Benito Almirante

Epidemiology and Predictors of Mortality in Cases of Candida Bloodstream Infection: Results from Population-Based Surveillance, Barcelona, Spain, from 2002 to 2003

ABSTRACT We conducted population-based surveillance for Candida bloodstream infections in Spain to determine its incidence, the extent of antifungal resistance, and risk factors for mortality. A case was defined as the first positive blood culture for any Candida spp. in a resident of Barcelona, from 1 January 2002 to 31 December 2003. We defined early mortality as occurring between days 3 to 7 after candidemia and late mortality as occurring between days 8 to 30. We detected 345 cases of candidemia, for an average annual incidence of 4.3 cases/100,000 population, 0.53 cases/1,000 hospital discharges, and 0.73 cases/10,000 patient-days. Outpatients comprised 11% of the cases, and 89% had a central venous catheter (CVC) at diagnosis. Overall mortality was 44%. Candida albicans was the most frequent species (51% of cases), followed by Candida parapsilosis (23%), Candida tropicalis (10%), Candida glabrata (8%), Candida krusei (4%), and other species (3%). Twenty-four isolates (7%) had decreased susceptibility to fluconazole (MIC ≥ 16 μg/ml). On multivariable analysis, early death was independently associated with hematological malignancy (odds ratio [OR], 3.5; 95% confidence interval [CI], 1.1 to 10.4). Treatment with antifungals (OR, 0.05; 95% CI, 0.01 to 0.2) and removal of CVCs (OR, 0.3; 95% CI, 0.1 to 0.9) were protective factors for early death. Receiving adequate treatment, defined as having CVCs removed and administration of an antifungal medication (OR, 0.2; 95% CI, 0.08 to 0.8), was associated with lower odds of late mortality; intubation (OR, 7.5; 95% CI, 2.6 to 21.1) was associated with higher odds. The incidence of candidemia and prevalence of fluconazole resistance are similar to other European countries, indicating that routine antifungal susceptibility testing is not warranted. Antifungal medication and catheter removal are critical in preventing mortality.

A bedside scoring system ("Candida score") for early antifungal treatment in nonneutropenic critically ill patients with Candida colonization.

Objective:To obtain a score for deciding early antifungal treatment when candidal infection is suspected in nonneutropenic critically ill patients. Design:Analysis of data collected from the database of the EPCAN project, an ongoing prospective, cohort, observational, multicenter surveillance study of fungal infection and colonization in intensive care unit (ICU) patients. Setting:Seventy-three medical-surgical ICUs of 70 teaching hospitals in Spain. Patients:A total of 1,699 ICU patients aged 18 yrs and older admitted for at least 7 days between May 1998 and January 1999 were studied. Interventions:Surveillance cultures of urine, tracheal, and gastric samples were obtained weekly. Patients were grouped as follows: neither colonized nor infected (n = 719), unifocal or multifocal Candida colonization (n = 883), and proven candidal infection (n = 97). The odds ratio (OR) for each risk factor associated with colonization vs. proven candidal infection was estimated. A logistic regression model was performed to adjust for possible confounders. The “Candida score” was obtained according to the logit method. The discriminatory power was evaluated by the area under the receiver operating characteristics curve. Measurements and main results:In the logit model, surgery (OR = 2.71, 95% confidence interval [CI], 1.45–5.06); multifocal colonization (OR = 3.04, 95% CI, 1.45–6.39); total parenteral nutrition (OR = 2.48, 95% CI, 1.16–5.31); and severe sepsis (OR = 7.68, 95% CI, 4.14–14.22) were predictors of proven candidal infection. The “Candida score” for a cut-off value of 2.5 (sensitivity 81%, specificity 74%) was as follows: parenteral nutrition, +0.908; surgery, +0.997; multifocal colonization, +1.112; and severe sepsis, +2.038. Central venous catheters were not a significant risk factor for proven candidal infection (p = .292). Conclusions:In a large cohort of nonneutropenic critically ill patients in whom Candida colonization was prospectively assessed, a “Candida score” >2.5 accurately selected patients who would benefit from early antifungal treatment.

