Endocarditis

Study objectives: To study the changing etiology of prosthetic valve endocarditis (PVE) and the impact of nosocomial acquisition of the infection on prognosis in a single hospital.

Methods: Retrospective review of 121 cases of PVE during a period of 34 years. Two different periods (the period from 1970 to 1986 [P1], and the period from 1987 to 2003 [P2]) were analyzed.

Results: During P1, 58 patients with PVE were treated (30 early PVE and 28 late PVE); during P2, 63 patients with PVE were treated (13 early PVE and 50 late PVE). The frequency of early-onset PVE decreased from 0.94% in P1 to 0.34% in P2 (p < 0.001), but the incidence rate of late-onset PVE did not change (0.33% and 0.42% per year, respectively). The microbiology of early PVE changed over the years: Gram-negative bacilli decreased from 40% during P1 to 7.7% in P2 (p = 0.033). Staphylococci remained the main causes of early PVE in both periods. The microbial etiology of late PVE also changed over the years with enterococci and Staphylococcus aureus as the leading causes during P2. Streptococcus viridans decreased from a leading position to a fourth position. Methicillin-resistant S aureus endocarditis appeared first in 1992. Eleven cases of late-onset PVE in P2 were hospital acquired (22%). In comparison, only two cases (7.1%) of hospital-acquired, late-onset PVE were seen in P1 (p = 0.11). Mortality of early-onset PVE decreased from 80% in P1 to 46% in P2 (p = 0.026). The overall mortality of late-onset PVE did not change between periods: 39% vs 34%. Mortality associated with nosocomial PVE in P2 was 63.6% (7 of 11 patients). In comparison, the mortality of community-acquired eases was 25.6% (10 of 39 patients; p = 0.03). In the multivariate analysis, the presence of comorbidities and hospital acquisition were associated with an excess of mortality (odds ratio [OR], 13.9; 95% confidence interval [CI], 1.23 to 158 [p = 0.033]; and OR, 10.8; 95% CI, 2.16 to 54.7 [p = 0.0037], respectively).

Conclusion: Although the mortality associated with early-onset PVE has significantly decreased, in this series the mortality of patients with late-onset PVE remained high due mainly to an increasing number of patients with comorbidities who acquired the infection during admission for other diseases.

Key words: cross infection; endocarditis; prosthesis-related infections

Abbreviations: CI = confidence interval; IQR = interquartile range; MRSA = methicillin-resistant Staphylococcus aureus; OR = odds ratio; P1 = period from 1970 to 1986; P2 = period from 1987 to 2003; PVE = prosthetic valve endocarditis

**********

Infectious endocarditis is a changing disease with predisposing risk factors, etiology, manifestations, and therapeutics in continuous evolution. (1,2) Since the late 1970s, several reports (3-6) have shown the effect of medical progress on the epidemiologic patterns of infectious endocarditis with emphasis on new populations at risk, increasing prevalence of resistant pathogens, or the role of advanced age on the incidence and outcome of endocarditis.

Although recognized for almost 30 years, invasive medical technology has been associated with an increasing number of cases of infectious endocarditis, and there is evidence that hospital-acquired endocarditis not associated with cardiac surgery is on the rise. (7-9) A comparison of cases of endocarditis seen from 1978 to 1992 with a historical control group conducted at our institution (8) revealed a tenfold increase in the number of patients with hospital-acquired endocarditis diagnosed in the 1980s; and more recently, a study (9) from the United Kingdom showed that the incidence of nosocomial endocarditis increased sevenfold between 1985 and 1996.

Endocarditis developing on prosthetic valves accounts for 15 to 32% of all cases of infectious endocarditis. (1,5,10-13) Advances in the management of prosthetic valve endocarditis (PVE) include a lesser incidence of early-onset infections, improvements in diagnosis by means of transesophageal echocardiography, and better outcome associated with combined medical/surgical treatment. (13-16) However, the changes that medical practice may have introduced in recent years in the epidemiology, prognosis, and risk factors for morality of PVE patients has not been assessed. Specifically, the impact of nosocomial acquisition of PVE has not been evaluated in details. In this study, we tried to elucidate the possible changes in the microbial etiology and prognosis of late-onset PVE associated with hospital-acquired infections.

