The incidence of infective endocarditis (IE) in the general population is estimated to be 0.6 to 11.6 cases per 100,000 person-years.1-7 A number of risk factors predispose patients to the development of IE, including injection drug use (IDU),3 female sex,8 prosthetic heart valves,9 structural heart disease,3,10 older age,11 and prior history of infective endocarditis.12 Previous studies have demonstrated higher rates of IE among HIV-infected IDUs than among HIV-negative IDUs5,8,13-15 and an inverse relationship between IE occurrence and CD4 lymphocyte count.8,15-17
The introduction of highly active antiretroviral therapy (HAART) has reduced morbidity, mortality, and hospitalizations associated with HIV infection.18-20 In addition, the epidemiology of HIV has changed, with more women, minorities, and IDUs becoming infected in the past decade.21,22
Studies in the pre-HAART and early HAART eras indicated an increased risk of IE in HIV-infected patients8 and higher mortality rates from IE among severely immunosuppressed HIV-infected patients.15-17,23,24 However, no studies have evaluated the long-term morbidity and mortality of IE in HIV-infected patients in the current era of HAART. Therefore, the following study aims to examine the changing incidence of IE between 1990 and 2002 and to examine risk factors for IE in the era of HAART in an urban HIV cohort. In addition, we sought to describe the clinical presentation and 1-year morbidity and mortality of IE in HIV-infected patients followed at Johns Hopkins Hospital with IE between 1996 and 2002.
Patient and Data Collection
The Johns Hopkins University AIDS Service provides comprehensive primary and subspecialty care for HIV-infected patients. At the time of registration into the clinic, patients undergo an evaluation by a physician or physician's assistant and a social worker. The clinic-based medical record maintains information on all confirmed medical and surgical diagnoses, hospitalizations, laboratory information, and a section for each visit to document prescribed therapy by treatment name, dose, and number of refills. The records are also updated when prescriptions are filled over the phone or mailed to patients.
An observational database has been maintained on clinic patients to obtain extensive information about patients followed in our clinic. Trained monitors use structured data collection forms to extract extensive demographic, clinical, laboratory, pharmaceutical, and psychosocial data as well as death information from patient charts and from the hospital's automated databases at baseline and every 6 months thereafter. Information on death was obtained from patients' charts and from a separate death registry maintained by the clinic. The names of those patients whose vital status was unknown for more than 12 months were searched for in death records of the Maryland Bureau of Vital Records and the National Death Index. Systematic problems were identified and corrected for all patients. Maintenance of our database and use of its contents for analysis of patient outcomes is approved by the Institutional Review Board of the Johns Hopkins University School of Medicine. Patients receiving longitudinal primary HIV care who were enrolled in our clinic before January 1, 2003, were eligible for inclusion in our analysis.
For our analysis, HIV transmission risk factors included injection drug use (IDU), men who had sex with men (MSM), and heterosexual transmission, which was defined as heterosexual activity either with a partner at high risk for HIV or with an HIV-infected individual. Risk factor assignment was not mutually exclusive, as patients could have multiple HIV risk factors. HAART was defined as (1) 3 or more nucleosides; (2) any use of 1 or more protease inhibitors (PIs) or a non-nucleoside reverse transcriptase inhibitor (NNRTI) in combination with 2 or more nucleoside RTIs (NRTIs); or (3) a PI + NNRTI combination. Patients were considered to be on HAART if they received any of these combinations.
Cases of infective endocarditis (IE) were defined according to the modified Duke Criteria.25 Only definite and possible cases of endocarditis were included. All charts were systematically reviewed by one of the authors (M.B.) to confirm IE diagnosis. Possible IE cases were reviewed by 2 board-certified infectious diseases physicians for confirmation (K.G., L.W.). Potential IE cases were identified between January 1, 1990, and December 31, 2002. Only first episodes of IE were used for the analysis. Potential case patients for which adequate information could not be obtained to confirm the IE diagnosis were excluded from the analysis.