Community-Onset Bacteremia Due to Extended-Spectrum β-Lactamase-Producing Escherichia coli: Risk Factors and Prognosis

BACKGROUND There is little clinical information about community-onset bloodstream infections (COBSIs) caused by extended-spectrum beta-lactamase (ESBL)-producing Escherichia coli (ESBLEC). We investigated the prevalence and risk factors for COBSI due to ESBLEC, and described their clinical features and the impact of COBSI caused by ESBLEC on 14-day mortality. METHODS Risk factors were assessed using a multicenter case-control-control study. Influence of ESBL production on mortality was studied in all patients with COBSI due to E. coli. Isolates and ESBLs were microbiologically characterized. Statistical analysis was performed using multivariate logistic regression. Thirteen tertiary care Spanish hospitals participated in the study. RESULTS We included 95 case patients with COBSI due to ESBLEC, which accounted for 7.3% of all COBSI due to E. coli. The ESBL in 83 of these (87%) belonged to the CTX-M family of ESBL, and most were clonally unrelated. Comparison with both control groups disclosed association with health care (odds ratio [OR], 2.1; 95% confidence interval [CI], 1.2-3.8), urinary catheter use (OR, 3.1; 95% CI, 1.5-6.5), and previous antimicrobial use (OR, 2.7; 95% CI, 1.5-4.9) as independent risk factors for COBSI due to ESBLEC. Mortality among patients with COBSI due to ESBLEC was lower among patients who received empirical therapy with beta-lactam/beta-lactam inhibitor combinations or carbapenems (8%-12%) than among those receiving cephalosporins or fluoroquinolones (24% and 29%, respectively). Mortality among patients with COBSI due to E. coli was associated with inappropriate empirical therapy irrespective of ESBL production. CONCLUSIONS ESBLEC is an important cause of COBSI due to E. coli. Clinicians should consider adequate empirical therapy with coverage of these pathogens for patients with risk factors.

Value of differential quantitative blood cultures in the diagnosis of catheter-related sepsis

A prospective study was performed to assess the value of differential quantitative blood cultures in the diagnosis of catheter-related sepsis when this condition is suspected on clinical grounds and to establish a reliable discriminative value for application without removal of the inserted catheter. A total of 107 central venous catheters from 64 patients were used for the study. Blood was obtained simultaneously through the suspected infected device and from a peripheral venipuncture. The catheter was removed and its tip cultured semiquantitatively. Catheter-related sepsis occurred in 17 patients. Using as cut-off value a colony count fourfold higher in blood drawn through the catheter than in simultaneously drawn peripheral blood, a sensitivity of 94 %, specificity of 100 % and positive predictive value of 100 % were obtained. A single bacterial count > 100 cfu/ml in the quantitative culture of the catheter blood specimen in the presence of a positive qualitative peripheral blood culture of the same organism was also highly suggestive of catheter-related sepsis. Differential quantitative blood culture is a reliable method for the diagnosis of catheter-associated sepsis without catheter removal.

Epidemiology, risk factors, and prognosis of Candida parapsilosis bloodstream infections: case-control population-based surveillance study of patients in Barcelona, Spain, from 2002 to 2003.

