0
Special Features |

A Systematic Review of Population-Based Studies of Infective Endocarditis* FREE TO VIEW

Imad M. Tleyjeh, MD, MSc; Ahmed Abdel-Latif, MD; Hazim Rahbi, MBChB; Christopher G. Scott, MSc; Kent R. Bailey, PhD; James M. Steckelberg, MD; Walter R. Wilson, MD; Larry M. Baddour, MD
Author and Funding Information

*From the Divisions of Infectious Diseases (Drs. Tleyjeh, Steckelberg, Wilson, and Baddour) and Biostatistics (Mr. Scott and Dr. Bailey), Mayo Clinic College of Medicine, Mayo Clinic, Rochester, MN; Department of Medicine (Dr. Abdel-Latif), Division of Cardiology, University of Louisville, Louisville, KY; and Leicester Medical School (Dr. Rahbi), University of Leicester, Leicester, UK.

Correspondence to: Imad M. Tleyjeh, MD, MSc, Department of Medicine, King Fahd Medical City, Riyadh, Saudi Arabia, Main Hospital, PO Box 59046, Riyadh, 11525, KSA; e-mail: tleyjeh.imad@mayo.edu



Chest. 2007;132(3):1025-1035. doi:10.1378/chest.06-2048
Text Size: A A A
Published online

Background: We sought to summarize and critically appraise the literature on the epidemiology of infective endocarditis (IE) in the general population.

Methods: We retrieved population-based IE surveys by searching MEDLINE and EMBASE. Two reviewers independently extracted relevant data. We performed a metaregression to determine if temporal trends of IE characteristics exist.

Results: Fifteen population-based investigations with 2,371 IE cases from seven countries (Denmark, France, Italy, the Netherlands, Sweden, United Kingdom, and United States) from 1969 to 2000 were eligible. Different case definitions and procedures were used to capture all IE cases, including census of existing diagnoses, record-linkage system, and direct contact survey. In the unadjusted regression, there was a decline in the proportion of IE patients with underlying rheumatic heart disease (RHD; 12%; 95% confidence interval [CI], − 21 to − 3%; p = 0.01) and an increase in the proportion of patients undergoing valve surgery (9%; 95% CI, 3 to 16%) per decade. After adjusting for country, the decline in IE cases with underlying RHD became nonsignificant, but the proportions of IE patients undergoing valve surgery increased 7% per decade (95% CI, − 4 to 14%; p = 0.06), and those with underlying prosthetic valve increased 7% per decade (95% CI, − 1 to 16%; p = 0.07). There were no significant temporal trends in the causative organisms.

Conclusion: Evidence from well-planned, representative IE epidemiologic surveys is scarce in many countries. Available studies suggest a changing distribution of underlying valvular heart disease in patients with IE and an increase in its surgical treatment.

Figures in this Article

The clinical features of infective endocarditis (IE) as described by Osler have undergone many changes in developed countries.12 IE was a disease that commonly affected patients with predisposing valvular abnormalities caused by rheumatic carditis, and viridans group streptococci were the most common pathogens. This presentation is still seen in developing countries where rheumatic heart disease (RHD) is prevalent. This group of at-risk patients is, however, being surpassed by new at-risk groups, including injection drug users, elderly people with valvular sclerosis, patients with cardiovascular prostheses, those with nosocomial exposure, and hemodialysis patients. Several studies2from passively reported case series suggest that Staphylococcus aureus is currently the most frequently identified pathogen. The apparent increase in the proportion of IE cases due to S aureus in institutional experiences from referral centers may be artifactual, due to changes in medical practice that affect likelihood and patterns of referral.3 Population-based investigations, in contrast, are more reliable in characterizing the epidemiology of IE because they avoid many of the biases that plague institutional surveys.

We performed a systematic review of all population-based IE surveys in order to more appropriately define the epidemiology of this disease. We also performed a metaregression to determine if temporal trends of IE characteristics exist. We outline the methodologic issues that we encountered and suggest directions for future research.

Our systematic review included the following procedures: (1) establishment of a complete series of existing population-based IE surveys; (2) collection of methodologic details and results; (3) selection of well-designed studies; and (4) synthesis of findings that characterize IE epidemiology and its temporal trends.

