. Sex did not modify the associated risk of dying from heart failure, IE and stroke, P = 0.97, P = 0.68 and P = 0.27 for interaction, respectively. Figure 3: View largeDownload slide Specific prespecified contributory causes of death. The cumulative incidence of cause of death from (A) heart failure, (B) infective endocarditis and (C) stroke for the 2 study groups. P-value tests for difference between curves using Gray’s test. Figure 3: View largeDownload slide Specific prespecified contributory causes of death. The cumulative incidence of cause of death from (A) heart failure, (B) infective endocarditis and (C) stroke for the 2 study groups. P-value tests for difference between curves using Gray’s test. Figure 4: View largeDownload slide Associated risk of dying from specific prespecified contributory causes of death for patients treated with surgical intervention during IE admission when compared with patients treated with medical therapy only. *Follow-up from 0 to 5 years. **Follow-up from 5 to 10 years. CI: cumulative incidence; FU: follow-up; HR: hazard ratio. Figure 4: View largeDownload slide Associated risk of dying from specific prespecified contributory causes of death for patients treated with surgical intervention during IE admission when compared with patients treated with medical therapy only. *Follow-up from 0 to 5 years. **Follow-up from 5 to 10 years. CI: cumulative incidence; FU: follow-up; HR: hazard ratio. Sensitivity analysis: a comparison between background population and in-hospital mortality rate Each study group was compared with sex- and age-matched controls from the background population (see Supplementary Material, Table S2 for baseline characteristics). Controls for the medically treated group had a mortality rate of 3.3%, 16.3% and 31.6% after 1, 5 and 10 years of follow-up, respectively (P < 0.0001 when compared with medically treated IE patients). Controls for the surgically treated group had a mortality rate of 1.7%, 8.3% and 18.6% at 1, 5 and 10 years of follow-up, respectively (P < 0.0001 when compared with surgically treated IE patients). The most common causes of death over the 10-year study period were CV diseases (40.3%), neoplasms (23.9%), respiratory disorders (6.2%) and other (29.6%) for the controls of the medically treated group, whereas for the controls of the surgically treated group, this was CV disease (35.6%), neoplasms (33.5%), respiratory disorders (6.4%) and other (24.5%). No major differences in the causes of death were observed when we conducted a sensitivity analysis with very strict matching criteria for the matched background population as shown in the Supplementary Material, Table S4a and b. When including patients who died during IE hospitalization, we identified a mortality rate of 23.4% among patients treated with medical therapy only and 12.6% among patients in the surgical group, see Supplementary Material, Table S5. DISCUSSION This study investigated differences in long-term mortality and causes of death in surviving IE patients treated with medical therapy only and patients treated with surgery during IE admission. The study had 3 major findings. First, over a 10-year follow-up, 63.1% died in the medically treated group, whereas 41.6% died in the surgery group. Second, the most common cause of death was CV disease in both groups. For non-CV causes of death, neoplasms were the most common in both groups. Third, differences were found between groups when looking at specific prespecified causes of death (heart failure, IE and stroke) where patients from the surgery group had a lower associated risk of dying from heart failure and stroke when compared with patients in the medically treated group. Several studies have shown differences in outcomes in patients undergoing medical therapy only and surgical treatment during IE admission [5–7, 19, 20]. Reasons for this difference are due to differences in patient characteristics and the severity of the IE case and less likely the intervention itself. Patients with valve destruction leading to heart failure, uncontrolled infection or large vegetations with high risk of emboli should undergo surgery if the patient has a low burden of comorbidity. On the other hand, patients treated with medical therapy only constitute a patient group with IE without the need for surgical treatment or a patient group with a need for surgical treatment where surgery is avoided due to a high burden of comorbidities. Hence, different approaches may be needed in the follow-up of IE survivors to secure patient care at a high level. Our study found no differences in the relative frequencies of the major causes of death between patients undergoing medical therapy only and patients undergoing surgery. This illustrates the need to address diseases known to be frequent causes of death in the background population in IE patients as well [21, 22]. Our study adds valuable knowledge to the clinical awareness of IE survivors. Furthermore, our results justify the close monitoring of patients surviving IE, which has also been suggested by other authors [4]. These findings are of considerable clinical interest as little attention has been given to IE survivors. To ensure the best prevention strategy, it is of importance to acknowledge the diseases