Embed Size (px)
Citation preview
current as of November 13, 2009. Online article and related content
http://jama.ama-assn.org/cgi/content/full/297/12/1354
. 2007;297(12):1354-1361 (doi:10.1001/jama.297.12.1354) JAMA
Andrew Wang; Eugene Athan; Paul A. Pappas; et al.
Prosthetic Valve EndocarditisContemporary Clinical Profile and Outcome of
Correction Contact me if this article is corrected.
Citations Contact me when this article is cited. This article has been cited 33 times.
Topic collections
Contact me when new articles are published in these topic areas.Cardiovascular Disease/ Myocardial Infarction; Infectious Diseases Bacterial Infections; Infectious Diseases, Other; Prognosis/ Outcomes;
the same issueRelated Articles published in
. 2007;297(12):1396.JAMAJanet M. Torpy et al. Heart Valve Infections
http://pubs.ama-assn.org/misc/[email protected]
http://jama.com/subscribeSubscribe
[email protected]/E-prints
http://jamaarchives.com/alertsEmail Alerts
by guest on November 13, 2009 www.jama.comDownloaded from
CLINICIAN’S CORNERORIGINAL CONTRIBUTION
Contemporary Clinical Profile and Outcomeof Prosthetic Valve EndocarditisAndrew Wang, MDEugene Athan, MDPaul A. Pappas, MSVance G. Fowler, Jr, MD, MHSLars Olaison, MDCarlos Pare, MDBenito Almirante, MDPatricia Muñoz, MDMarco Rizzi, MDChristoph Naber, MDMateja Logar, MDPierre Tattevin, MDDiana L. Iarussi, MDChristine Selton-Suty, MDSandra Braun Jones, MDJose Casabe, PhDArthur Morris, MDG. Ralph Corey, MDChristopher H. Cabell, MD, MHSfor the International Collaborationon Endocarditis-Prospective CohortStudy Investigators
HEART VALVE REPLACEMENT
surgery is performed in morethan 150 000 patients peryear in the United States and
Europe,1,2 and the number of pros-thetic valve implantations continues toincrease.3 Infection of a heart valve pros-thesis, or prosthetic valve endocardi-tis (PVE), is an uncommon but poten-tially lethal complication of heartvalve replacement surgery. Despite ad-
See also Patient Page.
CME available online atwww.jama.com
Author Affiliations: Department of Medicine, Duke Uni-versity Medical Center, Durham, NC (Drs Wang, Fowler,Corey, and Cabell); Department of Infectious Disease,The Geelong Hospital at the University of Melbourne,Melbourne, Australia (Dr Athan); Outcomes Researchand Assessment Group, Duke Clinical Research Insti-tute, Durham, NC (Mr Pappas); Department of Infec-tious Disease, Sahlgrenska Universitetssjukhuset/Ostra, Göteborg, Sweden (Dr Olaison); Department ofCardiology, University of Barcelona, Barcelona, Spain(Dr Pare); Infectious Diseases Department, Hospital Uni-versitari Vall d’Hebron, Barcelona, Spain (Dr Almi-rante); Department of Infectious Disease, Hospital Gen-eral Universitario Gregorio Maranon, Barcelona, Spain(Dr Muñoz); Department of Infectious Disease, Os-pedali Riuniti, Bergamo, Italy (Dr Rizzi); Cardiology Clinic,University Essen, Essen, Germany (Dr Naber); Depart-ment of Infectious Disease, Medical Center Ljublijana,
Ljublijana, Slovenia (Dr Logar); Department of Infec-tious Disease, Pontchaillou University, Rennes, France(Dr Tattevin); Department of Cardiology, II Universitadi Napoli, Naples, Italy (Dr Iarussi); Cardiology Ser-vice, CHU Nancy-Brabois, Nancy, France (Dr Selton-Suty); Laboratorio de Tecnicas No Invasivas, HospitalClinico Pont Universidad Catolica de Chile, Santiago,Chile (Dr Jones); Department of Cardiology, Institutode Cardiologia y Cirugia Cardiovascular, Fundacion Fa-valoro, Buenos Aires, Argentina (Dr Casabe); Diagnos-tic Medlab, Auckland City Hospital, Auckland, New Zea-land (Dr Morris).A List of the ICE Investigators appears at the end ofthis article.Corresponding Author: Andrew Wang, MD, Depart-ment of Medicine, Duke University Medical Center,DUMC 3428, Room 7624 A, Durham, NC 27710([email protected]).
Context Prosthetic valve endocarditis (PVE) is associated with significant mortality andmorbidity. The contemporary clinical profile and outcome of PVE are not well defined.
Objectives To describe the prevalence, clinical characteristics, and outcome of PVE,with attention to health care–associated infection, and to determine prognostic fac-tors associated with in-hospital mortality.
Design,Setting, andParticipants Prospective, observational cohort study conductedat 61 medical centers in 28 countries, including 556 patients with definite PVE as definedby Duke University diagnostic criteria who were enrolled in the International Collabora-tion on Endocarditis-Prospective Cohort Study from June 2000 to August 2005.
Main Outcome Measure In-hospital mortality.
Results Definite PVE was present in 556 (20.1%) of 2670 patients with infective en-docarditis. Staphylococcus aureus was the most common causative organism (128 pa-tients [23.0%]), followed by coagulase-negative staphylococci (94 patients [16.9%]). Healthcare–associated PVE was present in 203 (36.5%) of the overall cohort. Seventy-one per-cent of health care–associated PVE occurred within the first year of valve implantation,and the majority of cases were diagnosed after the early (60-day) period. Surgery wasperformed in 272 (48.9%) patients during the index hospitalization. In-hospital deathoccurred in 127 (22.8%) patients and was predicted by older age, health care–associated infection (62/203 [30.5%]; adjusted odds ratio [OR], 1.62; 95% confidenceinterval [CI], 1.08-2.44; P=.02), S aureus infection (44/128 [34.4%]; adjusted OR, 1.73;95% CI, 1.01-2.95; P=.05), and complications of PVE, including heart failure (60/183[32.8%]; adjusted OR, 2.33; 95% CI, 1.62-3.34; P�.001), stroke (34/101 [33.7%]; ad-justed OR, 2.25; 95% CI, 1.25-4.03; P=.007), intracardiac abscess (47/144 [32.6%];adjusted OR, 1.86; 95% CI, 1.10-3.15; P=.02), and persistent bacteremia (27/49 [55.1%];adjusted OR, 4.29; 95% CI, 1.99-9.22; P�.001).