Identification of Relevant Literature

A literature search was performed to identify all population-based IE surveys with the help of an experienced medical librarian. The search included MEDLINE, and EMBASE databases from inception until June 4, 2005. We searched the MEDLINE database using the search terms endocarditis, registry, and population surveillance. There was no restriction to language of publication. There were two studies in Danish, one in French, one in Italian, and one in Spanish that were eligible for our review. Physicians translated these articles and abstracted relevant data after discussion with one of the investigators (I.M.T.). Two investigators (I.M.T. and H.R.) independently reviewed abstracts of all identified references. Any study that could be relevant, based on findings that were described in the respective abstract, was reviewed in full text. We manually reviewed bibliographies of retrieved articles for additional citations and obtained the full text of all potentially relevant articles. We did not seek unpublished investigations.

A study had to be a population-based survey of IE, as defined below, to be included. We excluded studies restricted to specific subgroups such as injection drug users, patients with valvular heart disease, or patients with pacemaker endocarditis. For work published in several reports and/or languages that involved the same patient population, we included the report that had the most complete data. Surveys originating from the same population were included in the review part, and data for different time periods were extracted for the trend analysis.

Data Collection

A data collection form was developed and used to retrieve information on relevant clinical features and results of pertinent studies. Two reviewers (I.M.T. and A.A.L.) independently extracted data. Data included the following items: variables on the anatomy of the survey and different IE characteristics (sample size, demographics, incidence, proportions with underlying heart disease and injection drug users, proportions of different microorganisms, and outcome).

Assessment of Study Quality

Two reviewers (I.M.T. and A.A.L.) independently rated the methodologic quality of each study. We assessed the quality of each population-based survey based on four key features: adequacy of population definition, sampling techniques, disease definition, and completeness of case ascertainment.

We deemed the population definition to be inadequate if the residency status of all IE patients was not ascertained. Optimal sampling techniques include complete enumeration or random-sampling techniques. For the purpose of our study, we did not consider a specific case definition to be superior because case definitions have changed over time with the advent of echocardiography and improvement in blood culture techniques. Adequacy of case ascertainment was assessed on the basis of case-finding procedures, inclusion of postmortem diagnoses, and number of hospitals serving the population under study that participated in the study. Author statements about shortfall in case ascertainment were also considered an indication of inadequate case ascertainment. On the basis of these criteria, we excluded studies that had considerable shortfalls in case ascertainment and/or lacked a case definition. Reviewer disagreements were resolved by consensus after review of the article.

Statistical Analysis

We supplemented our systematic review with metaregression analyses to test the hypothesis that there are global temporal trends in IE characteristics by using weighted linear regression (metaregression) models. We included only the 15 best-quality studies in our analysis. The study by Nolsoe et al4 did not contribute results because it did not include data on all IE characteristics. Long-term studies were grouped in 5-year periods, when data were available, in order to better define the rates in each time frame (data available on request). For all studies, the time point used in the metaregression analysis and plots was taken as the midpoint of the study time interval or subinterval for the long-term studies. For example, if the time interval for a study was from 1980 to 1985, the time point used was 1982. Binomial SEs were calculated for each time period, and the reciprocals of these SEs were used as weights in the regression analysis. The analyses attempted to remove confounding of time period and country. This was done for each characteristic by assuming similar time trends across countries, but allowing the intercept to vary for each country. Therefore, country was included as a fixed effect in each regression model, and country-specific intercepts for the baseline year in the analysis, 1970, are reported. Because similar time trends are assumed for each country, the intercept gives the relative position of each country in the analysis. F tests were used to determine the overall significance of country effects in these metaregression analyses. Bubble plots are shown in Figures 2, 3 . In these plots, only studies from countries with multiple time points, and therefore contributing to the estimated slope in the adjusted analysis, are shown. Bubble sizes in these plots are proportional to weights used in the analysis, and the regression lines presented have an intercept that is calculated as the average across individual countries. Analyses were performed using statistical software (SAS version 8.2; SAS Institute; Cary, NC).

Eligible Studies

Our search identified 466 studies. Figure 1 outlines the selection process of eligible studies. Ultimately, 15 population-based investigations from seven different countries (Denmark, France, Italy, the Netherlands, Sweden, United Kingdom, and United States) were selected for inclusion. Table 1 summarizes the characteristics of identified studies, including population characteristics and study design.

Different procedures were used to capture all IE cases, including census of existing diagnoses, record-linkage system, and direct contact survey or a combination of these. Census of existing diagnoses was performed by searching one or more of the following: hospital diagnostic codes, autopsy registries, death certificates, national or county registries of discharge diagnoses, and records of specialized units. A record-linkage system that linked all diagnoses for the population it serves was only available in Olmsted County, MN. Sixteen studies used a direct contact survey design and involved sending a questionnaire to physicians and/or microbiologists that asked them to report relevant information on any confirmed and suspected IE case. Surveillance systems were applied in three surveys in which they actively recruited any suspected IE patient in participating hospitals.