ABSTRACT Candida parapsilosis has emerged as an important yeast species causing fungemia. We describe the incidence and epidemiology of C. parapsilosis fungemia. Data from active population-based surveillance in Barcelona, Spain, from January 2002 to December 2003 were analyzed. We focused on 78 episodes of C. parapsilosis fungemia, and we compared them with 175 Candida albicans controls. C. parapsilosis accounted for 23% of all fungemias. The annual incidences were 1 episode per 105 patients, 1.2 episodes per 104 discharges, and 1.7 episodes per 105 patient days. All isolates but one (99%) were fluconazole susceptible. Seventy-two isolates (92%) were inpatient candidemias. Forty-two episodes (51%) were considered catheter-related fungemia, 35 (45%) were considered primary fungemia, and 3 (4%) were considered secondary fungemia. Risk factors for candidemia were vascular catheterization (97%), prior antibiotic therapy (91%), parenteral nutrition (54%), prior surgery (46%), prior immunosuppressive therapy (38%), malignancy (27%), prior antifungal infection (26%), transplant recipient (16%), neutropenia (12%), and prior colonization (11%). Multivariate analysis of the differential characteristics showed that the factors that independently predicted the presence of C. parapsilosis fungemia were neonate patients (odds ratio [OR], 7.5; 95% confidence interval [CI], 2.1 to 26.8; P = 0.002), transplant recipients (OR, 9.2; 95% CI, 1.9 to 43.3; P = 0.005), patients with a history of prior antifungal therapy (OR, 5.4; 95% CI, 1.8 to 15.9; P = 0.002), and patients who received parenteral nutrition (OR, 2.2; 95% CI, 1.09 to 4.6; P = 0.028). The overall mortality rate was lower than that associated with C. albicans candidemia (23% versus 43%; P < 0.01). In summary, C. parapsilosis was responsible for 23% of all candidemias and was more frequent in neonates, in transplant recipients, and in patients who received parenteral nutrition or previous antifungal therapy, mainly fluconazole. The mortality rate was lower than that associated with C. albicans fungemia.

Effectiveness of Cloxacillin with and without Gentamicin in Short-Term Therapy for Right-Sided Staphylococcus aureus Endocarditis: A Randomized, Controlled Trial