Conclusions Prosthetic valve endocarditis accounts for a high percentage of all casesof infective endocarditis in many regions of the world. Staphylococcus aureus is nowthe leading cause of PVE. Health care–associated infection significantly influences theclinical characteristics and outcome of PVE. Complications of PVE strongly predict in-hospital mortality, which remains high despite prompt diagnosis and the frequent useof surgical intervention.JAMA. 2007;297:1354-1361 www.jama.com
1354 JAMA, March 28, 2007—Vol 297, No. 12 (Reprinted) ©2007 American Medical Association. All rights reserved.
by guest on November 13, 2009 www.jama.comDownloaded from
vances in its diagnosis and treatment,PVE is associated with substantial mor-bidity and mortality.4-8 Prosthetic valveendocarditis has been estimated to oc-cur at a rate of 0.3% to 1% per patient-year and to account for 1% to 5% of allcases of infective endocarditis (IE).9-12
Recent studies of predominantly na-tive valve infective endocarditis (NVIE)have found an increased rate of Staphy-lococcus aureus infection, with an as-sociation between health care contactand this causative organism.13-15 How-ever, the contemporary clinical pro-file and outcome of PVE have not beenevaluated in large, prospective stud-ies. Much of the current understand-ing of PVE has been based on studieslimited by small sample size, retrospec-tive design, or single-center experi-ences, and many of these investiga-tions antedated the routine use ofechocardiography in the diagnosis of IEand the contemporary and validatedDuke diagnostic criteria for IE.16,17
We hypothesized that the contem-porary profile of PVE may be signifi-cantly influenced by health care–associated infection. The objectives ofthis study were (1) to prospectively de-scribe the prevalence, clinical charac-teristics, and outcome of PVE, with at-tention to health care–associatedinfection; and (2) to determine prog-nostic factors associated with in-hospital mortality. To accomplish thesegoals, we analyzed data from the Inter-national Collaboration on Endocarditis-Prospective Cohort Study (ICE-PCS),a large, prospective, contemporary,multinational registry of endocarditis.
METHODSInternational Collaborationon Endocarditis-ProspectiveCohort Study
Data from the International Collabo-ration on Endocarditis (ICE) were usedfor this study. The background and in-clusion criteria of this prospective, mul-ticenter, international registry of IE havebeen reported previously.14,15,18 Be-tween June 2000 and August 2005,3250 patients from 61 centers in 28countries were enrolled. The ICE-PCS
database is maintained at the DukeClinical Research Institute, which is thecoordinating center for ICE studies. In-formed consent (oral or written) wasobtained from all patients according tolocal institutional review boards or eth-ics committee guidelines at all sites.Geographic regions participating inICE-PCS that enrolled patients withdefinite PVE included the following:United States (10 sites), South America(7 sites), Northern/Central Europe (14sites), Southern Europe/Middle East/South Africa (11 sites), and Australia/New Zealand/Asia (11 sites).
Patient Selectionand Data Collection
Patients were identified prospectivelyusing site-specific procedures to en-sure consecutive enrollment.15,18 Pa-tients were enrolled in ICE-PCS if theymet criteria for possible or definite IEbased on modified Duke criteria.16 Onlypatients with definite IE were in-cluded in the current study. To pre-serve the assumption of indepen-dence of observations, only the firstepisode of IE recorded for an indi-vidual patient was used in the analysis.
The method of data collection forICE-PCS has been previously re-ported.14 Briefly, a standard case re-port form was used at all sites to col-lect data. This case report form included275 variables and was developed by ICEaccording to standard definitions.18 Datawere collected during the index hos-pitalization and were then entered at thecoordinating center or by site investi-gators using an Internet-based data en-try system.
Definitions
Definitions of the variables included inthe ICE-PCS case report form have beenpreviously reported.14 “Intravascular ac-cess devices” were defined as an arterial-venous fistula or an indwelling vascu-lar catheter. A “long-term indwellingcentral venous catheter” was defined asa tunneled, cuffed catheter, or a sub-cutaneous port catheter. An “intravas-cular access device” was presumed tobe a possible source of IE if it was pres-
ent at the onset of IE symptoms. “Per-sistent bacteremia” was defined as pre-viously reported.16,17 An “intracardiacabscess” was defined as a thickened areaor mass with a heterogeneous echo-genic or echolucent appearance by ech-ocardiography, or the presence of pusby direct visualization at the time ofsurgery.19
“Early PVE” was defined as the diag-nosis of PVE within 60 days of pros-thetic valve implantation.20-22 Healthcare–associated infection in PVE was cat-egorized as either “nosocomial” or “non-nosocomial health care–associated in-fection.”23 “Nosocomial infection” wasdefined as PVE developing in a patienthospitalized for more than 48 hours be-fore the onset of signs or symptoms con-sistent with IE. “Non-nosocomial healthcare–associated infection” was defined asIE diagnosed within 48 hours of admis-sion in an outpatient with extensivehealth care contact as reflected by any ofthe following criteria: (1) received in-travenous therapy, wound care, or spe-cialized nursing care at home within the30 days prior to the onset of PVE; (2) at-tended a hospital or hemodialysis clinicor received intravenous chemotherapywithin the 30 days before the onset ofPVE; (3)washospitalized inanacutecarehospital for 2 or more days in the 90 daysbefore the onset of PVE; or (4) residedin a nursing home or long-term care fa-cility.23
Statistical Analysis
Continuous variables were presented asmedians with 25th and 75th percen-tiles. Categorical variables were pre-sented as frequencies and percentagesof the specified group. Univariable com-parisons of clinical characteristics weremade with the Wilcoxon rank-sum testor the �2 test as appropriate. Within thePVE group, analyses were performed toevaluate the nature of early PVE com-pared with late PVE and, more broadly,health care–associated PVE in com-parison to community-acquired PVE.Regional differences in PVE were as-sessed by dividing the participating sitesin ICE-PCS based on their geographicregion while integrating regions with
CONTEMPORARY CLINICAL PROFILE AND OUTCOME OF PROSTHETIC VALVE ENDOCARDITIS
©2007 American Medical Association. All rights reserved. (Reprinted) JAMA, March 28, 2007—Vol 297, No. 12 1355
by guest on November 13, 2009 www.jama.comDownloaded from
few PVE cases (eg, Africa) with otherproximal regions (Southern Europe andMiddle East).
An exploratory multivariable gener-alized estimating equation model wascreated to determine variables inde-pendently associated with in-hospitalmortality in PVE. Final parameter es-timates were converted to odds ratioswith corresponding 95% confidence in-tervals. To avoid overfitting the data,15 clinically relevant variables wereconsidered for the final generalized es-timating equation model using back-ward selection. Multicollinearity was as-sessed by examination of the estimatedparameter correlation matrix and nosignificant collinearity was found. Clini-cally relevant interactions were inves-tigated and none were found to be sig-nificant. Missing values were imputedto the most frequent category for cat-egorical variables and to the median forcontinuous variables. Plots of each
variable considered for the final regres-sion model against the outcome of in-terest were created to identify exces-sively influential observations and nonewere found. The fit of the final regres-sion model to the data was evaluatedby the Hosmer-Lemeshow test,24 whichpartitions the observations into 10equal-sized groups according to theirpredicted probabilities and comparesthe observed and expected mortalityacross the partitions.
Mortality rates for S aureus andhealth care–associated PVE wereevaluated by plotting the survival dis-tribution derived from Kaplan-Meierestimates, and differences in survivalwere assessed by the log-rank test.Event (or censoring) times for allpatients were measured from the dayof initial admission for PVE (time 0).For all tests, statistical significance wasdetermined at the .05 level. All statisti-cal analyses were performed using SAS
software version 8.2 (SAS Institute,Cary, NC).
RESULTSProsthetic valve endocarditis was di-agnosed in 556 (20.1%) of 2670 pa-tients with definite IE. Among pa-tients with PVE, a prosthetic aortic valvewas present in 384 patients (69.1%),prosthetic mitral valve or ring in 280patients (50.4%), prosthetic tricuspidvalve or ring in 52 patients (9.4%), andprosthetic pulmonic valve in 31 pa-tients (5.6%). Of the overall PVE co-hort, 127 (22.8%) patients died dur-ing hospital admission for PVE.