Quality Assessment

Eleven studies were excluded. Ten of these studies1928 had considerable shortfall in reporting or inadequate/unclear case-finding procedures. Nine of the 11 studies used the direct contact survey that led to inadequate case ascertainment. The other studies that used direct contact survey were more successful because they either supplemented it with active search for cases or they recruited more specialties to participate in the study and used active follow-up visits or reminders during the study period. One study29 used discharge codes and did not validate the diagnoses or checked for duplicates.

IE Characteristics and Temporal Trends

Table 2 summarizes the incidence, demographics, and characteristics of IE in different surveys.

Individual Studies That Examined Temporal Trends:

The only study18 designed to examine temporal trends of IE incidence and characteristics was the one conducted in Olmsted County, MN. The results of another survey10 from France restricted to the regions of Ile de France, Lorraine, and Rhone-Alpes were compared to an earlier survey9 from 1990 of the same regions. Table 3 summarizes the main findings of these two studies.

Metaregression:

Table 4 summarizes the results of different metaregression analyses based on 14 of the 15 included studies (Table 1); the study by Nolsoe et al4 was omitted from the analysis because it did not include data on all IE characteristics. Because country effects were considered to be important to control for, all analyses are presented as both unadjusted and after adjusting for country.

In the unadjusted regression analysis, there was an overall decline in the proportion of IE patients with underlying RHD and an increase in the proportion of patients undergoing valve surgery, the point estimates being a decrease of 12% (95% CI, − 21 to − 3%; p = 0.01) and an increase of 9% (95% CI, 3 to 16%) per decade, respectively.

There was significant country effect for the rheumatic heart disease variable (p value for F test = 0.05). This country effect refers to the estimated intercept (the proportion in 1970) for the each country and indicates where each country lies relative to the others. After adjusting for country, the decline in IE patients with underlying RHD became nonsignificant, but the proportions of IE patients undergoing valve surgery and those with underlying prosthetic valves both showed both a 7%-per-decade increase (95% CI, − 4 to 14%; p = 0.06; and 95% CI, − 1 to 16%; p = 0.07, respectively). There were no significant temporal trends in the proportion of IE cases due to viridans group streptococci or with underlying mitral valve prolapse. There did appear to be weak-to-moderate evidence of country effects when examining the temporal trends among most characteristics of IE. Distribution of proportions across countries and results of metaregression analyses for temporal trends in the proportions of viridans group streptococci and S aureus are shown in Figures 2, 3.

Our systematic review highlights several important aspects of IE epidemiology in the general population and the different limitations that have distorted the findings of available IE population-based surveys. It reveals significant variability and between-country heterogeneity in IE characteristics. Furthermore, it does not support the current belief that IE is undergoing a metamorphosis.

First, although all identified studies attempted to describe the epidemiology of IE in a given population, many failed to ascertain the residency status of all cases; failure to do so can lead to inclusion of referral cases. Second, the majority of identified studies intended complete enumeration of all IE cases in the population to avoid selection bias; however, many had considerable shortfall in case identification. Third, failure to use a validated case definition by several studies may have contributed to disease misclassification by including false-positive cases and excluding false-negative cases. Fourth, case ascertainment was inadequate in several studies.

There are advantages and disadvantages of different study designs. The direct contact survey design has some advantages over searching of existing diagnoses because it can cover a larger population and include a larger sample size. Nevertheless, this design is prone to underreporting and selection bias because of the voluntary nature of participation. The extent of underreporting is difficult to assess unless investigators compare survey results to those from searching existing diagnoses in the same population at the same time. Moreover, survey design may fail to include culture-negative IE cases or cases diagnosed on postmortem examination. Census of existing diagnoses, if performed in a comprehensive manner, is more reliable in capturing all cases; however, it is also prone to underestimation of the true number of cases if diagnostic coding procedures are not complete. Given the seriousness of IE, it is less likely that cases are not reported in discharge codes. Census of existing diagnoses relies on abstracting data retrospectively from medical records and is prone to missing or inaccurate information and lack of standard definitions, especially when describing underlying heart disease. In addition, this procedure is most often used in a small population and is limited by the small sample size and limited generalizability of the results. Record-linkage systems provide a unique opportunity to study IE epidemiology because they are more reliable in capturing all cases; however, they are of limited availability worldwide and can be affected by small sample size, information bias, and limited generalizability.