Tricuspid valve endocarditis is a common infection in intravenous drug users and is caused by Staphylococcus aureus in more than 80% of cases [1-4]. Right-sided endocarditis is less aggressive than left-sided disease [5]. The prognosis for patients with isolated tricuspid valve endocarditis caused by S. aureus is generally favorable, and the condition promptly responds to medical therapy [6]. For this reason, and because of the difficulty of hospitalizing patients with right-sided endocarditis for 4 to 6 weeks [7], there has been considerable interest in short courses of intravenous treatment for managing this condition [8]. Several recent studies [9-11] have shown that a combination of a penicillinase-resistant penicillin and an aminoglycoside given for 2 weeks is useful in most patients with uncomplicated, right-sided S. aureus endocarditis. However, the efficacy of a penicillinase-resistant penicillin as single-agent therapy given for 2 weeks has not yet been evaluated. Thus, the benefit of adding an aminoglycoside to short-course regimens is only theoretical [8]. We did a randomized clinical trial to compare the efficacy of cloxacillin alone with that of cloxacillin plus gentamicin as short-course therapy for right-sided S. aureus endocarditis in intravenous drug users. Methods Patients All patients who were suspected of having right-sided staphylococcal endocarditis, who were hospitalized at our institution between March 1988 and February 1993, and who admitted that they used intravenous drugs were considered for inclusion in our study. Patients who presented with tricuspid endocarditis caused by S. aureus according to the criteria shown in Table 1 were eligible. The institutional review board of our hospital approved the study, and all patients gave informed consent. Table 1. Diagnostic Criteria for Tricuspid Valve Infective Endocarditis Patients were excluded if they had any of the following: allergy to study medications, infection with methicillin-resistant S. aureus, confirmed or suspected left-sided endocarditis shown by clinical examination (indicated by characteristic regurgitation murmur or systemic embolization) or echocardiography (indicated by vegetation or valvular insufficiency), vegetation on the pulmonic valves shown by echocardiography, systemic complications requiring prolonged therapy (osteomyelitis, abscesses that could not be drained surgically), or antibiotic therapy given for longer than 48 hours before hospitalization. Clinically significant hemodynamic compromise, extensive pulmonary embolism, and large tricuspid vegetations were not considered exclusion criteria. Study Design and Treatment Our study was an open, randomized trial. Consecutive patients were randomly assigned to receive cloxacillin, 2 g intravenously every 4 hours for 14 days, alone or combined with gentamicin, 1 mg/kg of body weight intravenously every 8 hours for the first 7 days. Randomization was done using a random-number table in sets of 10: In each consecutive set, 5 patients received cloxacillin alone and 5 received combination therapy. A sealed envelope labeled with the randomization number was opened at the start of treatment. The decision to start treatment was based on microbiological criteria (positive blood cultures) or clinical criteria (fever and evidence of pulmonary embolism or poor general condition attributable to the infection). Thus, some patients were randomly assigned to treatment before the results of some procedures (echocardiography and isolation of microorganism) that might have uncovered exclusion criteria were available. Follow-up Initial patient assessment included medical history, physical examination, chest radiography, radionuclide body scanning with technetium pyrophosphate, electrocardiography, and the following laboratory tests: complete blood count, standard blood chemistries, urinalysis, serologic testing for human immunodeficiency virus (HIV), and lymphocyte subpopulation counts (when the HIV test result was positive). Standard laboratory tests and chest radiography were done every week. Two-dimensional transthoracic echocardiography was done within 4 days of hospital admission and at the end of treatment. Blood cultures were done at the time of hospital admission, 3 days after the initiation of treatment, and 2 days after the end of treatment. Aerobic and anaerobic cultures were done by using the BACTEC NR 660 System (Becton Dickinson, Mountain View, California). Follow-up visits were scheduled at 2 weeks and at 1, 3, and 6 months after the end of therapy. At the 2-week visit, a physical examination, standard laboratory tests, chest radiography, and two blood cultures were done. Subsequent blood cultures were done only if evidence suggested infection. Nine patients did not attend a follow-up visit; outcome information for these patients was obtained by telephone. Evaluation Criteria In all patients who met the exclusion criteria after random assignment, treatment was modified