To identify factors associated with IE in the current era of HAART, demographic, clinical, pharmacy, physical examination, and laboratory examination from the presentation were reviewed for all IE cases between January 1, 1996, and December 31, 2002. The CD4 and HIV-1 RNA levels at the time of IE occurrence were recorded. All charts were reviewed to assess for 1-year mortality and recurrence of IE. Recurrence of IE was defined as another episode of IE more than 30 days after completion of antibiotic treatment.
Statistical analyses were done with use of the STATA 8.0 (College Station, TX). A nested case-control analysis was used to assess risk factors associated with IE between 1996 and 2002. Four non-IE controls were randomly selected from the overall cohort for each case. Controls were matched on cohort enrollment date and duration of follow-up in the cohort. For this nested case-control study, conditional logistic regression analyses were used to estimate the bivariate and multivariate adjusted odds ratios associated with IE.
Logistic regression was used to identify demographic and clinical factors associated with 1-year mortality and 1-year recurrence of IE. Separate univariate regression analyses were performed to identify individual variables significantly associated with the outcome (P < 0.05). Possible interactions between pairs of significant variables were tested by combining variables in the logistic regression. There were no significant interactions between the demographic characteristics of race, age, or gender and HIV risk factor. All reported P values are 2-tailed.
Between 1990 and 1995, we identified 56 total cases of IE among 1113 study patients with 2734 years of follow-up, for an overall incidence of 20.5 per 1000 PY. Of these 56 cases, 45 were first episodes, for a first episode incidence of 16.5 per 1000 PY in the pre-HAART era. We identified 86 IE episodes in 4162 patients with 13,067 years of follow-up between January 1, 1996, and December 31, 2002, for an incidence of 6.6 per 1000 PY. Of those between 1996 and 2002, 58 were first episodes and 28 were recurrences for a first episode incidence of 4.4 per 1000 PY in the HAART era. Because prior endocarditis is a risk factor for subsequent IE, for the remainder of this analysis, we describe the 58 first episodes of IE. Of note, of the 58 cases, 42 (72.4%) met the modified Duke Criteria for definite IE cases and 16 (27.6%) for possible IE cases.
The majority of IE cases were African American (89.7%), male (65.5%), and had a median age of 40 years (range 28-57; Table 1). The major HIV risk factors in the patients were IDU (84.5%), heterosexual transmission (31.0%), and MSM (13.8%). The median CD4 at IE episode was 68 cells/mm3 (range 1-1224), with a median HIV-1 RNA of 78,288 copies/mL (range 0 to >750,000). Seventy-five percent were infected with hepatitis C virus. At the time of the IE episode, 31.0% of patients were on HAART, 87.9% were on Pneumocystis jiroveci pneumonia (PCP) prophylaxis, and 70.7% were on mycobacterium avium complex (MAC) prophylaxis. Other antibiotics at the time of IE diagnosis include ciprofloxacin (2%), clindamycin (2%), and metronidazole (2%). One patient was on corticosteroids at the time of IE diagnosis.
In the case-control analysis, cases with first episode of IE and controls did not differ by age, race, gender, or HAART usage (Table 2). Cases were more likely than controls to have a history of IDU (OR, 6.67; 95% CI: 3.04 to 14.6), to be more immunosuppressed with lower CD4 counts (P ≤ 0.001), and to have higher viral loads (P ≤ 0.001). Cases were also more likely to be chronically infected with hepatitis C virus (OR, 3.80; 95% CI: 1.87 to 7.74) than controls. Consistent with worsening immunosuppression, cases were more likely to be on PCP or MAC prophylaxis, than controls; however, after adjustment for need of opportunistic infection (OI) prophylaxis, there was no association between OI prophylaxis and development of IE.
In multivariate case-control analysis, cases were more likely to have a history of IDU (AOR, 8.71; 95% CI: 2.92 to 26.0), have CD4 counts <50 cells/mm3, and HIV-1 RNA >100,000 copies/mL (AOR, 3.88; 95% CI: 0.95 to 15.9) (Table 2).
