Methods

From 1996 through 2009, 292 consecutive patients with definite NVE according to modified Duke criteria were identified at Ca’ Foncello Hospital, Treviso, Italy. Clinical, microbiological, and imaging data were collected prospectively in a computerized database. Comorbidity was assessed by using the Charlson co-morbidity index. Each patient underwent ≥1 transthoracic and ≥transesophageal echocardiography (TEE) study. Three or more blood cultures were obtained from each patient. Delayed diagnosis was considered when the time interval from onset of symptoms to diagnosis was >30 days; failed antibiotic therapy was identified when bacteremia and/or fever with leukocytosis persisted after 1 week of intravenous antibiotic therapy; renal failure was considered when serum creatinine at admission was >1.4 mg/dl (mean reference value of the hospital laboratory ±2 SD); and worsened renal function was considered when normal baseline creatinine exceeded 1.4 mg/dl or when an increase ≥25% in abnormal baseline creatinine was observed during hospitalization. Recurrence was identified with either a repeat episode of NVE caused by the same microorganism as a previous episode or an infection with a different microorganism.

The study was divided into two 7-year consecutive periods: period 1 before implementation of the standardized protocol (1996 through 2002) and period 2 after the implementation of our protocol (2003 through 2009). The investigation protocol conformed to the ethical guidelines of the 1975 Declaration of Helsinki and was approved by the ethical committee of the hospital. Ca’ Foncello Hospital, Treviso, is a tertiary care hospital in northeast Italy serving a population of approximately 650,000 people. A formal policy for the care of IE was introduced in 2003 in which patients were referred to and managed by a preexisting team involving a cardiologist, a specialist in infectious diseases, a microbiologist, and a cardiac surgeon as soon as the diagnosis of IE was considered. The diagnosis of NVE was established when 2 main Duke criteria or 1 main and 2 minor criteria were fulfilled. Particular attention was given to the standardization of the blood cultures and echocardiography because they are the primary diagnostic criteria and may be misleading when performed or interpreted incorrectly. In patients who were already taking antibiotics, blood samples were drawn 48 hours after antibiotics had been temporarily withdrawn or at the valley of the plasmatic antibiotic concentration. Serology for Coxiella , Bartonella , and Chlamydia was performed in patients with suspected IE in whom blood cultures were still negative after 48-hour incubation.

Echocardiography in patients with suspected NVE was initially performed transthoracically by expert physicians who were aware of the clinical suspicion using systems provided with the harmonic imaging facility. TEE was performed soon after the transthoracic study if the quality of the study was poor or if complications were suspected. In the remaining conditions, patients underwent TEE within 48 hours of admission. In the echocardiographic report, vegetation characteristics (related to embolic risk; mobile left-heart vegetations >10 mm were considered at risk), the presence of abscess or paravalvular complication (such as pseudo-aneurysm and fistula), and the left ventricular function should have been reported.

After NVE was diagnosed and a complete echocardiogram was available, the initial multidisciplinary evaluation was performed (within 12 hours of admission/diagnosis) by a cardiologist, a cardiac surgeon, and a specialist in infectious diseases. This initial collaborative evaluation was intended to assess the immediate risk for death or serious sequelae that could be alleviated by immediate surgery. When patients were defined as low risk or deemed unfit for surgery (patients with severe co-morbidities, active malignancy, age >90 years, logistic EuroSCORE >30, major neurologic deficit or coma due to ischemic stroke, hemorrhagic stroke), medical therapy was indicated. Patients were defined as high risk when they presented with conditions associated with impending hemodynamic impairment (i.e., heart failure, severe multivalvular regurgitation without heart failure, aorto-cavitary fistula, severe aortic or mitral regurgitation with left ventricular ejection fraction <0.45) and in cases in which antibiotic therapy had little chance of curing the infection or had proved ineffective (i.e., fungal NVE, persistent bacteremia or fever with leukocytosis after 1 week of adequate antibiotic therapy, perivalvular abscess, or false aneurysm). In all these cases patients underwent urgent surgery (i.e., within 48 hours).

Another indication for early surgery was the prevention of systemic embolic events. Because the rate of emboli declines significantly in the first days after initiation of the antibiotic therapy, patients who needed immediate surgery were those who (1) presented with left-side high-risk (i.e., mobile, >10 mm) vegetations at first evaluation before antibiotic therapy had been initiated and (2) had a major systemic embolic event while assuming antibiotics if a high-risk vegetation persisted at TEE. Patients assigned to medical therapy were admitted to the department of infectious diseases. The patient’s clinical course was evaluated daily. C-reactive protein; blood cultures; and plasmatic concentrations of vancomycin, teicoplanin, and gentamicin were assessed every 3 days. Patients underwent transthoracic echocardiography every 5 days (TEE when the acoustic window was inadequate or when there was a discrepancy between clinical data and the findings of transthoracic echocardiography), and they were evaluated by the multidisciplinary team every week. If the patients developed conditions at high risk or had persistent bacteremia/fever after 1 week of appropriate antibiotic therapy (after excluding noncardiac causes of fever), on-treatment surgery was indicated. The choice of the antibiotic drug and the duration of treatment were decided by the infectious disease specialists according to published guidelines. All patients underwent a clinical, echocardiographic, and laboratory (including blood cultures) evaluation at 1, 3, 6, 12, 24, and 36 months after discharge.