MATERIALS AND METHODS

The hospital records of all patients with a discharge diagnosis or postmortem findings indicative of PVE treated at the Fundacion Jimenez Diaz from January 1970 to December 2003 were retrospectively reviewed. Our institution is a 600-bed, tertiary-care, university-affiliated hospital in Madrid providing medical care to a population of 235,000 persons. During the first 15 years of the study, it was a reference center for cardiac surgery. Ninety percent of the patients were seen by one or more of the authors, and the most relevant clinical data were recorded in "ad hoc" sheets from 1975 to 1986 and within a database from 1987 to 2003.

The following information was collected from medical records: age; gender; coexisting comorbidities diagnosed according to clinical criteria (diabetes mellitus, ischemic miocardiopathy, stroke, chronic lung disease, chronic liver disease, chronic renal failure, malignancy, HIV infection, and any other condition requiring admission for surgery); hospital-based diagnostic or therapeutic procedures performed on each patient (instrumentation, surgical procedures); type of prosthesis (mechanical vs biological) and its location within the heart; clinical signs and symptoms; microorganisms isolated from blood cultures or cultures of prosthetic valves or cardiac tissue; echocardiographic findings; medical and surgical treatment; operative findings; and outcome. Acute renal failure was defined as an increment of serum creatinine level > 2 mg/dL.

In order to assess differences in the epidemiology and clinical variables of PVE over the years, patients were classified into one of two periods of the same duration: the period from 1970 to 1986 (P1), or the period from 1987 to 2003 (P2). Patients of both groups were compared.

During P1, diagnosis of infectious endocarditis was performed using criteria proposed by Pelletier and Petersdof (17) and later by Von Reyn et al. (18) The Duke criteria were prospectively or retrospectively applied for the diagnosis of endocarditis during P2. (19) When echocardiographic examinations were available, these criteria were also retrospectively applied to cases seen during P1. Cases were considered early-onset PVE when they occurred within the first year after valve replacement and late-onset PVE when occurring later. (20) The choice of a period of 12 months after valve replacement instead of the most conventional of 60 clays was based on previous studies (21) that documented incubations periods of 8 months and 11 months of infections shown to be the result of surgical contamination; and so, we tried to avoid the inclusion of cases of delayed onset of early PVE as cases of late-onset PVE. Active endocarditis was a pathologic designation applied on the basis of autopsy or surgical findings of cardiac vegetations, ulcerative dehiscence of prosthesis, or abscesses and generally associated with positive blood culture results and positive culture results of cardiac tissue. (14) Only definitive cases of endocarditis were included.

Nosocomial endocarditis was defined as endocarditis developing > 72 h after admission in direct association with a hospital-based procedure performed during the current hospitalization or during another hospitalization within the preceding 8 weeks. (8) Since cases of early-onset PVE were commonly associated with cardiac surgery and hospital paraphernalia, nosocomial endocarditis was exclusively applied in cases of late-onset PVE acquired during hospitalization for another medical or surgical condition. New endocarditis, that is, development of endocarditis after the onset of bacteremia, was defined by the absence of clinical evidence of endocarditis at the time the bacteremia was first documented. (22) This category was exclusively applied to patients with nosocomial endocarditis.

Endocarditis was considered associated with an IV catheter when signs of endocardial infection were documented in a patient with concurrent or recent (< 8 weeks) catheter-associated bacteremia. Endocarditis was considered associated with a genitourinary procedure when it developed during hospitalization or within 8 weeks of a preceding hospitalization and was temporarily linked to manipulation of the urogenital tract (Foley catheterization, curettage, transurethral resection of the prostatic gland). Endocarditis-related mortality includes all deaths that occurred during the hospitalization as a probable consequence of infection, including operative mortality and those deaths that occurred within 30 days of discharge.