In those with endocarditis, common presenting symptoms included fever (62.1%), chills (31.0%), shortness of breath (25.9%), myalgia (15.5%), and chest pain (13.8%). The most common etiologic organisms of IE were Staphylococcus aureus (69%; of these, 11 (27.5%) were methicillin resistant), Enterococcus faecalis (6.9%), coagulase-negative staphylococci (6.9%), and Streptococcus viridans (5.1%) (Table 3).
Echocardiograms were performed on 48 (82.8%) of IE case patients. The remaining 10 IE case patients' charts were reviewed by 2 board-certified infectious diseases physicians (K.G., L.W.). Of these 10 IE cases, 1 patient died shortly after admission (<14 h) before an echocardiogram could be performed and met criteria for “possible” IE. The remainder met modified Duke Criteria for “definite” (7) or “possible” (2) IE and were treated presumptively without echocardiogram. Ninety-six percent of all echocardiograms performed in this study were transthoracic, and 2 (4%) of the echocardiograms performed were transesophageal. Among those with echocardiograms, vegetations were noted on 5 (10.4%) tricuspid valves, 10 (20.8%) aortic valves, 11 (22.9%) mitral valves, and 0 pulmonic valves. New regurgitation was noted in 22 (45.8%) tricuspid valves, 7 (14.6%) aortic valves, 20 (41.7%) mitral valves, and 2 (4.2%) pulmonic valves. Including vegetations and regurgitations, the mitral 25 (52.1%) and tricuspid valves 23 (47.9%) were the most frequently affected, followed by the aortic 12 (25.0%) and pulmonic valves 2 (4.2%) (Table 4). In 18 (37.5%) cases, 2 or more valves were involved. Fifteen (31.3%) cases had evidence of both left- and right-sided involvement. All cases involved native valves.
Vascular phenomena were noted in 31% of all cases. Thirteen patients (22.4%) developed pulmonary infarcts, 5 (8.6%) had major arterial emboli (3 cerebral infarcts, 2 renal emboli), and 3 (5.2%) were noted to have Janeway lesions. Immunologic phenomena were also noted in 6.9% of all cases. Two patients (3.5%) developed glomerulonephritis, and 2 (3.5%) had Osler nodes. Other complications included septic arthritis in 2 patients (3.5%) and severe arrhythmias in 1 patient (1.6%). Seven patients (12.1%) were admitted to the intensive care unit, 5 (7.8%) required surgery (1 valve replacement, 1 bilateral leg amputation, and 3 abscess debridement). Six (10.3%) died before discharge.
Within 1 year of IE diagnosis, 9 patients (16%) had an IE recurrence and 30 (52%) died. Combined, 34 (58%) patients either died or had IE recurrence. Recurrent cases of IE presented with a variety of organisms; 2 with multiple organisms. Six cases were due to S. aureus, 2 to S. viridans, and 1 to Eikenella corrodens. Of those who recurred, 4 had the same organism and 6 were confirmed by repeat echocardiogram to involve the same valve as the original IE episode. The median time from presentation to death was 105 days (range 0-342 days), and the median time to IE recurrence was 105 days (range 80-365 days). Known death diagnoses were varied but included endocarditis complications (2), hypotensive shock (1), complications from surgical removal of subdural hematoma and abscess (1), end-stage renal disease (1), septic myocardial infarcts (1), sepsis (1), and acute respiratory distress syndrome (1).
In univariate analysis, age of 40 years or greater was associated with 1-year mortality (OR, 4.22; 95% CI: 1.41 to 12.66) but not recurrence (Table 4). All recurrences occurred in African American patients; however, race was not associated with 1-year mortality. History of IDU, HCV infection, PCP prophylaxis, MAC prophylaxis, and HAART usage were not significantly associated with mortality or recurrence. Organism involved, valve involved, and left-sided vs. right-sidedness were not associated with mortality or recurrence. In multivariate analysis, adjusting for race, sex, IDU, and HAART usage, age of 40 years or greater was associated with 1-year mortality (AOR, 8.67; 95% CI 4.73 to 43.4) (Table 5), but not with 1-year recurrence (OR, 1.30; 95% CI: 0.31 to 5.44).