We observed significant between-country heterogeneity in IE characteristics as manifested by the wide variability in the proportions of IE cases with a specific underlying heart disease or caused by a specific organism. There was also an important country effect on the results of the adjusted metaregression analysis. This between-country heterogeneity may be related to different study design, case definition and ascertainment, and different populations at risk, and thus it precludes a definitive interpretation of the metaregression analyses to study global temporal trends in IE. This type of analysis is prone to bias because it compares different populations at different times. There was also a spread of studies across different countries, and there were limited observations to detect a significant effect with metaregression while controlling for country. Nevertheless, the directions of temporal trends obtained from these analyses are clinically plausible.

Results of the metaregression analyses and one individual temporal trend study15 suggest an increase in IE cases with underlying prosthetic value, a nonsignificant increase in cases with underlying mitral valve prolapse and a nonsignificant decrease in IE cases with underlying RHD, in developed countries. Although these trends did not reach statistical significance due to the limited number of studies in the meta-regression analyses, the direction of the trends are plausible. The decline in RHD likely accounts for the decline in the proportion of IE cases with this as a predisposing cardiac condition. The apparent increase in IE cases with mitral valve prolapse could be due to detection bias with the more frequent use of echocardiography.

Although the results of the metaregression analyses suggest no overall temporal trends in S aureus or viridans group streptococcal IE, a heterogeneous pattern of temporal change in IE-causative organisms exists across different countries. The direction of the overall trend was an increase in viridans group streptococcal IE and a decrease in S aureus. The proportions of S aureus and viridans group streptococcal IE ranged from 12.2 to 33.9% and 6.8 to 45% among 15 studies with no major limitations.

Our review provides the first summary of IE population-based studies based on findings from different geographic and temporal investigations. We performed a comprehensive search and included studies published in any language to capture surveys from different countries. Moreover, we critically appraised the quality of included studies to make our assessment and to help guide the future development of well-designed studies to investigate global characteristics of IE.

Despite the thoroughness of this systematic review, limitations to our understanding of global IE epidemiology remain. First, all population-based studies are prone to incomplete case ascertainment or disease misclassification. Second, different IE case definitions used over time and across different studies limit our ability to examine temporal trends or geographic patterns of IE. Third, many regions in the world are underrepresented in the IE literature. Several IE studies addressing IE epidemiology have been published from different countries, but only very few were population based. All 15 well-designed studies were from developed countries. Fourth, even within one country, few regions are represented. For example, in the United States, the only IE population-based study was from Olmsted County, MN. The relative uniformity of the racial and ethnic composition of Olmsted County potentially limits the ability to generalize the findings of the study to groups underrepresented in the population within and outside the United States. Finally, the country adjustment in the metaregression analysis lessens the power to detect trends; nevertheless, given the fact that we have different countries represented at different time points, it is important to take these country effects into account. This emphasizes the need for additional, large, and well-executed studies to further define these issues. Coordinated multinational studies covering the same time span and using the same methodology are particularly useful.

Our review provides the first summary of IE population-based studies from different geographic and temporal investigations. Evidence from well-planned representative epidemiologic surveys is scarce in many countries. The available evidence does not support the contention that IE epidemiology is undergoing major changes. More appropriately designed studies from different countries are needed to expand our understanding of the global epidemiology of IE.

Abbreviations: CI = confidence interval; IE = infective endocarditis; RHD = rheumatic heart disease