Categorical variables were reported as percentages and continuous variables as the mean [+ or -] SD or median with interquartile range (IQR). Differences between continuous variables were analyzed using Student t test or nonparametric test in samples without normal distribution. Differences between noncontinuous variables were analyzed using the [chi square] test or Fisher Exact Test when required. Significance was set at p < 0.05. Stepwise logistic regression analysis was applied to variables that yielded significant results in the univariate analysis to identify risk factors for mortality of late-onset PVE during P2.

RESULTS

One hundred twenty-one cases of PVE were studied: 58 patients during P1 and 63 patients during P2. All of them fulfilled the criteria for definite endocarditis. Table 1 summarizes the characteristics of these patients. The mean age significantly increased from 44.2 [+ or -] 12.9 to 60.6 [+ or -] 11.3 years. The use of bioprosthesis diminished during the second period; consequently, the number of eases of PVE occurring on bioprosthesis significantly decreased.

Incidence

Fifty-three of 58 eases of PVE during P1 were seen from 1975 to 1986. In this period, 3,193 valve replacements were performed with an overall frequency of PVE of 1.82%. The cumulative follow-up was 15,640 patient-years. The number of surgeries with valve replacement performed during P2 was 3,837, with an overall frequency of PVE of 1.64% (not significant [NS]). The cumulative follow-up was 14,737 patient-years (NS).

The frequency of early-onset PVE significantly decreased from 0.94% to 0.34% (p < 0.001). However, the overall frequency of late-onset PVE increased although not significantly from 0.88% in P1 to 1.30% in P2 (p = 0.17). The incidence rates of PVE during P1 and P2 were 0.33% and 0.42% per year, respectively (NS).

Microbial Etiology and Setting

Coagulase-negative staphylococci and Staphylococcus aureus were the leading causes of early-onset PVE in both periods (Table 2). During the first years of the study, cases of endocarditis caused by Serratia marcesens, Pseudomonas, and other nonfermentative Gram-negative bacilli were seen. Some of these cases occurred as part of small outbreaks associated with contamination of surgical devices. These nosocomial infections disappeared during the late 1970s and were not found during P2.

The microbial etiology of late-onset PVE also changed with S aureus, enteroeocci, and coagulase-negative staphylococci as the main causes during P2. Streptococcus viridans decreased from a leading position in P1 to a fourth position in P2. Methicillin-resistant S aureus (MRSA) endocarditis appeared first in 1992.

Eleven of 50 cases (22%) of late-onset PVE seen during P2 were hospital acquired. In comparison, only two cases (7.1%) of hospital-acquired, late-onset PVE were seen in the early years of the study (p = 0.11). Most of these nosocomial infections were caused by staphylococci including MRSA and enterococci (Table 3). Evidence for endocarditis was present at the time of the first positive blood culture result or developed shortly after in most patients. However, new endocarditis was diagnosed in five cases (cases 2, 4, 5, 8, and 9). Endometritis and Escherichia coli bacteremia after miscarriage developed in one patient (case 2). No murmurs or emboli were present at outset, and transthoracic echocardiography results were negative. Curettage and antibiotics did temporally clear bacteremia, but blood culture results were again positive 10 days later. Because of "breakthrough bacteremia," she was taken to the operating room for valve replacement. Signs of active endocarditis were found. Staphylococcal abscess of the prostatic sac after transurethral resection of the gland developed in one patient (case 5). Antimicrobial therapy was administered for 3 weeks, but after discontinuation bacteremia relapsed and signs of endocarditis were observed. Candidemia associated with parenteral nutrition developed in one patient (ease 9). No signs of endocarditis were found at outset, and transesophageal echocardiography results were negative. After 11 days of antifungal therapy, an acute popliteal emboli developed, which was removed. Pseudohyphi were observed, and the patient was taken to the operating room. Bulky vegetations were observed on the mitral valve.