TABLE 1 shows the clinical charac-teristics of individuals with PVE in com-parison to NVIE. Patients with PVEwere significantly older, less likely touse injection drugs, and more likely tohave health care–associated infectionand intracardiac abscess. In-hospitalmortality was significantly higheramong PVE cases despite similar ratesof complications and surgical interven-tion. Although patients with PVE hada higher rate of coagulase-negativestaphylococcal infection (16.9%) anda lower rate of S aureus infection(23.0%) than patients with NVIE (8.3%and 32.9%, respectively; P�.001 forboth), S aureus was the most commoncause of PVE (TABLE 2). For 384 pa-tients for whom time from valve im-plantation was available, the microbi-ology of early and late PVE are alsoshown. During the index hospitaliza-tion, surgical therapy was used in 272(48.9%) patients with PVE and 879(46.4%) patients with NVIE (P=.30).
A comparison of PVE across geo-graphic regions is shown in TABLE 3.Direct correlations were found be-tween the percentages of PVE due tostaphylococcal species (either S au-reus or coagulase-negative staphylo-coccus as causative microorganism) andpercentages of PVE cases (r=0.713,P�.001) as well as health care–associated PVE (r=0.623, P�.001). Pa-tients in the United States were signifi-cantly more likely to have non-nosocomial health care–associated PVE,an intravascular device source of infec-
Table 1. Comparison of Patients With Prosthetic Valve Endocarditis (PVE) and Native ValveInfective Endocarditis (NVIE) in the International Collaboration on Endocarditis-ProspectiveCohort Study*
CharacteristicPVE
(n = 556)NVIE
(n = 1895)P
Value
Age, mean (range), y 65.0 (49.9-74.3) 56.3 (41.1-69.9) �.001
Male 363 (65.3) 1299 (68.6) .17
Hemodialysis dependent 25 (4.5) 173 (9.1) �.001
Diabetes mellitus 100 (18.0) 292 (15.4) .13
Current injection drug use 10 (1.8) 235 (12.4) �.001
Chronic immunosuppressive therapy 24 (4.3) 127 (6.7) .05
Cancer 32 (5.8) 169 (8.9) .02
Other chronic illness 246 (44.2) 916 (48.3) .12
Previous endocarditis 112 (20.1) 91 (4.8) �.001
Health care–associated infection 203 (36.5) 587 (31.0) .01
Transesophageal echocardiography performed 467 (84.0) 1290 (68.1) �.001
Time from admission to transesophagealechocardiography, mean (range), d
2.0 (0-6.0) 3.0 (0-7.0) .13
Echocardiographic findingsVegetation 406 (73.0) 1703 (89.9) �.001
New regurgitation 257 (46.2) 1346 (71.0) �.001
Abscess 165 (29.7) 222 (11.7) �.001
Complications and outcomeHeart failure 183 (32.9) 616 (32.5) .90
Stroke 101 (18.2) 322 (17.0) .55
Other systemic embolization 83 (14.9) 468 (24.7) �.001
Surgery during admission 272 (48.9) 879 (46.4) .30
Persistent bacteremia 49 (8.8) 166 (8.8) .92
Duration of hospitalization, mean (range), d 33 (19-49) 29 (16-44) �.001
In-hospital death 127 (22.8) 310 (16.4) �.001*Data are presented as number and percentage unless otherwise indicated.
CONTEMPORARY CLINICAL PROFILE AND OUTCOME OF PROSTHETIC VALVE ENDOCARDITIS
1356 JAMA, March 28, 2007—Vol 297, No. 12 (Reprinted) ©2007 American Medical Association. All rights reserved.
by guest on November 13, 2009 www.jama.comDownloaded from
tion, hemodialysis treatment, S aureusinfection, and higher rates of compli-cated PVE (P�.005 for all compari-sons). In-hospital mortality rates werenot statistically different across the vari-ous regions.
Health care–associated PVE wasreported in 36.5% (203/556) of theoverall cohort. Of these patients withhealth care–associated PVE, 141(69.5%) were classified as havingnosocomial infection and 62 (30.5%)were class i f ied as having non-nosocomial health care–associatedinfection. An intravascular devicesource of infection was presumed in42.9% of health care−associated infec-tions. Health care–associated PVE wascharacterized by high rates of S aureus(34.0%, including methicillin-resistantS aureus in 13.3%) and coagulase-negative staphylococci infection(25.6%) and low rates of enterococcal(9.4%) and viridans streptococcal(4.9%) infections. The in-hospitalmorta l i ty ra te o f hea l th care–associated PVE was 30.5% (62/203),including 31.9% (45/141) of patientswith nosocomial PVE who died and27.4% (17/62) of patients with non-nosocomial, health care–associatedinfection who died (P=.52).
Prosthetic valve endocarditis asso-ciated with health care contact oc-curred at a median of 83.5 days (inter-quartile range, 39.5-543.0 days) from
valve implantation, with 71% diag-nosed within 1 year of prosthetic valveimplantation. Health care–associated
PVE occurred beyond the early periodin 62.1% (87/140) of patients. In com-parison to early PVE, health care–
Table 2. Causative Organisms for Total Cohort, Early, and Late Prosthetic Valve Endocarditis(PVE)
Causative OrganismTotal, No. (%)
(n = 556)Early PVE, No. (%)
(n = 53)Late PVE, No. (%)
(n = 331)
Staphylococccus aureus 128 (23.0) 19 (35.9) 61 (18.4)
Methicillin-sensitive S aureus 82 (14.7) 8 (15.1) 43 (13.0)
Methicillin-resistant S aureus 36 (6.5) 10 (18.9) 11 (3.3)
Coagulase-negative staphylococci 94 (16.9) 9 (17.0) 66 (19.9)
Enterococcus spp 71 (12.8) 4 (7.5) 42 (12.7)
Viridans streptococci 67 (12.1) 1 (1.9) 34 (10.3)
Culture negative 62 (11.2) 9 (17.0) 41 (12.4)
Streptococcus bovis 29 (5.2) 1 (1.9) 22 (6.7)
Fungal 23 (4.1) 5 (9.4) 11 (3.3)
Polymicrobial 10 (1.8) 0 6 (1.8)
HACEK spp* 8 (1.4) 0 7 (2.1)
Escherichia coli 7 (1.3) 1 (1.9) 3 (0.9)
Streptococcus agalactiae 5 (0.9) 0 3 (0.9)
Propionibacterium acnes 4 (0.7) 0 3 (0.9)
Streptococcus group G 4 (0.7) 0 3 (0.9)
Propionibacterium NOS 3 (0.5) 0 2 (0.6)
Pseudomonas aeruginosa 3 (0.5) 1 (1.9) 1 (0.3)
Streptococcus anginosus 3 (0.5) 0 2 (0.6)
Streptococcus NOS 3 (0.5) 0 2 (0.6)
Streptococcus pneumoniae 3 (0.5) 0 3 (0.9)
Listeria monocytogenes 2 (0.4) 0 2 (0.6)
Micromonas micros 2 (0.4) 0 2 (0.6)
Mycobacteria spp 2 (0.4) 0 1 (0.3)
Serratia marcescens 2 (0.4) 1 (1.9) 0
Streptococcus gallolyticus 2 (0.4) 0 0
Streptococcus group B 2 (0.4) 0 0
Streptococcus group C 2 (0.4) 0 1 (0.3)Abbreviation: NOS, not otherwise speciated.*The HACEK group includes Haemophilus spp, Actinobacillus actinomycetemcomitans, Cardiobacterium hominis, Eikenella
corrodens, and Kingella kingae bacteria.