Figure Jump LinkFigure 2. Top, A: Distribution of proportions of S aureus IE among culture-positive IE cases across different countries: X-axis (year), Y-axis (proportion of viridans streptococcal IE among culture-positive IE cases). Each country is represented by a different shape and color. Bottom, B: Metaregression analyses: overall temporal trends in S aureus IE among culture-positive IE cases: X-axis (year), Y-axis (proportion of S aureus IE among culture-positive IE cases). Only studies from countries with multiple time points, and therefore contributing to the estimated slope in the adjusted analysis, are shown. Bubble sizes in these plots are proportional to weights used in the analysis, and the regression lines presented have an intercept that is calculated as the average across individual countries.Grahic Jump Location
Figure Jump LinkFigure 3. Top, A: Distribution of proportions of viridans streptococcal IE among culture-positive IE cases across different countries: X-axis (year), Y-axis (proportion of viridans streptococcal IE among culture-positive IE cases). Each country is represented by a different shape. Bottom, B: Metaregression analyses: overall temporal trends in viridans streptococcal IE among culture-positive IE cases: X-axis (year), Y-axis (proportion of viridans streptococcal IE among culture-positive IE cases). Only studies from countries with multiple time points, and therefore contributing to the estimated slope in the adjusted analysis, are shown. Bubble sizes in these plots are proportional to weights used in the analysis and the regression lines presented have an intercept which is calculated as the average across individual countries.Grahic Jump Location
Figure Jump LinkFigure 1. Flow diagram of eligible studies.Grahic Jump Location
Table Graphic Jump Location
Table 1. Study and Population Characteristics of Eligible Studies
Table Graphic Jump Location
Table 1A. Continued

* N/A = not applicable; D = definite; D/P = definite/possible; D/P/P = definite/probable/possible; D/P/P/R = definite/probable/possible/rejected; CTS = cardiothoracic surgeons; ID = infectious diseases physicians.

 

Durack et al.31

 

Li et al.32

Table Graphic Jump Location
Table 2. Demographics and Clinical Characteristics of Eligible Studies*
* 

IVDA = IV drug abuse.

 

Data on 74 of 80 subjects.

 

1.7 in the first 5 years and 3.6 the next 5 years.

§ 

Average incidence.

 

Male/female gender, No.

Table Graphic Jump Location
Table 3. Individual Studies That Examined Temporal Trends*
* 

UHD = underlying heart disease.

Table Graphic Jump Location
Table 4. Results of Metaregression Analyses for Temporal Trends of IE Characteristics*
* 

Data are presented as change in proportion (95% CI). NA = data not available.

We thank Bjorg Thorsteinsdottir, MD, for help with translating non-English studies.

Mylonakis, E, Calderwood, SB (2001) Infective endocarditis in adults.N Engl J Med345,1318-1330. [PubMed] [CrossRef]
 
Moreillon, P, Que, YA Infective endocarditis.Lancet2004;363,139-149. [PubMed]
 
Steckelberg, JM, Melton, LJ, III, Ilstrup, DM, et al Influence of referral bias on the apparent clinical spectrum of infective endocarditis.Am J Med1990;88,582-588. [PubMed]
 
Nolsoe, C, Jensen, LT, Hartzner, K, et al Occurrence of infectious endocarditis in Denmark.Ugeskr Laeger1987;149,3458-3459. [PubMed]
 
Nissen, H, Nielsen, PF, Frederiksen, M, et al Native valve infective endocarditis in the general population: a 10-year survey of the clinical picture during the 1980s.Eur Heart J1992;13,872-877. [PubMed]
 
Benn, M, Hagelskjer, LH, Tvede, M Infective endocarditis, 1984 through 1993: a clinical and microbiological survey.J Int Med1997;242,15-22
 
Foghsgaard, J, Pedersen, SA, Launbjerg, J Incidence and diagnosis of infectious endocarditis in Frederiksborg county, 1990–2000.Ugeskr Laeger2004;166,2446-2450. [PubMed]
 
Pedersen, S, Foghsgaard, J, Launbjerg, J Infectious endocarditis in Frederiksborg Amt, 1990–2000: clinical findings and prognostic aspects [in Danish].Ugeskr Laeger2004;166,2441-2446. [PubMed]
 
Delahaye, F, Goulet, V, Lacassin, F, et al Characteristics of infective endocarditis in prance in 1991: a 1-year survey.Eur Heart J1995;16,394-401. [PubMed]
 
Hoen, B, Alla, F, Selton-Suty, C, et al Changing profile of infective endocarditis: results of a 1-year survey in France.JAMA2002;288,75-81. [PubMed]
 
Rossi, L, Castello, C, Franceschini, L, et al Clinico-epidemiological aspects of infectious endocarditis in a present-day Italian population.G Ital Cardiol1986;16,30-37. [PubMed]
 
van der Meer, JT, Thompson, J, Valkenburg, HA, et al Epidemiology of bacterial endocarditis in the Netherlands: I. Patient characteristics.Arch Intern Med1992;152,1863-1868. [PubMed]
 
Hogevik, H, Olaison, L, Andersson, R, et al Epidemiologic aspects of infective endocarditis in an urban population: a 5-year prospective study.Medicine1995;74,324-339. [PubMed]
 