Mortality

The overall mortality of PVE during P1 was 60.3% and decreased to 36.5% during P2 (p = 0.009). Examined by groups, the mortality of early-onset PVE significantly decreased from 80% in P1 to 46% in P2 (p = 0.026). However, mortality associated with late-onset PVE did not significantly change between periods: 39% and 34%, respectively. Mortality associated with nosocomial late-onset PVE in P2 was 63.6% (7 of 11 patients). In comparison, the mortality of patients with community-acquired disease was 25.6% (10 of 39 patients; p = 0.03)

Patients were followed up a median of 12.6 months (IQR, 1.9 to 48.6 months). Sixty percent were still alive 60 months after the diagnosis of endocarditis. Only three patients needed valve replacement due to dysfunction of the prosthetic valve 4, 8, and 9 years after diagnosis of endocarditis.

Risk Factors for Mortality of Late-Onset PVE Patients Observed During P2

The median duration of symptoms until diagnosis was 21 days (IQR, 7 to 32 days). Mortality was associated with a shorter period of symptoms (median, 14 days vs 28 days; p = 0.037). S aureus followed by enterococci and coagulase-negative staphylococci were the microrganisms with the shortest periods of symptoms (7 days: IQR, 2 to 2.9 days; 9.5 days: IQR, 7 to 14 days; and 12 days: IQR, 1 to 29 days, respectively). Patients with PVE caused by S viridans had a symptomatic period of 30 days (IQR, 28 to 120 days). Staphylococcal infection was associated with a greater risk of mortality (odds ratio [OR], 3.62; 95% confidence interval [CI], 1.23 to 10.6; p = 0.018).

The presence of one comorbidity was associated with mortality (OR, 3.2; 95% CI, 1.11 to 9.39; p = 0.031), and the presence of two or more comorbidities was associated with a still-higher risk of mortality (OR, 10.8; 95% CI, 1.17 to 99.5; p = 0.035). The single comorbidity that independently exhibited an increased risk of mortality was cardiovascular disease defined by a history of previous stroke or myocardial ischemia, not related to the episode of endocarditis (OR, 13.7; 95% CI, 1.53 to 123.1; p = 0.019). Acute renal failure occurred in 14% (7 of 50 cases) and cardiac failure was found in 62% (31 of 50 cases) of late-onset PVE. Emboli were observed 22% (11 of 50 cases). Only renal failure was significantly associated with mortality (OR, 3.93; 95% CI, 1.008 to 153; p = 0.048).

The role of treatment modality (surgical vs medical) on mortality was also assessed. Valve replacement was performed in 39 of 50 patients (78%) with late-onset PVE. Most (70%) had signs of active endocarditis found during surgery, and 38.4% had positive culture findings of prosthesis or cardiac tissue. Indications for surgery were congestive heart failure in 20 cases, prosthetic dehiscence in 8 patients, and poor response to antibiotics--"breakthrough bacteremia" or unremitting fever--in 11 patients. However, 11 patients with late-onset PVE were treated with antimicrobials alone. Seven of these patients--three patients with endocarditis due to S viridans and four patients with endocarditis due to enterococci--were cured with antimicrobial therapy alone. Surgery was not indicated or was refused by patients or relatives in another four instances, and all died.

In summary, of 39 patients with late-onset PVE in whom valve replacement was performed, 13 patients (33.3%) died, in comparison with 4 of 11 patients (36.3%) who were considered for medical treatment alone. Valve replacement was not an independent factor associated with survival in cases of late-onset PVE in this series. In the multivariate analysis, the presence of cardiovascular disease and the nosocomial acquisition of late-onset PVE remained the only independent risk factors for mortality during P2 (OR, 13.9; 95% CI, 1.23 to 158; p = 0.033; and OR, 10.8; 95% CI, 2.16 to 54.7, respectively [p = 0.0037]).