Table 3. Regional Comparison of Clinical Characteristics of Prosthetic Valve Endocarditis (PVE)*
OverallUnitedStates
SouthAmerica
Australia/New Zealand/
AsiaNorth/Central
Europe
Southern Europe/Middle East/South Africa
PValue
No. of sites 53 10 7 11 14 11
No. of PVE cases 556 116 66 120 109 145
PVE per total infective endocarditis cases reported, % 20.1 20.9 11.9 21.6 19.6 26.1
Early PVE, % 13.8 21.1 10.3 11.7 12.7 13.0 .38
Health care–associated infection 203 (36.5) 52 (44.8) 22 (33.3) 38 (31.7) 45 (41.3) 46 (31.7) .11
Non-nosocomial health care–associated infection 62 (11.2) 25 (21.6) 5 (7.6) 12 (10.0) 11 (10.1) 9 (6.2) .002
Presumed intravascular device source 87 (15.7) 32 (27.6) 5 (7.6) 11 (9.2) 17 (15.6) 22 (15.2) �.001
Hemodialysis 25 (4.5) 15 (12.9) 2 (3.0) 2 (1.7) 3 (2.8) 3 (2.1) �.001
Staphylococcus aureus 128 (23.0) 38 (32.8) 7 (10.6) 28 (23.3) 30 (27.5) 25 (17.3) .003
Coagulase-negative staphylococci 94 (16.9) 20 (17.2) 7 (10.6) 14 (11.7) 17 (15.6) 36 (24.8) .03
Persistent bacteremia 49 (8.8) 22 (19.0) 3 (4.6) 7 (5.8) 7 (6.4) 10 (6.9) �.001
Congestive heart failure 183 (32.9) 53 (45.7) 27 (40.9) 23 (19.2) 30 (27.5) 50 (34.5) �.001
In-hospital mortality 127 (22.8) 21 (18.1) 16 (24.2) 21 (17.5) 27 (24.8) 42 (29.0) .13*Data are presented as number and percentage unless otherwise indicated. P values are results of �2 analysis of each variable across all regions.
CONTEMPORARY CLINICAL PROFILE AND OUTCOME OF PROSTHETIC VALVE ENDOCARDITIS
©2007 American Medical Association. All rights reserved. (Reprinted) JAMA, March 28, 2007—Vol 297, No. 12 1357
by guest on November 13, 2009 www.jama.comDownloaded from
associated PVE beyond the early pe-riod was found to have a higher rate ofcoagulase-negative staphylococcal in-fection (early [17.0%] vs late, healthcare–associated PVE [39.1%]; P=.006);similar, high rates of presumed intra-vascular device source (32.1% vs 39.1%;P=.40) and S aureus infection (35.9%vs 29.9%; P=.46); and lower rates ofpersistent bacteremia (20.8% vs 5.8%;P = .007) and in-hospital mortality(47.2% vs 23.0%; P=.003).
TABLE 4 depicts the unadjusted andadjusted analyses of PVE characteris-tics related to in-hospital mortality.Goodness of fit of the multivariablemodel was found by the Hosmer-Lemeshow test (�² = 8.72, P = .27).When the composite variable, healthcare–associated infection, was re-moved from the model and the indi-vidual criteria of nosocomial infec-tion, presumed intravascular devicesource, and hemodialysis included,none was found to be independently as-sociated with mortality. Mortality re-sults for patients with S aureus PVE andhealth care–associated infection areshown in the FIGURE.
COMMENTSurgical treatment of valvular heart dis-ease is an increasingly common car-diac intervention and as a result, thenumber of patients at risk of develop-ing PVE is growing.3 In addition,changes in health care delivery may in-fluence the epidemiology of PVE. Thecurrent study, representing the largeststudy of PVE to date, offers several im-portant insights regarding this diseasein the current era.
Prosthetic valve endocarditis ac-counted for over 20% of all IE cases inthis prospective, multicenter, interna-tional registry, reflecting a consider-ably higher proportion of IE com-pared with earlier reports,9,10 but similarto the findings of a recent 1-year sur-vey of IE in France.13 An increased sen-sitivity of diagnostic modalities, includ-ing use of the Duke criteria16,25 for allcases and transesophageal echocardi-ography in 84% of cases, likely en-hanced the detection of PVE.
Also, in contrast to previous esti-mates,26 we found that S aureus, ratherthan coagulase-negative staphylococ-cal infection, was the most common
causative microorganism for PVE. Instudies of predominantly NVIE, S au-reus has been similarly found to be themost common cause of NVIE,14 and astrong association between an in-crease in S aureus IE and health carecontact has been noted.14,27,28 How-ever, few data regarding health care–associated PVE, aside from analyses ofearly PVE, have been reported and havegenerally been limited to small samplesizes and retrospective, single-center ex-periences.29,30 In 1993, an associationbetween nosocomial bacteremia andnew PVE in 18 patients was de-scribed.31 Although one third of new en-docarditis cases were attributable to in-travascular devices,31 the study wasperformed before routine diagnostic useof transesophageal echocardiographyfor evaluating staphylococcal bacter-emia. More recently, among 50 cases oflate-onset PVE at a single Spanish cen-ter, increases in staphylococcal infec-tion and late, hospital-acquired PVE in-creased were observed.30
A causal relationship between healthcare contact and PVE has been in-ferred and well described for PVE oc-curring within 60 days of valve implan-tation.20-22 Initial reports estimated thatearly PVE accounted for approxi-mately 35% to 50% of all cases ofPVE22,32,33 and noted its association withgram-positive microorganisms (par-ticularly coagulase-negative staphylo-cocci and S aureus) presumed to benosocomial in origin. Early PVE ac-counted for only 14% of all PVE casesin the ICE-PCS cohort but continuedto be associated with an extremely highmortality rate. This lower rate of earlyPVE was similar to other recent re-ports30,34 and may be related to a num-ber of factors, including improve-ments in surgical techniques, hygiene,and infection control.35
The current study demonstrates thathealth care–associated PVE beyond theearly period, representing the majorityof health care–associated infection, sig-nificantly influences the epidemiology ofPVE. In a previous investigation, Cal-derwood et al reported that the cumu-lative hazard of developing PVE was
Table 4. Relationship Between Prosthetic Valve Endocarditis Characteristics and In-HospitalMortality
Variable No.
In-HospitalMortality,No. (%)
Odds Ratio (95% CI)
Unadjusted Adjusted
Age, y�65 277 42 (15.2) 1 [Reference] 1 [Reference]
65-75 151 38 (25.2) 1.71 (1.01-2.90) 1.82 (1.09-3.03)
�75 128 47 (36.7) 3.10 (1.80-5.32) 3.73 (2.10-6.61)
Male 363 76 (20.9) 0.73 (0.47-1.12)
Diabetes mellitus 100 28 (28.0) 1.40 (0.92-2.13)
Prior infective endocarditis 112 21 (18.8) 0.74 (0.49-1.12)
Hemodialysis 25 10 (40.0) 2.31 (1.12-4.77)
Presumed intravascular device source 87 30 (34.5) 1.86 (1.03-3.38)
Health care–associated infection 203 62 (30.5) 1.83 (1.22-2.74) 1.62 (1.08-2.44)
Staphylococcus aureus 128 44 (34.4) 2.12 (1.25-3.60) 1.73 (1.01-2.95)
Coagulase-negative staphylococci 94 24 (25.5) 1.13 (0.81-1.58)
Mitral valve prosthesis 280 64 (22.9) 0.98 (0.70-1.38)
Time since valve implantationin 30-d intervals
1.00 (0.99-1.01)
Persistent bacteremia 49 27 (55.1) 1.67 (1.07-2.27) 4.29 (1.99-9.22)
Congestive heart failure 183 60 (32.8) 2.29 (1.59-3.32) 2.33 (1.62-3.34)
Intracardiac abscess 144 47 (32.6) 2.10 (1.22-3.60) 1.86 (1.10-3.15)
Stroke 101 34 (33.7) 2.10 (1.25-3.53) 2.25 (1.25-4.03)Abbreviation: CI, confidence interval.