This study has several important findings. First, in an urban HIV-infected cohort, IE is a common diagnosis and has significant morbidity and mortality. The incidence of 6.6 per 1000 PY is significantly greater than the 1 per 100,000 in the general population,3,6 and is consistent with other studies in IDUs8 and those with prosthetic valves at 10 per 1000 PY.26,27 In addition, the factors most strongly associated with IE episode were IDU and worsened immunosuppression, as measured by CD4 count <50 cells/mm3 and HIV-1 RNA >100,000 copies/mL at IE episode. Of those with IE, the majority of patients had a classic IE presentation of fever and chills, S. aureus bacteremia. Notably, there was a high rate of left-sided IE. The 1-year recurrence rate of 16% and mortality rate of 52% are higher than previously reported.
Incidence and Risk Factors for IE
Interestingly, the incidence of IE decreased significantly between the pre-HAART era and current era of HAART. The reasons for this decrease are unclear. Although HAART is not known to have a direct effect on B cells, other bacterial infections, such as bacteremias, urinary tract infections, and pneumonia have also been reported to decrease in the HAART era.28-30 In a recent study of this same cohort, the incidence of S. aureus bacteremia did not change significantly between 2000 and 2004; however, in this study, HAART usage was not associated with a decreased risk of bacteremia.31 Therefore, it is possible that, although HAART does not decrease overall bacteremia rates, it may decrease the propensity of the organism to cause endocarditis. Of note, there was no association to suggest an antimicrobial protective effect of PCP or MAC prophylaxis on development of IE.
In our study, patients with greater immunosuppression represented by lower CD4 counts were more likely to develop IE than those with higher CD4 counts, which is consistent with findings from the pre-HAART era.14 As in previous studies, there was no association of HAART usage with the development of IE.8 It should be noted, however, that many patients in this study who had CD4 counts <200 cells/mm3, which would be an indication for HAART therapy by current HIV guidelines, were not on HAART. This may be due to complex medical decision-making issues between the patient and the provider that we were unable to assess in this study. Regardless, lower CD4 count was associated with increased risk of IE.
Other studies have found an association of sex and development of IE, with some demonstrating increased rates in men,32 and others, in women.8,33 Like van der Meer,34,35 we did not find an association of female sex and IE. Of note, although our work was based on an urban HIV cohort, many of the women in our study had heterosexual risk and not IDU as an HIV risk factor. The van der Meer study34,35 was a nationwide prospective study of IE among those in the Netherlands, and only 7.3% of the population were IDUs.
There are several potential explanations for the high rate of left-sided involvement we observed in our cohort; however, our study suggests that left-sided endocarditis may be underdiagnosed in most studies of IDUs with HIV. Although IDU is a common risk factor for right-sided endocarditis, both clinical and autopsy investigations of IDUs with endocarditis have revealed similar rates of mitral and tricuspid valve involvement (46% vs. 32% and 44% vs. 43%, respectively).36,37 Furthermore, the relative frequencies of aortic and pulmonic involvement in our study were consistent with the findings from previous clinical and autopsy studies.36,37 Echocardiographic studies of injection drug users without a history of endocarditis have revealed evidence of predisposing valvular abnormalities (including focal thickening, valve prolapse and regurgitation) in both the tricuspid and mitral valves, suggesting a mechanism for left-sided infection.38 Also, S. aureus, which was highly prevalent in our cohort, may cause endocarditis even in the absence of an underlying lesion, as suggested by its high adherence ratio compared with nonpathogenic bacteria.39
Clinical Presentation and 1-Year Outcomes of IE
These HIV patients, like those in previous studies, had a febrile response to IE40 and presented with a higher ratio of left-sided endocarditis and S. aureus infections than HIV-negative IDUs.17,24,41 There was a lower than expected rate of S. viridans. This is most likely due to the large proportion of IDUs in the cohort. IDUs skin and nasal mucous membranes are known to have a high rate of colonization with S. aureus which may reach the bloodstream through percutaneous injection.42 The patients in this study had a relatively high rate of peripheral involvement with pulmonary emboli, cerebrovascular accidents, and renal emboli requiring intensive care admissions and intubations. Even with the large number of peripheral sequelae of endocarditis, only one patient underwent valve replacement, which has been shown to have low operative risk in HIV-infected patients and has not been shown to worsen the short-term prognosis of HIV.43,44
The rate of IE recurrence in HIV-infected patients has not been well-characterized. Our study demonstrates a high 1-year recurrence rate (16%). Of these, 44% recurred with the same organism. Although 7 patients left the hospital against medical advice before completion of antibiotic treatment, only 1 of those patients experienced a recurrence; therefore, this does not fully explain the high recurrence rate. An alternative explanation for the high recurrence rate is that many patients continued to use illicit drugs after their first IE event. Of those alive at 1 year, 20 (66%) were actively using drugs and 4 (13%) were in drug treatment.