Skehan, JD, Murray, M, Mills, PG Infective endocarditis: incidence and mortality in the North East Thames Region.Br Heart J1988;59,62-68. [PubMed]
 
Whitby, M, Fenech, A Infective endocarditis in adults in Glasgow, 1976–81.Int J Cardiol1985;7,391-403. [PubMed]
 
Smith, RH, Radford, DJ, Clark, RA, et al Infective endocarditis: a survey of cases in the South-East region of Scotland, 1969–72.Thorax1976;31,373-379. [PubMed]
 
Griffin, MR, Wilson, WR, Edwards, WD, et al Infective endocarditis: Olmsted County, Minnesota, 1950 through 1981.JAMA1985;254,1199-1202. [PubMed]
 
Tleyjeh, IM, Steckelberg, JM, Murad, HS, et al Temporal trends in infective endocarditis: a population-based study in Olmsted County, Minnesota.JAMA2005;293,3022-3028. [PubMed]
 
Goulet, V, Etienne, J, Fleurette, J, et al Infectious endocarditis in France: epidemiological characteristics.Presse Med1986;15,1855-1858. [PubMed]
 
Oyonarte, M, Montagna, R, Braun, S, et al Infectious endocarditis: morbi-mortality in Chile: results of the National Cooperative Study of Infective Endocarditis (1998–2002).Rev Med Chil2003;131,237-250. [PubMed]
 
Bayliss, R, Clarke, C, Oakley, CM, et al Incidence, mortality and prevention of infective endocarditis.J R Coll Physicians Lond1986;20,15-20. [PubMed]
 
Berlin, JA, Abrutyn, E, Strom, BL, et al Incidence of infective endocarditis in the Delaware Valley, 1988–1990.Am J Cardiol1995;76,933-936. [PubMed]
 
Cecchi, E, Forno, D, Imazio, M, et al New trends in the epidemiological and clinical features of infective endocarditis: results of a multicenter prospective study.Ital Heart J2004;5,249-256. [PubMed]
 
King, JW, Nguyen, VQ, Conrad, SA Results of a prospective statewide reporting system for infective endocarditis.Am J Med Sci1988;295,517-527. [PubMed]
 
Cheng, A, Athan, E, Appelbe, A, et al The changing profile of bacterial endocarditis as seen at an Australian provincial centre.Heart, Lung and Circulation2002;11,26-31
 
Hricak, V, Kovacik, J, Marx, P, et al Etiology and risk factors of 180 cases of native valve endocarditis: report from a 5-year national prospective survey in Slovak Republic.Diagn Microbiol Infect Dis1998;31,431-435. [PubMed]
 
Krcmery, V, Gogova, M, Ondrusova, A, et al Etiology and risk factors of 339 cases of infective endocarditis: report from a 10-year national prospective survey in the Slovak Republic.J Chemother2003;15,579-583. [PubMed]
 
Schnurr, LP, Ball, AP, Geddes, AM, et al Bacterial endocarditis in England in the 1970’s: a review of 70 patients.Q J Med1977;46,499-512. [PubMed]
 
Fonager, K, Lindberg, J, Thulstrup, AM, et al Incidence and short-term prognosis of infective endocarditis in Denmark, 1980–1997.Scand J Infect Dis2003;35,27-30. [PubMed]
 
Von Reyn, CF, Levy, BS, Arbeit, RD, et al Infective endocarditis: an analysis based on strict case definitions.Ann Intern Med1981;94,505-518. [PubMed]
 
Durack, DT, Lukes, AS, Bright, DK New criteria for diagnosis of infective endocarditis: utilization of specific echocardiographic findings.Am J Med1994;96,200-209. [PubMed]
 
Li, JS, Sexton, DJ, Mick, N, et al Proposed modifications to the Duke criteria for the diagnosis of infective endocarditis.Clin Infect Dis2000;30,633-638. [PubMed]
 