CONTEMPORARY CLINICAL PROFILE AND OUTCOME OF PROSTHETIC VALVE ENDOCARDITIS
1358 JAMA, March 28, 2007—Vol 297, No. 12 (Reprinted) ©2007 American Medical Association. All rights reserved.
by guest on November 13, 2009 www.jama.comDownloaded from
highest within the initial 12 months af-ter valve replacement surgery, influ-enced by an increased incidence of co-agulase-negative staphylococcal infectionfrom 2 to 12 months after valve implan-tation.12 Our results confirm that the firstyear after implantation is a vulnerable pe-riod for prosthetic valve infection withboth S aureus and coagulase-negativestaphylococci. Possible explanations forthe time course of health care–associated PVE include greater expo-sure to health care contact and lack ofcomplete endothelialization of the pros-thetic valve early after implantation.
Although PVE was diagnosedpromptly and treated commonly withsurgical intervention, the morbidity andin-hospital mortality rates were veryhigh. Staphylococcus aureus and healthcare–associated infection were found tobe independently predictive of in-hospital mortality. These results con-firm earlier studies that reported a sig-nificant association between S aureusinfection and mortality in both nativeand prosthetic valve IE4,36,37 and addsprognostic significance to its high preva-lence in this study. Prosthetic valveendocarditis occurring in associationwith health care contact may repre-sent a composite of prognostic factors,such as virulence of causative organ-ism and host factors (eg, end-stage renaldisease or other conditions requiringintravascular catheter access) andemphasizes the need for preventivestrategies, particularly those related tointravascular catheter access,38,39 toreduce this potential causative factor.
The strongest predictors of mortal-ity were well-recognized complica-tions of PVE: persistent bacteremia,heart failure, intracardiac abscess, andstroke. Persistent bacteremia, as wellas health care–associated infection,has been found to be independentlyassociated with S aureus IE.14 Theseserious complications may be evidentearly in the course of the disease andare common indications for surgicalintervention. Heart failure, one of thestrongest predictors of in-hospitalmortality in the current study, hasbeen found to predict 6-month mor-
tality in NVIE40 and may identify asubgroup of patients for whom cardiacsurgery is associated with a significantsurvival benefit.41 Our previous studydemonstrated that hospital survivalrates were similar between compli-cated PVE treated with surgery anduncomplicated PVE after adjustmentfor those characteristics associatedwith surgical intervention,4 but addi-tional studies are needed to definemore clearly the role, timing, andeffect of surgery in PVE.
Although the ICE-PCS study de-sign was a large, prospective, multina-tional registry of definite IE cases, thisinvestigation has certain limitations.This is an observational study involv-ing centers with voluntary participa-tion and thus, population sampling wasnot obtained, limiting any epidemio-logical inferences. Specifically, be-
cause this cohort only included pa-tients with definite IE and not patientswith normal prosthetic valves, we wereunable to evaluate risk factors for de-veloping PVE. The definition of non-nosocomial health care–associated in-fection, although previously validatedas similar to nosocomial bloodstreaminfections with regard to source of in-fection, microbiology, antibiotic sus-ceptibility, and prognosis,23 may be im-precise, particularly regarding recenthospitalization as the presumed sourceof infection. Because each geographicregion was represented by a few cen-ters, regional characterization was basedonly on those participating centers. Fi-nally, the primary end point of mortal-ity was assessed at hospital discharge,and thus, does not reflect PVE out-come at a specific time interval from di-agnosis.
Figure. In-hospital Mortality as a Function of Causative Organism and Health CareAssociation
0.6
0.2
0.1
0.3
0.4
0.5
0
No. at RiskStaphylococcus aureusOther Organisms
0
128417
10 20 30 40 50 60 70 80 90
120 98 73 53 39 29 20
HR; 1.77 (95% CI, 1.21-2.57)Log-Rank P = .003
9 8388 316 238 175 110 78 56 41 30
Days Since Initial Hospitalization
Mor
talit
y, P
ropo
rtio
n
PVE Caused by Staphylococcus aureus or Other Organisms
Staphylococcus aureusOther Organisms
Health Care–Associated InfectionCommunity-Acquired Infection
0.6
0.2
0.1
0.3
0.4
0.5
0
No. at RiskHealth Care–Associated InfectionCommunity-Acquired Infection
0
199346
10 20 30 40 50 60 70 80 90
183 147 114 87 67 49 34
HR; 1.54 (95% CI, 1.07-2.21)Log-Rank P = .02
22 14325 267 197 141 82 58 42 28 24
Days Since Initial Hospitalization
Mor
talit
y, P
ropo
rtio
n
Health Care–Associated or Community-Acquired PVE
HR indicates hazard ratio; CI, confidence interval; PVE, prosthetic valve endocarditis.
CONTEMPORARY CLINICAL PROFILE AND OUTCOME OF PROSTHETIC VALVE ENDOCARDITIS
©2007 American Medical Association. All rights reserved. (Reprinted) JAMA, March 28, 2007—Vol 297, No. 12 1359
by guest on November 13, 2009 www.jama.comDownloaded from
In conclusion, PVE accounts for ahigh percentage of all cases of IE inmany regions of the world. Staphylo-coccus aureus is now the leading causeof PVE. Health care–associated infec-tion significantly influences the clini-cal characteristics and outcome of PVE.Despite prompt diagnosis and the com-mon use of surgical treatment in PVE,morbidity and mortality remain high,emphasizing the need for further stud-ies of preventive and therapeutic strat-egies for this serious disease.