Previous studies done in the pre-HAART and early HAART eras have demonstrated that overall IE mortality between HIV-infected and noninfected IDUs are similar.15,17,24 Another recent smaller study in Atlanta demonstrated a 30% incidence of death in HIV-infected patients with IE.45 Our study, however, shows an unexpectedly high 1-year mortality rate in this population that may be due to a number of causes, including continued IDU. Consistent with previous studies, our cohort showed high rates of left-sided involvement (53.5%) and S. aureus (69.0%), both of which are associated with increased mortality.15,17 Furthermore, there was a low rate of surgical intervention in our cohort and a high rate of embolization.
In our analysis, all recurrences were in African American patients. In addition, we also found that older patients were more likely to die than younger patients, although we found no other demographic factor associated with 1-year recurrence or mortality. Several studies have suggested that advanced stage and severe immunosuppression may be associated with worsened prognosis of endocarditis in HIV patients.15-17,24 Interestingly, we found no association of recurrence or mortality with immunosuppression, high viral load, or use of HAART. Our ability to detect a difference may have been limited, however, by our population being very immunosuppressed, with a median CD4 count of 68 cells/mm3.
Our study did have several potential limitations. First, these results are based on patients from a single institution. The Johns Hopkins Hospital is the primary hospital for these HIV patients, and we care for a high proportion of African Americans and IDUs. Therefore, our results may not be reflective of the all HIV clinics; however, our findings may be generalizable to other urban HIV care sites.46 Also, it is possible that we have overestimated the incidence of endocarditis, as only 73% of all cases met the modified Duke Criteria for “definite” IE; however, all other cases met criteria for “possible” IE, and the physicians responsible for these patients treated them clinically as if they had IE. Furthermore, 96% of all echocardiograms performed in this study were transthoracic and would therefore be likely to underestimate rather than overestimate IE case finding.47 In addition, in our analysis, we exclusively evaluated first episodes of endocarditis. Risk factors for a recurrent episode may in fact be different than risk factors for a first episode. Due to relatively small sample size, we were unable to assess for differences in type of PCP prophylaxis (atovaquone and/or dapsone compared to trimethoprim-sulfamethoxazole) and the association with risk of developing IE. In addition, our analysis may have missed IE episodes that occurred at other hospitals. A recent analysis of Medicaid claims of our patients suggests that 96% of all admissions occur at our hospital, suggesting excellent capture of hospitalizations. The authors confirmed that each event was a first episode by chart review. Finally, to evaluate the effect of HAART, we limited our study to only include the 58 episodes of IE that occurred in our clinic since January 1, 1996. The cases that occurred before introduction of HAART may, in fact, be different than those studied in our population.
These data from an urban, HIV-infected population demonstrate a significant decline in the incidence of IE between the pre-HAART and current HAART eras. In addition, IE occured at a rate near that of patients with prosthetic valves. IDUs and those with advanced HIV immunosuppression were more likely to develop IE than non-IDUs or those with elevated CD4 counts. These endocarditis patients presented classically with high rates of fever, valvular regurgitation, and S. aureus bacteremia. In addition, there were high rates of 1-year IE recurrence and 1-year mortality among HIV patients diagnosed with IE, indicating the need for more aggressive follow-up, especially in those over 40 years of age. Future studies investigating the utility of IE prophylaxis in HIV patients with a history of IE may be warranted.