accordingly. In the intention-to-treat analysis, these patients were considered to have had treatment failure. Three primary end points were considered to indicate treatment failure: 1) death during treatment, 2) continued clinical or microbiological evidence of active infection after 2 weeks of therapy, and 3) relapse of staphylococcal infection. Clinical recovery was defined as the disappearance of clinical evidence of infection and the absence of radiologic abnormalities (pleural effusion or active pulmonary abscesses) at 14 days of treatment. Bacteriologic recovery was defined as a negative blood culture obtained 48 hours after the end of treatment. Successful therapy was defined as clinical and microbiological recovery without subsequent relapse. Patients who showed signs of active infection after 14 days of treatment continued to receive cloxacillin for 2 more weeks. Relapses were treated with cloxacillin alone for 4 weeks. To differentiate between relapse and reinfection, we considered clinical criteria (resumption of drug use and duration of symptom-free interval after stopping treatment) and comparison of S. aureus isolates by phage-typing. Phage-typing was done at the National Reference Center of Microbiology, Virology and Immunology, Majadahonda, Madrid, by using the international set of phages applied sequentially at the standard test dilution and a 1:100 test dilution; typing was also done after heat treatment. Reverse phagotyping was done as described elsewhere [12]. The duration of fever and appearance of complications during treatment were also evaluated. In patients without previous renal dysfunction, acute renal insufficiency was defined as a serum creatinine level greater than 176.8 mol/L. In patients with previous renal dysfunction, renal insufficiency was defined as an increase in the serum creatinine level of more than 50%. Statistical Analysis We did two analyses. An intention-to-treat analysis was done for all randomly assigned patients, and an efficacy analysis was done after we excluded patients who were found, after treatment began, to have met an exclusion criterion. We used the Fisher exact test to compare categorical baseline characteristics and outcomes of the two groups, and we used the Student t-test or the Mann-Whitney U test to compare continuous characteristics and outcomes. In the primary efficacy analysis, we compared the rates of successful treatment. Because we assumed that treatment would be successful in 95% of patients who received combination therapy [9], we estimated that almost 38 evaluable patients had to receive each regimen in order for us to detect a difference in efficacy of 20% or more, with a power of 80% at a one-sided level of 0.05. Because we anticipated that 10% to 15% of patients might have to be excluded after random assignment, 45 patients were assigned to each group. Two-tailed P values were used for all calculations; thus, the power of the study to find no differences in the efficacy of the two regimens was 70%. Results Patients During the 5-year study period, 45 patients were randomly assigned to each treatment group. Of these 90 patients, 6 (13%) of those assigned to receive cloxacillin alone and 8 (18%) of those assigned to receive combination therapy were later excluded for the following reasons. Three patients (1 receiving cloxacillin alone and 2 receiving combination therapy) had endocarditis caused by Streptococcus viridans; 7 patients (3 receiving cloxacillin alone and 4 receiving combination therapy) had involvement of the mitral, aortic, or pulmonic valves; 3 patients (1 receiving cloxacillin alone and 2 receiving combination therapy) were allergic to penicillin or had methicillin-resistant S. aureus infection; and 1 patient receiving cloxacillin alone had osteomyelitis. Two patients (1 in each group) refused further in-hospital treatment 5 and 8 days after enrollment, respectively. Thus, 74 patients (38 receiving cloxacillin alone and 36 receiving combination therapy) remained available for the efficacy analysis. Clinical, radiologic, and echocardiographic findings and laboratory values were similar in the two groups. The exclusion of 16 patients from the efficacy analysis after randomization did not change these similarities (Table 2). Table 2. Selected Baseline Characteristics of 74 Patients Evaluable for the Efficacy Analysis* Efficacy of Therapy Table 3 shows the outcome of all patients randomly assigned to treatment, as determined by the intention-to-treat analysis. A successful overall outcome was seen in 34 of the 45 patients (76% [95% CI, 61% to 87%]) assigned to receive cloxacillin only and 31 of the 45 patients (69% [CI, 53% to 82%]) assigned to receive combination therapy (P > 0.2). Table 3. Outcome of All Randomly Assigned Patients according to Treatment Regimen (Intention-to-Treat Analysis) Table 4 shows the outcomes for the 74 patients that were available for the efficacy analysis. Treatment was successful in 34 of the