DISCUSSION

The comparison of two consecutive series of patients with PVE revealed the changing epidemiology of the infection and the impact of medical practice on the microbiology and prognosis of this infection. Firstly, the incidence of early-onset PVE did decrease. The lesser incidence of early-onset PVE may be attributed to improved surgical techniques, shorter surgical procedures and admission in the ICUs, better hospital hygiene, and antimicrobial prophylaxis. The complete disappearance of early-onset PVE caused by Gram-negative bacteria is remarkable. A comparison of cases of PVE published in the 1980s with more recent studies (23-28) also showed a decrease in the frequency of this type of infection. However, we observed that the incidence of late-onset PVE increased, although not significantly, during the second part of the study. The longer survival of patients after cardiac surgery and the growing population of individuals with prosthetic valves may some way explain this trend. Nevertheless, we believe that the increasing age of these patients, the common presence of comorbidities and, mostly, the increasing number of cases of hospital-acquired endocarditis were of paramount importance in the occurrence of new cases of late-onset PVE. The changing microbial spectrum of late-onset PVE in our institution, with lesser number of cases caused by streptococci and other community-acquired pathogens and the increased frequency of S aureus and enterococci, supports the idea that the growing incidence of nosocomial infections has an important role in the epidemiology of PVE at the present time.

The frequency of nosocomial endocarditis not associated with cardiac surgery is increasing both in Europe and the United States. (8,9,29-32) Fowler et al (30) found that endocarditis caused by S aureus was hospital acquired in 46% of their patients, and the incidence of new endocarditis following staphylococcal bacteremia in the subset of patients with a prosthetic valve was 17%. In a prospective study (31) of bacteremia, new endocarditis developed in 8% of patients with hospital-acquired bacteremia due to Enterococcus faecalis. Bacteremic patients with prosthetic valves were at notable risk for endocarditis, even when they received antibiotic therapy before endocarditis developed and regardless of the duration of such therapy. (22) S aureus, coagulase-negative staphylococci, enterococci, and some Gram-negative bacilli are most frequently isolated in patients with nosocomial endocarditis. (8,9,30,32) Corynebacterium amycolatum and other nondiphtheriae corynebacterium are also occasionally found. (33) In addition, patients with prosthetic heart valves who acquired nosocomial candidemia are at notable risk of having candida PVE or having candida PVE develop months or years later. (34) The importance of these data could be better interpreted taking into account the evolution of the incidence of bacteremia and fungemia in US hospitals in the last 2 decades. Between 1979 and 2000, there was an annualized increase in the incidence of sepsis of 8.7%, from approximately 164,000 cases to nearly 660,000 cases. (35) The rate of sepsis due to fungal organisms increased by 207%, with Gram-positive bacteria becoming the predominant pathogens after 1987. (35) Hence, an increased incidence of bacteremia and fungemia in patients at risk, mainly elderly people with underlying valve disorders or prosthetic valves, may result in a higher frequency of nosocomial endocarditis. (8,9)

The mortality of patients with early-onset PVE significantly diminished in the second period. The better outcome of these patients was probably associated with improvements in diagnosis by means of transesophageal echocardiography, more frequent use of surgical treatment, and the lesser number of infections caused by Gram-negative bacilli. However, the unexpected finding of mortality of patients of late-onset PVE that did not favorably change between periods is most worrisome.

Several reports (13,24,28,36,37) have shown that patients with PVE had a greater risk of mortality when the infection was caused by microorganisms other than streptococci, when valve replacement was not performed, when heart failure and other cardiac complications developed, or when cerebral emboli or renal failure occurred. Some of these complications were found in our patients with late-onset PVE, but the presence of comorbidities, particularly cardiovascular disease and nosocomial acquisition of the infection, were the most important determinants of mortality.

As shown by others (32) and by our results, the mortality of patients with nosocomial endocarditis is higher than that observed among cases of community-acquired endocarditis. The presence of other cardiac or noncardiae underlying diseases that may limit the ability of the patient to survive a severe infection such as endocarditis or precludes valve replacement are serious obstacles for success. In addition, antimicrobial resistance of nosocomial pathogens must also be taken into account. MRSA, enterococci, corynebacteria, Gram-negative bacilli, and Candida spp were commonly found in this and other series of nosocomial endocarditis. (8,9,22) Fortunately, long-term prognosis was excellent, most patients did not experience adverse clinical events, and only a minority needed reinterventions.