The comparative distribution of clinical, microbiological, and echocardiographic variables and of clinical events was assessed in the 2 subgroups in a bivariate fashion. The impact of the management protocol on in-hospital mortality was assessed using the Cox proportional hazard model. Because data are observational and subjects were not randomized to 1 of the 2 groups, a propensity 2-treatment score analysis was used to control for underlying bias. The fit of the propensity model to the data was assessed using the concordance index. The matching tolerance was a propensity score difference of 0.20. To adjust further for potential confounding, a Cox proportional hazard regression analysis was used to assess the 3-year mortality between the matched groups. Psmatch2 module (STATA11, StataCorp, College Station, Texas) was used for this purpose. All tests were 2-sided and differences were considered statistically significant at p <0.05. Statistical analysis was carried out using also SPSS software V 12.0 (SPSS Inc., Chicago, Illinois).

Results

During the whole study period, 307 patients had NVE. After review of each case by 2 expert investigators, 15 cases were excluded for the following reasons: possible or rejected NVE (10) and incomplete case report form (5).The final sample was made up of 292 patients with definite NVE. One hundred two cases of NVE were included during period 1 (1996 through 2002), and 190 cases were included in period 2 (2003 through 2008), following the implementation of the management protocol. Comparing period 1 with period 2 ( Table 1 ), the number of patients increased, with more patients referred from community hospitals; mean age increased, with more patients aged >70 years; and more patients had Charlson co-morbidity index >2. Oral Streptococci were less frequently identified as the causative agent; there was a significant increase in NVE caused by group D Streptococci . Colonoscopy was performed in 55 of 61 patients with NVE caused by these germs and identified advanced colorectal neoplastic lesions in 44 (80%) of them. Culture-negative NVE declined significantly from period 1 to period 2. The pattern of infection did not alter in terms of infected valve and paravalvular involvement. In period 2, patients underwent more echocardiographic studies during hospitalization and less frequently experienced worsened renal function ( Table 2 ). The surgery rate during the index hospitalization was higher during period 2 (43% vs 31%), but this difference did not attain statistical significance. Indications for early surgery were more common in period 2 with more patients undergoing urgent surgery. Consequently, elective surgery after the completion of planned antibiotic therapy declined in period 2 from 28% to 9% (p <0.0001). Although the type of surgery did not change, the mortality rate in patients undergoing operation during the active phase of the disease significantly decreased during period 2 from 47% to 13% (p <0.001). Overall in-hospital mortality declined from 28% to 13% (p = 0.02). No difference in mortality was observed between patients with infection involving the mitral valve or the aortic valve (17 vs 21%, p = 0.53). Causes of death during the study are presented in Table 3 ; deaths from septic shock significantly decreased in period 2. In multivariate analysis ( Table 4 ), failed antibiotic therapy, onset of multiorgan failure, and the presence of abscess at echocardiography were independent predictors of 3-year mortality. After adjustment for relevant prognostic factors (gender, heart failure, delayed diagnosis, diabetes mellitus, failed antibiotic therapy, oral Streptococci , Staphylococcus aureus ) and patient characteristics that had changed significantly between the 2 periods (age, Charlson co-morbidity index, causative organism such as group D Streptococci and oral Streptococci , culture-negative NVE, number of patients referred from other hospitals), the treatment during period 2 remained independently predictive of 3-year survival (odds ratio [OR]: 0.38, 95% CI: 0.11 to 0.76, p = 0.03). The beneficial effect remained when the calendar year was added to the Cox model (OR: 0.41, 95% CI: 0.19 to 0.88, p = 0.03). Finally, using matching, we obtained 2 cohorts of 40 patients each ( Table 5 ). Within these 80 matched patients, multiorgan failure syndrome and the presence of abscess were independent predictors of 3-year mortality, and the management of NVE during period 2 was independently correlated with 3-year survival (OR: 0.71; 95% CI: 0.08 to 0.91; Table 6 ).

Table 4

Multivariate analysis of 3-year mortality (Cox multivariate analysis)

	Adjusted OR	95% CI	p Value
Age	1.3	0.5–2.1	0.47
Age >70 yrs	1.9	0.6–5.2	0.67
Male gender	0.7	0.2–5.2	0.54
Diagnosis >30 days	2.1	0.5–8.8	0.31
Predisposing cardiac disease	4.1	0.76–48	0.97
Diabetes mellitus	3.4	0.9–18	0.07
Renal failure	5.1	0.5–9.9	0.09
Immunosuppression	15.1	0.9–18	0.07
Charlson co-morbidity index	3.4	0.5–18	0.19
Charlson co-morbidity index ≥ 2	4.7	0.6–25	0.08
Staphylococcus aureus	2.8	0.8–10.1	0.11
Group D Streptococci	1.5	0.6–5.3	0.71
Heart failure	3.2	0.4–20.5	0.22
Failed antibiotic therapy	16.1	5.1–48	<0.001
Multiorgan failure	29	8.1–114	<0.001
High-risk vegetations	1.7	0.8–2.5	0.07
Abscess	7.2	2.8–21.5	<0.001
Left ventricular ejection fraction <0.45	3.5	0.9–7.1	0.11
Surgery	0.71	0.33–4.5	0.10
Calendar year	0.78	0.6–1.3	0.38
Period 2 (2003–2009)	0.41	0.19–0.88	0.03

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related

Related posts:

1. Comparison of Usefulness of Secondary Prevention of Coronary Disease in Patients <80 Versus ≥80 Years of Age
2. Effect of Positive Well-Being on Incidence of Symptomatic Coronary Artery Disease
3. Evaluation of the Appropriateness and Outcome of In-Hospital Telemetry Monitoring
4. Effectiveness of Percutaneous Closure of Patent Foramen Ovale for Hypoxemia

Stay updated, free articles. Join our Telegram channel

Join