1. Smith RH, Radford DJ, Clark RA, et al. Infective endocarditis: a survey of cases in the South-East region of Scotland, 1969-72. Thorax
2. Williams RC, Kunkel HG. Rheumatoid factors and their disappearance following therapy in patients with subacute bacterial endocarditis. Arthritis Rheum
3. Griffin MR, Wilson WR, Edwards WD, et al. Infective endocarditis. Olmsted County, Minnesota, 1950 through 1981. JAMA
4. Hickey AJ, MacMahon SW, Wilcken DE. Mitral valve prolapse and bacterial endocarditis: when is antibiotic prophylaxis necessary? Am Heart J
5. Berlin JA, Abrutyn E, Strom BL, et al. Incidence of infective endocarditis in the Delaware Valley, 1988-1990. Am J Cardiol
6. King JW, Nguyen VQ, Conrad SA. Results of a prospective statewide reporting system for infective endocarditis. Am J Med Sci
7. Haupt BJ, Graves E. Detailed Diagnoses and Procedures for Patients Discharged From Short Stay Hospitals: United States, 1979
. DHHD Publication No. PHS-82-1274-1. Hyattsville, MD: National Center for Health Statistics; 1982.
8. Wilson LE, Thomas DL, Astemborski J, et al. Prospective study of infective endocarditis among injection drug users. J Infect Dis
9. Bayer AS. Infective endocarditis. Clin Infect Dis
10. Cherubin CE, Neu HC. Infective endocarditis at the Presbyterian Hospital in New York City from 1938-1967. Am J Med
11. Cantrell M, Yoshikawa TT. Infective endocarditis in the aging patient. Gerontology
12. Tornos MP, Permanyer-Miralda G, Olona M, et al. Long-term complications of native valve infective endocarditis in non-addicts. A 15-year follow-up study. Ann Intern Med
13. Scheidegger C, Zimmerli W. Incidence and spectrum of severe medical complications among hospitalized HIV-seronegative and HIV-seropositive narcotic drug users. AIDS
14. Manoff SB, Vlahov D, Herskowitz A, et al. Human immunodeficiency virus infection and infective endocarditis among injecting drug users. Epidemiology
15. Nahass RG, Weinstein MP, Bartels J, et al. Infective endocarditis in intravenous drug users: a comparison of human immunodeficiency virus type 1-negative and -positive patients. J Infect Dis
16. Cicalini S, Forcina G, De Rosa FG. Infective endocarditis in patients with human immunodeficiency virus infection. J Infect
17. Ribera E, Miro JM, Cortes E, et al. Influence of human immunodeficiency virus 1 infection and degree of immunosuppression in the clinical characteristics and outcome of infective endocarditis in intravenous drug users. Arch Intern Med
18. Palella FJ Jr, Delaney KM, Moorman AC, et al. Declining morbidity and mortality among patients with advanced human immunodeficiency virus infection. HIV Outpatient Study Investigators. N Engl J Med
19. Mocroft A, Vella S, Benfield TL, et al. Changing patterns of mortality across Europe in patients infected with HIV-1. EuroSIDA Study Group. Lancet
20. Gebo KA, Fleishman JA, Moore RD. Hospitalizations for metabolic conditions, opportunistic infections, and injection drug use among HIV patients: trends between 1996 and 2000 in 12 states. J Acquir Immune Defic Syndr
21. Centers for Disease Control and Prevention. National HIV Surveillance Report
, Midyear Edition, 1999
. Atlanta, GA: CDC; 1999;11:3.