Because of the substantial morbidity and mortality of nosocomial endocarditis involving prosthetic valves, major emphasis must be put on prevention. The review of cases of this and other series have shown that many may have been prevented. (8,9,22) Prevention should largely be aimed at first identifying the patients at risk. Individuals who are admitted to the hospital with conditions whose diagnosis or treatment require instrumentation of the urogenital tract, surgical procedures, invasive studies, parenteral nutrition, or prolonged IV therapy should be identified at the time of admission. Central venous lines should be avoided, and all invasive procedures should be performed under strict aseptic conditions. Until new recommendations on antimicrobial prophylaxis specifically designed for prevention of nosocomial endocarditis are issued, antibiotics should be administered following the most recent recommendations. (38) Unfortunately, it seems likely that the number of eases of nosocomially acquired PVE will continue to increase.

REFERENCES

(1) Mylonakis E, Calderwood SB. Infective endocarditis in adults. N Engl J Med 2001 1; 345:1318-1330

(2) Moreillon P, Que YA. Infective endocarditis. Lancet 2004; 363:139-149

(3) Hogevik H, Olaison L, Andersson R, et al. Epidemiologic aspects of infective endocarditis in an urban population: a 5-year prospective study. Medicine (Baltimore) 1995; 74: 324-339

(4) Spies C, Madison JR, Schatz IJ. Infective endocarditis in patients with end-stage renal disease. Arch Intern Med 2004; 164:71-75

(5) Bouza E, Menasalvas A, Munoz P, et al. Infective endocarditis: a prospective study at the end of the twentieth century; new predisposing conditions, new etiologic agents, and still a high mortality. Medicine (Baltimore) 2001; 80:298-307

(6) Werner GS, Schulz R, Fuchs JB, et al. Infective endocarditis in the elderly in the era of echocardiography: clinical features, and prognosis compared with younger patients. Am J Med 1996; 100:90-97

(7) Watanakunakorn C. Infective endocarditis as a result of medical progress. Am J Med 1978; 64:917-919

(8) Fernandez Guerrero ML, Verdejo C, Azofra J, et al. Hospital-acquired infectious endocarditis not associated with cardiac surgery: an emerging problem. Clin Infect Dis 1995; 20: 16-23

(9) Lamas CC, Eykin SJ. Hospital acquired native endocarditis: analysis of 22 cases presenting over 11 years. Heart 1998; 79:442-447

(10) Watanakunakorn C, Burkert T. Infective endocarditis at a large community teaching hospital, 1980-1990: a review of 210 episodes. Medicine 1993; 12:90-102

(11) Wallace SM, Walton BI, Kharbanda RK, et al. Mortality from infective endocarditis: clinical predictors of outcome. Heart 2002; 88:53-60

(12) Sandre RM, Shafran SD. Infective endocarditis: review of 135 cases over 9 years. Clin Infect Dis 1996; 22:276-286

(13) Wolff M, Witchitz S, Chastang C, et al. Prosthetic valve endocarditis in the ICU: prognostic factors of overall survival in a series of 122 cases and consequences for treatment decision. Chest 1995; 108:688-694

(14) Mullany CJ, Chua YL, Schaff HV, et al. Early and late survival after surgical treatment of culture-positive active endocarditis. Mayo Clin Proc 1995; 70:517-725

(15) John MDV, Hibberd PL, Karchmer AW, et al. Staphylococcus aureus prosthetic valve endocarditis: optimal management and risk factors for death. Clin Infect Dis 1998; 26:1302-1309

(16) Vlessis AA, Khaki A, Grunkenmeier GL, et al. Risk, diagnosis and management of prosthetic valve endocarditis: a review. J Heart Valve Dis 1997; 6:443-465

(17) Pelletier LL Jr, Petersdorf RG. Infective endocarditis: a review of 125 cases from the University of Washington Hospitals, 1963-72. Medicine (Baltimore) 1977; 56:287-313

(18) Von Reyn CF, Levy BS, Arbeit RD, et al. Infective endocarditis: an analysis based on strict case definitions. Ann Intern Med 1981; 94:505-518