Author Contributions: Dr Wang had full access to allof the data in the study and takes responsibility forthe integrity of the data and the accuracy of the dataanalysis.Study concept and design: Wang, Athan, Pappas,Fowler, Corey, Cabell.Acquisition of data: Athan, Fowler, Olaison, Pare,Almirante, Munoz, Rizzi, Naber, Logar, Tattevin, Iarussi,Selton-Suty, Jones, Morris.Analysis and interpretation of data: Wang, Athan,Pappas, Fowler, Tattevin, Casabe, Morris, Corey,Cabell.Drafting of the manuscript: Athan, Pappas, Fowler,Pare, Iarussi, Casabe, Morris, Corey, Cabell.Critical revision of the manuscript for important in-tellectual content: Wang, Athan, Pappas, Fowler,Olaison, Almirante, Munoz, Rizzi, Naber, Logar,Tattevin, Selton-Suty, Jones, Morris, Corey, Cabell.Statistical analysis: Wang, Pappas, Cabell.Obtained funding: Wang, Corey, Cabell.Administrative, technical, or material support: Fowler,Naber, Selton-Suty, Corey.Study supervision: Wang, Athan, Fowler, Corey.Financial Disclosures: None reported.International Collaboration on Endocarditis (ICE)Registry Investigators 2007: David Gordon, MBBS,FRACP, FRCPA, PhD, Uma Devi, MD (Flinders Medi-cal Centre, Adelaide, Australia); Denis Spelman, MD(Alfred Hospital, Amiens, France); Jan T.M. van derMeer, MD, PhD (University of Amsterdam, Amster-dam, The Netherlands); Carol Kauffman, MD, Su-zanne Bradley, MD, William Armstrong, MD (Ann Ar-bor VA Medical Center, Ann Arbor, Mich); EfthymiaGiannitsioti, MD, Helen Giamarellou, MD, PhD (At-tikon University General Hospital, Athens, Greece); Sta-matios Lerakis, MD, FAHA, FACC, FASE, FCCP (EmoryUniversity, Atlanta, Ga); Ana del Rio, MD, AsuncionMoreno, MD, Carlos A. Mestres, MD, PhD, FETCS,Carlos Pare, MD, Cristina Garcia de la Maria, MD, ElisaDe Lazzario, BSc, Francesc Marco, MD, Jose M. Ga-tell, MD, Jose M. Miro, MD, PhD, Manel Almela, MD,Manuel Azqueta, MD, Maria Jesus Jimenez-Exposito, MD, Natividad de Benito, MD, Noel Perez,MD (Hosp. Clinic-IDIBAPS: University of Barcelona,Barcelona, Spain); Benito Almirante, MD, Nuria Fern-andez-Hidalgo, MD, Pablo Rodriguez de Vera, MD,Pilar Tornos, MD, Vicente Falco, MD, Xavier Clar-amonte, MD, Yolanda Armero, MD (Hospital Univer-sitari Vall d’Hebron, Barcelona, Spain); Nisreen Si-dani, RN, MSN, Souha Kanj-Sharara, MD, FACP, ZeinaKanafani, MD, MS (American University of BeirutMedical Center, Beirut, Lebanon); Annibale Raglio, MD,DTM&H, Antonio Goglio, MD, Fabrizio Gnecchi, MD,Fredy Suter, MD, Grazia Valsecchi, MD, Marco Rizzi,MD, Veronica Ravasio, MD (Ospedali Riuniti di Ber-gamo, Bergamo, Italy); Bruno Hoen, MD, PhD, Cather-ine Chirouze, MD, Efthymia Giannitsioti, MD, Joel Le-roy, MD, Patrick Plesiat, MD, Yvette Bernard, MD,(University Medical Center of Besancon, Besancon,
France); Anna Casey, Peter Lambert, BSc, PhD, DSc,Richard Watkin, MRCP, Tom Elliott, BM, BS, BMed-Sci, PhD, DSc, FRCPath (Queen Elizabeth Hospital, Bir-mingham, England); Mukesh Patel, MD, William Dis-mukes, MD (University of Alabama at Birmingham);Angelo Pan, MD, Giampiero Caros, MD (Spedali Civili-Università di Brescia, Brescia, Italy); Amel Brahim Mathi-ron, Christophe Tribouilloy, MD, PhD, MD, ThomasGoissen, MD (South Hospital Amiens, Bron Cedex,France); Armelle Delahaye, Francois Delahaye, MD,MPH, FESC, Francois Vandenesch, MD, PhD (Hopi-tal Louis Pradel, Bron Cedex, France); Carla Vizzotti,MD, Francisco M. Nacinovich, MD, Marcelo Marin,MD, Marcelo Trivi, MD, Martin Lombardero, MD (In-stituto Cardiovascular, Buenos Aires, Argentina); Clau-dia Cortes, MD, Jose Horacio Casabe, MD (Institutode Cardiologıa y Cirugıa Cardiovascular, Buenos Aires,Argentina); Javier Altclas, MD, Silvia Kogan, MD (Sana-torio Mitre, Buenos Aires, Argentina); Liliana Clara, MD,Marisa Sanchez, MD (Hospital Italiano, Buenos Aires,Argentina); Anita Commerford, MD, Cass Hansa, MD,Eduan Deetlefs, MD, Mpiko Ntsekhe, MD, PatrickCommerford, MD (Groote Schuur Hospital, CapeTown, South Africa); Dannah Wray, MD, MHS, LisaL. Steed, PhD, Preston Church, MD, Robert Cantey,MD (Medical University of South Carolina, Charles-ton); Arthur Morris, MD, FRCPA, David Holland, MD,David Murdoch, MD, DTM&H, FRACP, FRCPA,FACTM, Katherine Graham, MD, Kerry Read, MD, Ni-gel Raymond, MD, Paul Bridgman, MD, RichardTroughton, MD, Selwyn Lang, MD, Stephen Cham-bers, MD (Canterbury Health Laboratories,Christchurch, New Zealand); Despina Kotsanas, BSc(Hons), Tony M. Korman, MD (Southern Health, Clay-ton, Australia); Gail Peterson, MD, Jon Purcell, BS, PaulM. Southern, Jr, MD (UT-Southwestern Medical Cen-ter, Dallas, Tex); Manisha Shah, MD, Roger Bedimo,MD, MS (Dallas VA Medical Center, Dallas, Tex); Ar-jun Reddy, Donald Levine, MD, Gaurav Dhar, MD(Wayne State University, Detroit, Mich); Alanna Han-lon-Feeney, Margaret Hannan, MD, BCh BAO, MSc,MRCPath, FRCPI, Sinead Kelly, MD (Mater Hospi-tals, Dublin, Ireland); Andrew Wang, MD, Chris-topher H. Cabell, MD, MHS, Christopher W. Woods,MD, MPH, Daniel J. Sexton, MD, Danny Benjamin,Jr, MD, MPH, PhD, G. Ralph Corey, MD, Jay R.McDonald, MD, John J. Engemann, MD, L. Barth Rel-ler, MD, Laura Drew, RN, BSN, Martin Stryjewski, MD,MHS, Susan Morpeth, MBChB, Tahaniyat Lalani, MD,Vance Fowler, Jr, MD, MHS, Vivian Chu, MD (DukeUniversity Medical Center, Durham, NC); BahramMazaheri, PhD, Carl Neuerburg, Christoph Naber, MD(University Essen, Essen, Germany); Eugene Athan,MD, Margaret Henry, BSc (Hons), PhD, Owen Har-ris, MD (Barwon Health, Geelong, Australia); Eric Al-estig, MD, Lars Olaison, MD, PhD, Lotta Wikstrom,Ulrika Snygg-Martin, MD (Sahlgrenska