22. Karon JM, Fleming PL, Steketee RW, et al. HIV in the United States at the turn of the century: an epidemic in transition. Am J Public Health
23. Wallace SM, Walton BI, Kharbanda RK, et al. Mortality from infective endocarditis: clinical predictors of outcome. Heart
24. Pulvirenti JJ, Kerns E, Benson C, et al. Infective endocarditis in injection drug users: importance of human immunodeficiency virus serostatus and degree of immunosuppression. Clin Infect Dis
25. Li JS, Sexton DJ, Mick N, et al. Proposed modifications to the Duke criteria for the diagnosis of infective endocarditis. Clin Infect Dis
26. Ivert TS, Dismukes WE, Cobbs CG, et al. Prosthetic valve endocarditis. Circulation
27. Rutledge R, Kim BJ, Applebaum RE. Actuarial analysis of the risk of prosthetic valve endocarditis in 1,598 patients with mechanical and bioprosthetic valves. Arch Surg
28. Tumbarello M, Tacconelli E, Donati KG, et al. HIV-associated bacteremia: how it has changed in the highly active antiretroviral therapy (HAART) era. J Acquir Immune Defic Syndr
29. Gaetano Donati K, Tumbarello M, Tacconelli E, et al. Impact of highly active antiretroviral therapy (HAART) on the incidence of bacterial infections in HIV-infected subjects. J Chemother
30. Sullivan JH, Moore RD, Keruly JC, et al. Effect of antiretroviral therapy on the incidence of bacterial pneumonia in patients with advanced HIV infection. Am J Respir Crit Care Med
31. Burkey M, Lucas GM, Moore RD, et al. The incidence of and risk factors for MRSA bacteremia in an urban HIV cohort in the HAART era. Presented at: 13th Conference on Retroviruses and Opportunistic Infections; February 5-9, 2006; Denver, CO.
32. Watanakunakorn C. Changing epidemiology and newer aspects of infective endocarditis. Adv Intern Med
33. Spijkerman IJ, van Ameijden EJ, Mientjes GH, et al. Human immunodeficiency virus infection and other risk factors for skin abscesses and endocarditis among injection drug users. J Clin Epidemiol
34. van der Meer JT, Thompson J, Valkenburg HA, et al. Epidemiology of bacterial endocarditis in The Netherlands. I. Patient characteristics. Arch Intern Med
35. van der Meer JT, Thompson J, Valkenburg HA, et al. Epidemiology of bacterial endocarditis in The Netherlands. II. Antecedent procedures and use of prophylaxis. Arch Intern Med
36. Mathew J, Addai T, Anand A, et al. Clinical features, site of involvement, bacteriologic findings, and outcome of infective endocarditis in intravenous drug users. Arch Intern Med
37. Dressler FA, Roberts WC. Infective endocarditis in opiate addicts: analysis of 80 cases studied at necropsy. Am J Cardiol
38. Pons-Llado G, Carreras F, Borras X, et al. Findings on Doppler echocardiography in asymptomatic intravenous heroin users. Am J Cardiol
39. Gould K, Ramirez-Ronda CH, Holmes RK, et al. Adherence of bacteria to heart valves in vitro. J Clin Invest
40. Robinson DJ, Lazo MC, Davis T, et al. Infective endocarditis in intravenous drug users: does HIV status alter the presenting temperature and white blood cell count? J Emerg Med
41. Valencia ME, Guinea J, Soriano V, et al. Study of 164 episodes of infectious endocarditis in drug addicts: comparison of HIV positive and negative patients. Rev Clin Esp
42. Tuazon CU, Sheagren JN. Increased rate of carriage of Staphylococcus aureus
among narcotic addicts. J Infect Dis
43. Aris A, Pomar JL, Saura E. Cardiopulmonary bypass in HIV-positive patients. Ann Thorac Surg
44. Chong T, Alejo DE, Greene PS, et al. Cardiac valve replacement in human immunodeficiency virus-infected patients. Ann Thorac Surg
45. Smith DT, Sherwood M, Crisel R, et al. A comparison of HIV-positive patients with and without infective endocarditis: an echocardiographic study-The Emory Endocarditis Group experience. Am J Med Sci
46. Holmberg SD. The estimated prevalence and incidence of HIV in 96 large US metropolitan areas. Am J Public Health
47. Shively BK, Gurule FT, Roldan CA, et al. Diagnostic value of transesophageal compared with transthoracic echocardiography in infective endocarditis. J Am Coll Cardiol