that are the most frequent causes of death among these patients. The differences in patient characteristics are reflected in the substantial differences in the mortality rate found in our study (61.3% vs 41.6% over a 10-year study period for the medically treated group and the surgery group, respectively). In line with the results of our study, several other studies have found an excess mortality in IE survivors when compared with sex- and age-matched controls [2, 4]; however, up until now, it has been unknown as to which disease category is the most frequent cause of death. A Dutch study investigating long-term follow-up in 191 patients with IE found a mortality rate of 41% after 10-year follow-up and found the major causes of death to be congestive heart failure and malignancy [2]. A German study identified a mortality rate of 42.8% after 12 years of follow-up in IE patients who underwent valve surgery where long-term cause of death primarily was of extracardiac origin [10]. The mortality rate identified in the German study is in line with our results (41.6%); however, our data show that the most common cause of death was a CV disease rather than an extracardiac origin. An explanation for this difference may be found in the different populations studied. Our study is strengthened by investigating a nationwide cohort with a large sample size where very few patients emigrated during follow-up. We found that patients undergoing surgery for IE were associated with a lower risk of dying from heart failure and stroke when compared with patients treated with medical therapy only. A more complex clinical picture with renal failure, sepsis and surgical contraindications in the medically treated group may be the reasons for these findings. A large population-based study from Taiwan showed that IE survivors carried a significantly higher long-term risk of major adverse cardiac events when compared with matched controls from the background population [23]. These findings along with the results of our study show a need for increased clinical awareness and patient care in IE survivors. Limitations Our study has several limitations. First, our results are found in the Danish Cause of Death Registry where the cause of death is assessed by a single physician, and the autopsy rate is less than 10% [15]. Validation of such a registry is difficult; however, a validation study of the Danish Cause of Death registry identified a sensitivity of the myocardial infarction diagnosis at 92.8% [24]. Second, clinical characteristics during IE admission were not available. Important characteristics such as microbiological aetiology, echocardiographic findings and differentiation of left- or right-sided IE or prosthetic valve endocarditis would have given a more complete picture of the study population as it is known that Staphylococcus aureus IE is associated with worse outcomes [25]. Furthermore, data on antimicrobial therapy, antimicrobial resistance and duration of antimicrobial therapy were not accessible through the administrative registries. Third, in view of the long-term follow-up, time-dependent confounding factors may have influenced the results. CONCLUSION In conclusion, no major differences were found in the relative frequencies of CV and non-CV cause of death between patients treated with medical therapy only and patients treated with surgery. Patients undergoing surgical treatment for IE were associated with a lower risk of dying from heart failure and stroke when compared with patients treated medically only. No difference was found between surgically and medically treated patients in the associated risk of dying from IE subsequent to index hospitalization. SUPPLEMENTARY MATERIAL Supplementary material is available at EJCTS online. Conflict of interest: none declared. REFERENCES 1 Cresti A , Chiavarelli M , Scalese M , Nencioni C , Valentini S , Guerrini F et al. Epidemiological and mortality trends in infective endocarditis, a 17-year population-based prospective study . Cardiovasc Diagn Ther 2017 ; 7 : 27 – 35 . Google Scholar CrossRef Search ADS PubMed 2 Mokhles MM , Ciampichetti I , Domburg R , van Cheng JM , Bogers AJJC , Witsenburg M. Infective endocarditis in a tertiary referral hospital: long-term follow up . J Heart Valve Dis 2012 ; 21 : 118 – 24 . Google Scholar PubMed 3 Mokhles MM , Ciampichetti I , Head SJ , Takkenberg JJ , Bogers AJ. Survival of surgically treated infective endocarditis: a comparison with the general Dutch population . Ann Thorac Surg 2011 ; 91 : 1407 – 12 . Google Scholar CrossRef Search ADS PubMed 4 Thuny F , Giorgi R , Habachi R , Ansaldi S , Le Dolley Y , Casalta J-P et al. Excess mortality and morbidity in patients surviving infective endocarditis . Am Heart J 2012 ; 164 : 94 – 101 . Google Scholar CrossRef Search ADS PubMed 5 Ternhag A , Cederström A , Törner A , Westling K. A nationwide cohort study of mortality risk and long-term prognosis in infective endocarditis in Sweden . PLoS One 2013 ; 8 : e67519 . Google Scholar CrossRef Search ADS PubMed 6 Delahaye F , Alla F , Béguinot I , Bruneval P , Doco-Lecompte T , Lacassin F et al. In-hospital mortality of infective endocarditis: prognostic factors and evolution over an 8-year period . Scand J Infect Dis 2007 ; 39 : 849 – 57 . Google Scholar