Univer-sitetssjukhuset/Ostra, Göteborg, Sweden); JohnsonFrancis, MD, DM, K. Venugopal, MD, DM, Lathi NairMD, DM, Vinod Thomas, MD, DM (Medical CollegeCalicut, Kerla, India); Jaruwan Chaiworramukkun, MD,Orathai Pachirat, MD, Ploenchan Chetchotisakd, MD,Tewan Suwanich, MD (Khon Kaen University, KhonKaen, Thailand); Adeeba Kamarulzaman, MBBS,FRACP, Syahidah Syed Tamin, MD (University of Ma-laya Medical Center, Kuala Lumpur, Malaysia); ManicaMueller Premru, MD, PhD, Mateja Logar, MD, PhD,Tatjana Lejko-Zupanc, MD, PhD (Medical Center Ljub-lijana, Ljublijana, Slovenia); Christina Orezzi, John Klein,MD (St Thomas’ Hospital, London, England); EmilioBouza, MD, PhD, Mar Moreno, MD, PhD, Marta Ro-drıguez-Creixems, MD, PhD, Mercedes Marın, MD,Miguel Fernandez, MD, Patricia Munoz, MD, PhD, Ro-cıo Fernandez, Victor Ramallo, MD (Hospital Gen-eral Universitario Gregorio Maranon, Madrid, Spain);Didier Raoult, MD, PhD, Franck Thuny, MD, GilbertHabib, MD, FACC, FESC, Jean-Paul Casalta, MD,Pierre-Edouard Fournier, MD (Faculte de Medecine de
Marseille, Marseille, France); Natalia Chipigina, PhD,Ozerecky Kirill, MD, Tatiana Vinogradova, MD, PhD,Vadim P. Kulichenko, PhD (Russian Medical State Uni-versity, Moscow, Russia); O.M. Butkevich, PhD (Learn-ing Medical Centre of Russian Presidential Affairs Gov-ernment, Moscow, Russia); Christine Lion, MD,Christine Selton-Suty, MD, Francois Alla, MD, PhD,Helène Coyard, Thanh Doco-Lecompte, MD (CHUNancy-Brabois, Nancy, France); Diana Iarussi, MD,Emanuele Durante-Mangoni, MD, PhD, Marie Fran-coise Tripodi, MD, Riccardo Utili, MD (II Università diNapoli, Naples, Italy); A. Sampath Kumar, MD, Gau-tam Sharma, MD (All India Institute of Medical Sci-ences, New Delhi, India); Stuart A. Dickerman, MD(New York University Medical Center, New York, NY);Alan Street, Damon Peter Eisen, MBBS, MD, FRACP,Emma Sue McBryde, MBBS, FRACP, PhD, LeeanneGrigg (Royal Melbourne Hospital, Parkville, Austra-lia); Elias Abrutyn, MD (Drexel University College ofMedicine, Philadelphia, Pa); Christian Michelet, MD,PhD, Pierre Tattevin, MD, Pierre Yves Donnio, PhD(Pontchaillou University, Rennes, France); ClaudioQuerido Fortes, MD (Hospital Universitario Clem-entino Fraga Filho/UFRJ, Rio de Janeiro, Brazil); Ja-meela Edathodu, MRCP, Mashael Al-Hegelan, MD(King Faisal Specialist Hospital & Research Center, Ri-yadh, Saudi Arabia); Bernat Font, MD, Ignasi An-guera, MD, PhD, Joan Raimon Guma, MD (Hospitalde Sabadell, Sabedell, Spain); M. Cereceda, MD, MiguelJ. Oyonarte, MD, Rodrigo Montagna Mella, MD (Hos-pital Clinico Universidad de Chile, Santiago, Chile); Pa-tricia Garcia, MD, Sandra Braun Jones, MD (Hosp.Clınico Pont. Universidad Catolica de Chile, San-tiago, Chile); Auristela Isabel de Oliveira Ramos, MD,(Instituto Dante Pazzanese de Cardiologia, Sao Paulo,Brazil); Marcelo Goulart Paiva, MD, Regina Apare-cida de Medeiros Tranchesi, MD (Hospital 9 de Ju-lho, Sao Paulo, Brazil); Lok Ley Woon, BSN, Luh-NahLum, BSN, Ru-San Tan, MBBS, MRCP (National HeartCentre, Singapore, Singapore); David Rees, MD, PamKornecny, MD, Richard Lawrence, MD, Robyn De-ver, MD (St. George Hospital, Sydney, Australia); Jef-frey Post, MD, Phillip Jones, MD, Suzanne Ryan, MHSc,GCDM (The University of New South Wales, Syd-ney, Australia); John Harkness, MD, Michael Fene-ley, MD (St. Vincent’s, Sydney, Australia); Ethan Ru-binstein MD, LLB, Jacob Strahilewitz, MD (Tel AvivUniversity School of Medicine, Tel Aviv, Israel); AdinaIonac, MD, PhD, Cristian Mornos, MD, Stefan Dra-gulescu, MD, PhD (Victor Babes University of Medi-cine and Pharmacy, Timisoar, Romania); Davide Forno,MD, Enrico Cecchi, MD, Francesco De Rosa, MD, Mas-simo Imazio, MD, FESC, Rita Trinchero, MD (MariaVittoria Hospital, Torino, Italy); Franz Wiesbauer, MD,Rainer Gattringer, MD (Vienna General Hospital, Vi-enna, Austria); Ethan Rubinstein, MD, LLB, Greg Deans,MD (University of Manitoba, Winnipeg, Manitoba);Arjana Tambic Andrasevic, MD, PhD, Bruno Barsic,MD, PhD, Igor Klinar, MD, Josip Vincelj, MD, PhD,FESC, Suzana Bukovski, MD, Vladimir Krajinovic, MD(Univ. Hospital for Infectious Diseases, Zagreb, Croatia).None of the ICE Investigators received compensa-tion for their work in the study.ICE Coordinating Center (Durham, NC): ChristopherCabell, MD, MHS, Judy Stafford, MS, Khaula Baloch,BA, Paul Pappas, MS, Thomas Redick, MPH, Tina Har-ding, RN, BSN.ICE Steering Committee: A.W. Karchmer, MD, ArnieBayer, MD, Bruno Hoen, MD, PhD, Christopher H. Ca-bell, MD, MHS, Daniel J. Sexton, MD, David T. Du-rack, MD, DPhil, FACP, FRCP, FRACP, Elias Abrutyn,MD, Ethan Rubinstein, MD, LLB, G. Ralph Corey, MD,Jose M. Miro, MD, PhD, Phillipe Moreillon, Susan-nah Eykyn, Vance Fowler, Jr, MD, MHS, Lars Olai-son, MD, PhD.ICE Publications Committee: Arnie Bayer, MD, BrunoHoen, MD,PhD, Christopher H. Cabell, MD, MHS, EliasAbrutyn, MD, Eugene Athan, MD, Jose M. Miro, MD,
CONTEMPORARY CLINICAL PROFILE AND OUTCOME OF PROSTHETIC VALVE ENDOCARDITIS
1360 JAMA, March 28, 2007—Vol 297, No. 12 (Reprinted) ©2007 American Medical Association. All rights reserved.
by guest on November 13, 2009 www.jama.comDownloaded from
PhD, G. Ralph Corey, MD, Paul Pappas, MS, VanceFowler, Jr, MD, MHS.Funding/Support: This study was funded by an Ameri-can Heart Association Grant-in-Aid (0455802U) to DrWang.Role of the Sponsor: The American Heart Associa-tion had no role in the design and conduct of the study;in the collection, management, analysis, and inter-pretation of the data; or in the preparation, review,or approval of the manuscript.