Invasive Pneumococcal Disease in Patients Infected with HIV: Still a Threat in the Era of Highly Active Antiretroviral Therapy

We studied all human immunodeficiency virus (HIV)-infected patients with invasive pneumococcal disease who received their diagnosis during 1996-2002 to investigate the incidence of this disease in the highly active antiretroviral therapy era and to study the influence of CD4 lymphocyte count on the clinical presentation and outcome of disease. The overall incidence of invasive pneumococcal disease was 11.3 cases per 100,000 person-years in adult patients without known HIV infection and 677 cases per 100,000 person-years in HIV-infected patients. This incidence remained stable over the study period. Clinical presentation, severity of illness, and number of recurrent episodes were similar in patients with CD4+ cell counts of >200 or < or =200 cells/ microL. Patients receiving trimethoprim-sulfamethoxazole (TMP-SMZ) were more likely to present with TMP-SMZ-resistant pneumococci than were those who were not receiving this agent (76.7% vs. 43.6%; P=.007). The mortality rate was high (21%).

Neurological Complications of Infective Endocarditis Risk Factors, Outcome, and Impact of Cardiac Surgery: A Multicenter Observational Study

Background— The purpose of this study was to assess the incidence of neurological complications in patients with infective endocarditis, the risk factors for their development, their influence on the clinical outcome, and the impact of cardiac surgery. Methods and Results— This was a retrospective analysis of prospectively collected data on a multicenter cohort of 1345 consecutive episodes of left-sided infective endocarditis from 8 centers in Spain. Cox regression models were developed to analyze variables predictive of neurological complications and associated mortality. Three hundred forty patients (25%) experienced such complications: 192 patients (14%) had ischemic events, 86 (6%) had encephalopathy/meningitis, 60 (4%) had hemorrhages, and 2 (1%) had brain abscesses. Independent risk factors associated with all neurological complications were vegetation size ≥3 cm (hazard ratio [HR] 1.91), Staphylococcus aureus as a cause (HR 2.47), mitral valve involvement (HR 1.29), and anticoagulant therapy (HR 1.31). This last variable was particularly related to a greater incidence of hemorrhagic events (HR 2.71). Overall mortality was 30%, and neurological complications had a negative impact on outcome (45% of deaths versus 24% in patients without these complications; P<0.01), although only moderate to severe ischemic stroke (HR 1.63) and brain hemorrhage (HR 1.73) were significantly associated with a poorer prognosis. Antimicrobial treatment reduced (by 33% to 75%) the risk of neurological complications. In patients with hemorrhage, mortality was higher when surgery was performed within 4 weeks of the hemorrhagic event (75% versus 40% in later surgery). Conclusions— Moderate to severe ischemic stroke and brain hemorrhage were found to have a significant negative impact on the outcome of infective endocarditis. Early appropriate antimicrobial treatment is critical, and transitory discontinuation of anticoagulant therapy should be considered.Background— The purpose of this study was to assess the incidence of neurological complications in patients with infective endocarditis, the risk factors for their development, their influence on the clinical outcome, and the impact of cardiac surgery. Methods and Results— This was a retrospective analysis of prospectively collected data on a multicenter cohort of 1345 consecutive episodes of left-sided infective endocarditis from 8 centers in Spain. Cox regression models were developed to analyze variables predictive of neurological complications and associated mortality. Three hundred forty patients (25%) experienced such complications: 192 patients (14%) had ischemic events, 86 (6%) had encephalopathy/meningitis, 60 (4%) had hemorrhages, and 2 (1%) had brain abscesses. Independent risk factors associated with all neurological complications were vegetation size ≥3 cm (hazard ratio [HR] 1.91), Staphylococcus aureus as a cause (HR 2.47), mitral valve involvement (HR 1.29), and anticoagulant therapy (HR 1.31). This last variable was particularly related to a greater incidence of hemorrhagic events (HR 2.71). Overall mortality was 30%, and neurological complications had a negative impact on outcome (45% of deaths versus 24% in patients without these complications; P <0.01), although only moderate to severe ischemic stroke (HR 1.63) and brain hemorrhage (HR 1.73) were significantly associated with a poorer prognosis. Antimicrobial treatment reduced (by 33% to 75%) the risk of neurological complications. In patients with hemorrhage, mortality was higher when surgery was performed within 4 weeks of the hemorrhagic event (75% versus 40% in later surgery). Conclusions— Moderate to severe ischemic stroke and brain hemorrhage were found to have a significant negative impact on the outcome of infective endocarditis. Early appropriate antimicrobial treatment is critical, and transitory discontinuation of anticoagulant therapy should be considered. # Clinical Perspective {#article-title-46}

Contemporary Epidemiology and Prognosis of Health Care–Associated Infective Endocarditis