(19) Durack DT, Lukes AS, Bright DK. New criteria for diagnosis of infective endocarditis: utilization of specific echocardiographic findings; Duke Endocarditis Service. Am J Med 1994; 96:200-209

(20) Karchmer AW, Archer GL, Dismukes WE. Staphylococcus epidermidis causing prosthetic valve endocarditis: microbiologic and clinical observations as guides to therapy. Ann Intern Med 1983; 98:447-455

(21) Archer GL, Vishniavsky N, Stiver HG. Plasmid pattern analysis of Staphylococcus epidermidis isolates from patients with prosthetic valve endocarditis. Infect Immun 1982; 35: 627-632

(22) Fang G, Keys TF, Gentry LO, et al. Prosthetic valve endocarditis resulting from nosocomial bacteremia: a prospective, multicenter study. Ann Intern Med 1993; 119:560-567

(23) Wilson WR, Danielson GK, Giuliani ER, et al. Prosthetic valve endocarditis. Mayo Clin Proc 1982; 57:155-161

(24) Calderwood SB, Swinski LA, Karchmer AW, et al. Prosthetic valve endocarditis: analysis of factors affecting outcome of therapy. J Thorac Cardiovasc Surg 1986; 92:776-783

(25) Aoyagi S, Oryoji A, Nishi Y, et al. Long-term results of valve replacement with the St. Jude Medical valve. J Thorac Cardiovasc Surg 1994; 108:1021-1029

(26) Vlessis AA, Khaki A, Grunkeimer GL, et al. Risk, diagnosis and management of prosthetic valve endocarditis: a review. J Heart Valve Dis 1997; 6:443-665

(27) Sidhu P, O'Kane H, Ali N, et al. Mechanical or bioprosthetic valves in the elderly: a 20-year comparison. Ann Thorac Surg 2001; 71:S257-S60

(28) Tornos P, Almirante B, Olona M, et al. Clinical outcome and long-term prognosis of late prosthetic valve endocarditis: a 20-year experience. Clin Infect Dis 1997; 24:381-386

(29) Berlin JA, Abrutyn E, Strom BL, et al. Incidence of infective endocarditis in the Delaware Valley, 1988-1990. Am J Cardiol 1995; 76:933-936

(30) Fowler VG, Sanders LL, Kong LK, et al. Infective endocarditis due to Staphylococcus aureus: 59 prospectively identified cases with follow-up. Clin Infect Dis 1999; 28:106-114

(31) Fernandez Guerrero ML, Herrero L, Bellver MT, et al. Nosocomial enterococcal endocarditis: a serious hazard for hospitalized patients with enterococcal bacteremia. J Intern Med 2002; 252:510-515

(32) Gilleece A, Fenelon L. Nosocomial infective endocarditis. J Hosp Infect 2000; 46:83-88

(33) Knox KL, Holmes AH. Nosocomial endocarditis caused by Corynebacterium amycolatum and other nondiphtheriae corynebacteria. Emerg Infect Dis 2002; 8:97-99

(34) Nasser RM, Melgar GR, Longworth DL, et al. Incidence and risk of developing fungal prosthetic valve endocarditis after nosocomial candidemia. Am J Med 1997; 103:25-32

(35) Martin GS, Mannino DM, Eaton S, et al. The epidemiology of sepsis in the United States from 1979 through 2000. N Engl J Med 2003; 348:1546-1554

(36) John MDV, Hibberd PL, Karchmer AW, et al. Staphylococcus aureus prosthetic valve endocarditis: optimal management and risk factors for death. Clin Infect Dis 1998; 26:1302-1309

(37) Tornos P, Almirante B, Mirabet S, et al. Infective endocarditis due to Staphylococcus aureus: deleterious effect of anticoagulant therapy. Arch Intern Med 1999; 159:473-475

(38) Dajani AS, Bisno AL, Chung KJ, et al. Prevention of bacterial endocarditis: recommendations by the American Heart Association. JAMA 1990; 264:2919-2922

COPYRIGHT 2005 American College of Chest Physicians
COPYRIGHT 2005 Gale Group