Figures

Figure Jump LinkFigure 2. Top, A: Distribution of proportions of S aureus IE among culture-positive IE cases across different countries: X-axis (year), Y-axis (proportion of viridans streptococcal IE among culture-positive IE cases). Each country is represented by a different shape and color. Bottom, B: Metaregression analyses: overall temporal trends in S aureus IE among culture-positive IE cases: X-axis (year), Y-axis (proportion of S aureus IE among culture-positive IE cases). Only studies from countries with multiple time points, and therefore contributing to the estimated slope in the adjusted analysis, are shown. Bubble sizes in these plots are proportional to weights used in the analysis, and the regression lines presented have an intercept that is calculated as the average across individual countries.Grahic Jump Location
Figure Jump LinkFigure 3. Top, A: Distribution of proportions of viridans streptococcal IE among culture-positive IE cases across different countries: X-axis (year), Y-axis (proportion of viridans streptococcal IE among culture-positive IE cases). Each country is represented by a different shape. Bottom, B: Metaregression analyses: overall temporal trends in viridans streptococcal IE among culture-positive IE cases: X-axis (year), Y-axis (proportion of viridans streptococcal IE among culture-positive IE cases). Only studies from countries with multiple time points, and therefore contributing to the estimated slope in the adjusted analysis, are shown. Bubble sizes in these plots are proportional to weights used in the analysis and the regression lines presented have an intercept which is calculated as the average across individual countries.Grahic Jump Location
Figure Jump LinkFigure 1. Flow diagram of eligible studies.Grahic Jump Location

Tables

Table Graphic Jump Location
Table 1. Study and Population Characteristics of Eligible Studies
Table Graphic Jump Location
Table 1A. Continued

* N/A = not applicable; D = definite; D/P = definite/possible; D/P/P = definite/probable/possible; D/P/P/R = definite/probable/possible/rejected; CTS = cardiothoracic surgeons; ID = infectious diseases physicians.

 

Durack et al.31

 

Li et al.32

Table Graphic Jump Location
Table 2. Demographics and Clinical Characteristics of Eligible Studies*
* 

IVDA = IV drug abuse.

 

Data on 74 of 80 subjects.

 

1.7 in the first 5 years and 3.6 the next 5 years.

§ 

Average incidence.

 

Male/female gender, No.

Table Graphic Jump Location
Table 3. Individual Studies That Examined Temporal Trends*
* 

UHD = underlying heart disease.

Table Graphic Jump Location
Table 4. Results of Metaregression Analyses for Temporal Trends of IE Characteristics*
* 

Data are presented as change in proportion (95% CI). NA = data not available.

References

Mylonakis, E, Calderwood, SB (2001) Infective endocarditis in adults.N Engl J Med345,1318-1330. [PubMed] [CrossRef]
 
Moreillon, P, Que, YA Infective endocarditis.Lancet2004;363,139-149. [PubMed]
 
Steckelberg, JM, Melton, LJ, III, Ilstrup, DM, et al Influence of referral bias on the apparent clinical spectrum of infective endocarditis.Am J Med1990;88,582-588. [PubMed]
 
Nolsoe, C, Jensen, LT, Hartzner, K, et al Occurrence of infectious endocarditis in Denmark.Ugeskr Laeger1987;149,3458-3459. [PubMed]
 
Nissen, H, Nielsen, PF, Frederiksen, M, et al Native valve infective endocarditis in the general population: a 10-year survey of the clinical picture during the 1980s.Eur Heart J1992;13,872-877. [PubMed]
 
Benn, M, Hagelskjer, LH, Tvede, M Infective endocarditis, 1984 through 1993: a clinical and microbiological survey.J Int Med1997;242,15-22
 
Foghsgaard, J, Pedersen, SA, Launbjerg, J Incidence and diagnosis of infectious endocarditis in Frederiksborg county, 1990–2000.Ugeskr Laeger2004;166,2446-2450. [PubMed]
 
Pedersen, S, Foghsgaard, J, Launbjerg, J Infectious endocarditis in Frederiksborg Amt, 1990–2000: clinical findings and prognostic aspects [in Danish].Ugeskr Laeger2004;166,2441-2446. [PubMed]
 
Delahaye, F, Goulet, V, Lacassin, F, et al Characteristics of infective endocarditis in prance in 1991: a 1-year survey.Eur Heart J1995;16,394-401. [PubMed]
 
Hoen, B, Alla, F, Selton-Suty, C, et al Changing profile of infective endocarditis: results of a 1-year survey in France.JAMA2002;288,75-81. [PubMed]
 
Rossi, L, Castello, C, Franceschini, L, et al Clinico-epidemiological aspects of infectious endocarditis in a present-day Italian population.G Ital Cardiol1986;16,30-37. [PubMed]
 
van der Meer, JT, Thompson, J, Valkenburg, HA, et al Epidemiology of bacterial endocarditis in the Netherlands: I. Patient characteristics.Arch Intern Med1992;152,1863-1868. [PubMed]
 