BACKGROUND The aim of this study was to describe the characteristics of health care-associated infective endocarditis (HAIE) and to establish the risk factors for mortality. METHODS We conducted a prospective, observational cohort study. HAIE was defined according to the following conditions: (1) symptom onset >48 h after hospitalization or within 6 months after hospital discharge; or (2) ambulatory manipulations causing endocarditis. RESULTS Eighty-three episodes of HAIE (accounting for 28.4% of all cases of endocarditis) were diagnosed. Compared with patients with community-acquired endocarditis, patients with HAIE were older (median age +/- standard deviation, 65.3 +/- 16.4 years vs. 57.8 +/- 17.0 years; P = .001), were in poorer health before disease onset (Charlson index, 2.5 +/- 2.3 vs. 1.7 +/- 2.1; P = .006), had more staphylococcal (55.4% vs. 28.3% of cases) and enterococcal infections (22.9% vs. 7.7% of cases; P < .005), underwent fewer surgeries (22.9% vs. 45.9% of cases; P < .005), and experienced a higher rate of in-hospital (45.8% vs. 22.0%) and 1-year mortality (59.5% vs. 29.6%; P < .005). In the HAIE cohort, independent predictors of in-hospital death were stroke (odds ratio [OR], 8.95; 95% confidence interval [CI], 2.04-39.31; P = .004), congestive heart failure (OR, 5.48; 95% CI, 1.77-17.03; P = .003), surgery indicated but not performed (OR, 3.74; 95% CI, 1.22-11.45; P = .021), and enterococcal infection (OR, 0.18; 95% CI, 0.04-0.78; P = .022). Independent predictors of 1-year mortality were surgery indicated but not performed (OR, 7.81; 95% CI, 2.06-29.67; P = .003), acute renal failure (OR, 7.18; 95% CI, 1.32-39.18; P = .023), and enterococcal infection (OR, 0.18; 95% CI, 0.04-0.81; P = .026). For the series overall (292 episodes), HAIE was an independent predictor of in-hospital (OR, 2.83; 95% CI, 1.34-5.98; P = .007) and 1-year mortality (OR, 2.59; 95% CI, 1.25-5.39; P = .011). CONCLUSIONS HAIE is an important health problem associated with considerable mortality. New strategies to prevent HAIE should be assessed.

Long-Term Complications of Native Valve Infective Endocarditis in Non-Addicts: A 15-Year Follow-up Study

OBJECTIVE To describe the incidence and clinical manifestations of long-term cardiac complications of endocarditis. DESIGN Cohort study. SETTING University-affiliated tertiary medical center. PATIENTS One hundred twelve consecutive patients, survivors from a series of 140 non-addicted patients with a first episode of infective endocarditis on native valves hospitalized from 1975 to 1990. Thirty-two patients had had valve replacement during the active phase of the infection, and the remaining 80 patients received medical treatment alone. MEASUREMENTS Relapse, recurrence, need for late cardiac surgery, and cardiac mortality. RESULTS Relapses occurred in three patients (2.7%) and recurrences in five patients (4.5%, incidence density at 15 years, 0.0030 per patient-year). Late cardiac surgery was needed by 47% of the patients treated medically during the active phase, and most had surgery in the first 2 years of follow-up (incidence density, 0.25 per patient-year at 2 years). Aortic valve involvement (relative risk, 2.66; 95% CI, 1.15 to 6.17) and end-diastolic diameter greater than 60 mm (relative risk, 1.04; 95% CI, 1.03 to 2.43) were associated with the need for late surgery in univariate analysis. Multiple logistic regression analysis showed aortic valve involvement to be an independent predictor of the need for late surgery (relative risk, 3.04; CI, 1.23 to 7.54). Only 2 of the 32 patients who had surgery during the active infection needed a second operation during follow-up. At the end of follow-up, the number of patients who had surgery after the onset of the infection was 86 (60% of the whole series). Cardiac death occurred in 16 patients; most deaths were sudden or postoperative and occurred in the first 2 years of follow-up (incidence density, 0.047 per patient-year at 2 years). Independent predictors of death were not found. Survival was 90% at 2 years, 88% at 5 years, 81% at 10 years, and 61% at 15 years. CONCLUSIONS Survival after infective endocarditis is fair (81% probability of survival at 10 years), and the most common types of cardiac death are sudden and postoperative. Aortic valve involvement is an independent predictor of the need for late cardiac surgery. The rate of recurrences is not negligible (incidence density at 15 years, 0.0030 per patient-year).