REFERENCES
1. Ghosh P, Unger F. Cardiac surgery and catheterbased coronary interventions in Europe 2002. Car-diovascular Forum Online Web site. http://www.ehi.at/journal/index.htm. Accessed February 15, 2007.2. Vongpatanasin W, Hillis LD, Lange RA. Prostheticheart valves. N Engl J Med. 1996;335:407-416.3. Darouiche RO. Treatment of infections associ-ated with surgical implants. N Engl J Med. 2004;350:1422-1429.4. Wang A, Pappas P, Anstrom KJ, et al. The use andeffect of surgical therapy for prosthetic valve infec-tive endocarditis: a propensity analysis of a multi-center, international cohort. Am Heart J. 2005;150:1086-1091.5. Akowuah EF, Davies W, Oliver S, et al. Prostheticvalve endocarditis: early and late outcome followingmedical or surgical treatment. Heart. 2003;89:269-272.6. Edwards MB, Ratnatunga CP, Dore CJ, Taylor KM.Thirty-day mortality and long-term survival follow-ing surgery for prosthetic endocarditis: a study fromthe UK heart valve registry. Eur J Cardiothorac Surg.1998;14:156-164.7. Tornos P, Almirante B, Olona M, et al. Clinical out-come and long-term prognosis of late prosthetic valveendocarditis: a 20-year experience. Clin Infect Dis.1997;24:381-386.8. Grover FL, Cohen DJ, Oprian C, Henderson WG,Sethi G, Hammermeister KE. Determinants of the oc-currence of and survival from prosthetic valve endo-carditis: experience of the Veterans Affairs Coopera-tive Study on Valvular Heart Disease. J ThoracCardiovasc Surg. 1994;108:207-214.9. Varstela E. Personal follow-up of 100 aortic valvereplacement patients for 1081 patient years. Ann ChirGynaecol. 1998;87:205-212.10. Sidhu P, O’Kane H, Ali N, et al. Mechanical or bio-prosthetic valves in the elderly: a 20-year comparison.Ann Thorac Surg. 2001;71(suppl):S257-S260.11. Vlessis AA, Khaki A, Grunkemeier GL, Li HH, StarrA. Risk, diagnosis and management of prosthetic valveendocarditis: a review. J Heart Valve Dis. 1997;6:443-465.
12. Calderwood SB, Swinski LA, Waternaux CM,Karchmer AW, Buckley MJ. Risk factors for the de-velopment of prosthetic valve endocarditis. Circulation.1985;72:31-37.13. Hoen B, Alla F, Selton-Suty C, et al. Changing pro-file of infective endocarditis: results of a 1-year sur-vey in France. JAMA. 2002;288:75-81.14. Fowler VG Jr, Miro JM, Hoen B, et al. Staphylo-coccus aureus endocarditis: a consequence of medi-cal progress. JAMA. 2005;293:3012-3021.15. Cabell CH, Jollis JG, Peterson GE, et al. Chang-ing patient characteristics and the effect on mortalityin endocarditis. Arch Intern Med. 2002;162:90-94.16. Li JS, Sexton DJ, Mick N, et al. Proposed modi-fications to the Duke criteria for the diagnosis of in-fective endocarditis. Clin Infect Dis. 2000;30:633-638.17. Durack DT, Lukes AS, Bright DK. New criteria fordiagnosis of infective endocarditis: utilization of spe-cific echocardiographic findings. Am J Med. 1994;96:200-209.18. Cabell CH, Abrutyn E. Progress toward a globalunderstanding of infective endocarditis: early lessonsfrom the International Collaboration on Endocarditisinvestigation. Infect Dis Clin North Am. 2002;16:255-272.19. Daniel WG, Mugge A, Martin RP, et al. Improve-ment in the diagnosis of abscesses associated with en-docarditis by transesophageal echocardiography. N EnglJ Med. 1991;324:795-800.20. Slaughter L, Morris JE, Starr A. Prosthetic valvu-lar endocarditis: a 12-year review. Circulation. 1973;47:1319-1326.21. Block PC, DeSanctis RW, Weinberg AN, AustenWG. Prosthetic valve endocarditis. J Thorac Cardio-vasc Surg. 1970;60:540-548.22. Wilson WR, Jaumin PM, Danielson GK, GiulianiER, Washington JA II, Geraci JE. Prosthetic valveendocarditis. Ann Intern Med. 1975;82:751-756.23. Friedman ND, Kaye K, Stout J, et al. Health care-associated blood stream infections in adults: a reasonto change the accepted definition of community-acquired infections. Ann Intern Med. 2002;137:791-797.24. Lemeshow S, Hosmer DW Jr. A review of good-ness of fit statistics for use in the development of lo-gistic regression models. Am J Epidemiol. 1982;115:92-106.25. Perez-Vazquez A, Farinas MC, Garcia-Palomo JD,Bernal JM, Revuelta JM, Gonzalez-Macias J. Evalua-tion of the Duke criteria in 93 episodes of prostheticvalve endocarditis: could sensitivity be improved? ArchIntern Med. 2000;160:1185-1191.26. Mylonakis E, Calderwood SB. Infective endocar-ditis in adults. N Engl J Med. 2001;345:1318-1330.27. Cheng A, Athan E, Appelbe A, McDonald M. Thechanging profile of bacterial endocarditis as seen at
an Australian provincial centre. Heart Lung Circ. 2002;11:26-31.28. Martın-Davila P, Fortun J, Navas E, et al. Noso-comial endocarditis in a tertiary hospital: an increas-ing trend in native valve cases. Chest. 2005;128:772-779.29. Castillo JC, Anguita MP, Torres F, et al. Long-term prognosis of early and late prosthetic valveendocarditis. Am J Cardiol. 2004;93:1185-1187.30. Rivas P, Alonso J, Moya J, de Gorgolas M, Mar-tinell J, Fernandez Guerrero ML. The impact of hospital-acquired infections on the microbial etiology and prog-nosis of late-onset prosthetic valve endocarditis. Chest.2005;128:764-771.31. Fang G, Keys TF, Gentry LO, et al. Prosthetic valveendocarditis resulting from nosocomial bacteremia: aprospective, multicenter study. Ann Intern Med. 1993;119:560-567.32. Dismukes WE, Karchmer AW, Buckley MJ, Aus-ten WG, Swartz MN. Prosthetic valve endocarditis:analysis of 38 cases. Circulation. 1973;48:365-377.33. Masur H, Johnson WD Jr. Prosthetic valveendocarditis. J Thorac Cardiovasc Surg. 1980;80:31-37.34. Truninger K, Attenhofer Jost CH, Seifert B, et al.Long term follow up of prosthetic valve endocarditis:what characteristics identify patients who were treatedsuccessfully with antibiotics alone? Heart. 1999;82:714-720.35. Geraci JE, Dale AJ, McGoon GD. Bacterial endo-carditis and endarteritis following cardiac operations.Wis Med J. 1963;62:302-315.36. Chu VH, Cabell CH, Benjamin DK Jr, et al. Earlypredictors of in-hospital death in infective endocarditis.Circulation. 2004;109:1745-1749.37. Kuyvenhoven JP, van Rijk-Zwikker GL, HermansJ, Thompson J, Huysmans HA. Prosthetic valve en-docarditis: analysis of risk factors for mortality. Eur JCardiothorac Surg. 1994;8:420-424.38. Maki DG, Kluger DM, Crnich CJ. The risk of blood-stream infection in adults with different intravasculardevices: a systematic review of 200 published pro-spective studies. Mayo Clin Proc. 2006;81:1159-1171.39. O’Grady NP, Alexander M, Dellinger EP, et al.Guidelines for the prevention of intravascular catheter-related infections. Infect Control Hosp Epidemiol. 2002;23:759-769.40. Hasbun R, Vikram HR, Barakat LA, BuenconsejoJ, Quagliarello VJ. Complicated left-sided native valveendocarditis in adults: risk classification for mortality.JAMA. 2003;289:1933-1940.41. Vikram HR, Buenconsejo J, Hasbun R, Quagliar-ello VJ. Impact of valve surgery on 6-month mortal-ity in adults with complicated, left-sided native valveendocarditis: a propensity analysis. JAMA. 2003;290:3207-3214.
CONTEMPORARY CLINICAL PROFILE AND OUTCOME OF PROSTHETIC VALVE ENDOCARDITIS
©2007 American Medical Association. All rights reserved. (Reprinted) JAMA, March 28, 2007—Vol 297, No. 12 1361
by guest on November 13, 2009 www.jama.comDownloaded from