Hogevik, H, Olaison, L, Andersson, R, et al Epidemiologic aspects of infective endocarditis in an urban population: a 5-year prospective study.Medicine1995;74,324-339. [PubMed]
 
Skehan, JD, Murray, M, Mills, PG Infective endocarditis: incidence and mortality in the North East Thames Region.Br Heart J1988;59,62-68. [PubMed]
 
Whitby, M, Fenech, A Infective endocarditis in adults in Glasgow, 1976–81.Int J Cardiol1985;7,391-403. [PubMed]
 
Smith, RH, Radford, DJ, Clark, RA, et al Infective endocarditis: a survey of cases in the South-East region of Scotland, 1969–72.Thorax1976;31,373-379. [PubMed]
 
Griffin, MR, Wilson, WR, Edwards, WD, et al Infective endocarditis: Olmsted County, Minnesota, 1950 through 1981.JAMA1985;254,1199-1202. [PubMed]
 
Tleyjeh, IM, Steckelberg, JM, Murad, HS, et al Temporal trends in infective endocarditis: a population-based study in Olmsted County, Minnesota.JAMA2005;293,3022-3028. [PubMed]
 
Goulet, V, Etienne, J, Fleurette, J, et al Infectious endocarditis in France: epidemiological characteristics.Presse Med1986;15,1855-1858. [PubMed]
 
Oyonarte, M, Montagna, R, Braun, S, et al Infectious endocarditis: morbi-mortality in Chile: results of the National Cooperative Study of Infective Endocarditis (1998–2002).Rev Med Chil2003;131,237-250. [PubMed]
 
Bayliss, R, Clarke, C, Oakley, CM, et al Incidence, mortality and prevention of infective endocarditis.J R Coll Physicians Lond1986;20,15-20. [PubMed]
 
Berlin, JA, Abrutyn, E, Strom, BL, et al Incidence of infective endocarditis in the Delaware Valley, 1988–1990.Am J Cardiol1995;76,933-936. [PubMed]
 
Cecchi, E, Forno, D, Imazio, M, et al New trends in the epidemiological and clinical features of infective endocarditis: results of a multicenter prospective study.Ital Heart J2004;5,249-256. [PubMed]
 
King, JW, Nguyen, VQ, Conrad, SA Results of a prospective statewide reporting system for infective endocarditis.Am J Med Sci1988;295,517-527. [PubMed]
 
Cheng, A, Athan, E, Appelbe, A, et al The changing profile of bacterial endocarditis as seen at an Australian provincial centre.Heart, Lung and Circulation2002;11,26-31
 
Hricak, V, Kovacik, J, Marx, P, et al Etiology and risk factors of 180 cases of native valve endocarditis: report from a 5-year national prospective survey in Slovak Republic.Diagn Microbiol Infect Dis1998;31,431-435. [PubMed]
 
Krcmery, V, Gogova, M, Ondrusova, A, et al Etiology and risk factors of 339 cases of infective endocarditis: report from a 10-year national prospective survey in the Slovak Republic.J Chemother2003;15,579-583. [PubMed]
 
Schnurr, LP, Ball, AP, Geddes, AM, et al Bacterial endocarditis in England in the 1970’s: a review of 70 patients.Q J Med1977;46,499-512. [PubMed]
 
Fonager, K, Lindberg, J, Thulstrup, AM, et al Incidence and short-term prognosis of infective endocarditis in Denmark, 1980–1997.Scand J Infect Dis2003;35,27-30. [PubMed]
 
Von Reyn, CF, Levy, BS, Arbeit, RD, et al Infective endocarditis: an analysis based on strict case definitions.Ann Intern Med1981;94,505-518. [PubMed]
 
Durack, DT, Lukes, AS, Bright, DK New criteria for diagnosis of infective endocarditis: utilization of specific echocardiographic findings.Am J Med1994;96,200-209. [PubMed]
 
Li, JS, Sexton, DJ, Mick, N, et al Proposed modifications to the Duke criteria for the diagnosis of infective endocarditis.Clin Infect Dis2000;30,633-638. [PubMed]
 
NOTE:
Citing articles are presented as examples only. In non-demo SCM6 implementation, integration with CrossRef’s "Cited By" API will populate this tab (http://www.crossref.org/citedby.html).

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging & repositioning the boxes below.

Find Similar Articles
CHEST Journal Articles
PubMed Articles
Guidelines
  • CHEST Journal
    Print ISSN: 0012-3692
    Online ISSN: 1931-3543