CrossRef Search ADS PubMed 7 Kiefer T , Park L , Tribouilloy C , Cortes C , Casillo R , Chu V et al. Association between valvular surgery and mortality among patients with infective endocarditis complicated by heart failure . JAMA 2011 ; 306 : 2239 – 47 . Google Scholar CrossRef Search ADS PubMed 8 Thuny F , Avierinos J-F , Tribouilloy C , Giorgi R , Casalta J-P , Milandre L et al. Impact of cerebrovascular complications on mortality and neurologic outcome during infective endocarditis: a prospective multicentre study . Eur Heart J 2007 ; 28 : 1155 – 61 . Google Scholar CrossRef Search ADS PubMed 9 Tornos P , Iung B , Permanyer-Miralda G , Baron G , Delahaye F , Gohlke-Bärwolf C et al. Infective endocarditis in Europe: lessons from the Euro heart survey . Heart 2005 ; 91 : 571. Google Scholar CrossRef Search ADS PubMed 10 Spiliopoulos K , Giamouzis G , Haschemi A , Karangelis D , Antonopoulos N , Fink G et al. Surgical management of infective endocarditis: early and long-term mortality analysis. single-center experience and brief literature review . Hellenic J Cardiol 2014 ; 55 : 462 – 74 . Google Scholar PubMed 11 Schmidt M , Pedersen L , Sørensen HT. The Danish Civil Registration System as a tool in epidemiology . Eur J Epidemiol 2014 ; 29 : 541 – 9 . Google Scholar CrossRef Search ADS PubMed 12 Schmidt M , Schmidt SAJ , Sandegaard JL , Ehrenstein V , Pedersen L , Sørensen HT. The Danish National Patient Registry: a review of content, data quality, and research potential . Clin Epidemiol 2015 ; 7 : 449 – 90 . Google Scholar CrossRef Search ADS PubMed 13 Thygesen SK , Christiansen CF , Christensen S , Lash TL , Sørensen HT. The predictive value of ICD-10 diagnostic coding used to assess Charlson comorbidity index conditions in the population-based Danish National Registry of Patients . BMC Med Res Methodol 2011 ; 11 : 83 . Google Scholar CrossRef Search ADS PubMed 14 Kildemoes HW , Sørensen HT , Hallas J. The Danish National Prescription Registry . Scand J Public Health 2011 ; 39 : 38 – 41 . Google Scholar CrossRef Search ADS PubMed 15 Helweg-Larsen K. The Danish Register of causes of death . Scand J Public Health 2011 ; 39 : 26 – 9 . Google Scholar CrossRef Search ADS PubMed 16 Merkler AE , Chu SY , Lerario MP , Navi BB , Kamel H. Temporal relationship between infective endocarditis and stroke . Neurology 2015 ; 85 : 512 – 16 . Google Scholar CrossRef Search ADS PubMed 17 Alagna L , Park LP , Nicholson BP , Keiger AJ , Strahilevitz J , Morris A et al. Repeat endocarditis: analysis of risk factors based on the International Collaboration on Endocarditis—Prospective Cohort Study . Clin Microbiol Infect 2014 ; 20 : 566 – 75 . Google Scholar CrossRef Search ADS PubMed 18 Locally Written SAS Macros—Division of Biomedical Statistics and Informatics—Mayo Clinic Research. http://www.mayo.edu/research/departments-divisions/department-health-sciences-research/division-biomedical-statistics-informatics/software/locally-written-sas-macros (30 March 2018, date last accessed). 19 Bannay A , Hoen B , Duval X , Obadia J-F , Selton-Suty C , Le Moing V et al. The impact of valve surgery on short- and long-term mortality in left-sided infective endocarditis: do differences in methodological approaches explain previous conflicting results? Eur Heart J 2011 ; 32 : 2003 – 15 . Google Scholar CrossRef Search ADS PubMed 20 Heiro M , Helenius H , Hurme S , Savunen T , Metsärinne K , Engblom E et al. Long-term outcome of infective endocarditis: a study on patients surviving over one year after the initial episode treated in a Finnish teaching hospital during 25 years . BMC Infect Dis 2008 ; 8 : 49. Google Scholar CrossRef Search ADS PubMed 21 GBD 2013 Mortality and Causes of Death Collaborators . Global, regional, and national age–sex specific all-cause and cause-specific mortality for 240 causes of death, 1990–2013: a systematic analysis for the Global Burden of Disease Study 2013 . Lancet 2015 ; 385 : 117 – 71 . CrossRef Search ADS PubMed 22 Feigin VL , Forouzanfar MH , Krishnamurthi R , Mensah GA , Connor M , Bennett DA et al. Global and regional burden of stroke during 1990-2010: findings from the Global Burden of Disease Study 2010 . Lancet Lond Engl 2014 ; 383 : 245 – 54 . Google Scholar CrossRef Search ADS 23 Shih C-J , Chu H , Chao P-W , Lee Y-J , Kuo S-C , Li S-Y et al. Long-term clinical outcome of major adverse cardiac events in survivors of infective endocarditis: a nationwide population-based study . Circulation 2014 ; 130 : 1684 – 91 . Google Scholar CrossRef Search ADS PubMed 24 Madsen M , Davidsen M , Rasmussen S , Abildstrom SZ , Osler M. The validity of the diagnosis of acute myocardial infarction in routine statistics: a comparison of mortality and hospital discharge data with the Danish MONICA registry . J Clin Epidemiol 2003 ; 56 : 124 – 30 . Google Scholar CrossRef Search ADS PubMed 25 Chu VH , Cabell CH , Benjamin DK , Kuniholm EF , Fowler VG , Engemann J et al. Early predictors of in-hospital death in infective endocarditis . Circulation 2004 ; 109 : 1745 – 9 . Google Scholar CrossRef Search ADS PubMed © The Author(s) 2018. Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

European Journal of Cardio-Thoracic SurgeryOxford University Press

Published: Apr 10, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off