Abstract
Infective endocarditis (IE) is uncommon in children with congenital heart disease (CHD), and limited data exist regarding in-hospital outcomes of those requiring cardiac surgery. We retrospectively reviewed hospitalizations from the pediatric health information system from January 2016 to November 2024. Among 357,031 unique pediatric patients with CHD (≤ 21 years), 2108 patients (0.5%) were hospitalized for IE. The index hospitalizations during which cardiac surgery was performed were included. We aimed to characterize in-hospital outcomes among CHD children undergoing cardiac surgery for IE, with sub-analysis of those who required surgery versus medical therapy. The incidence of IE in CHD was 5.6 cases per 1000. Microorganism coding was unavailable in 41.3%; Staphylococcus (24.1%) and Streptococcus (22.3%) were the most common. Stroke occurred in 10.7%. Prosthetic valve endocarditis was present in 9.4%. Overall, 30% of patients required cardiac surgery, most commonly pulmonary valve or conduit procedures (36.6%) and tricuspid valve surgery (16.7%). Two-valve surgery was performed in 13.3%. In-hospital mortality was comparable between surgical and medical patients (10.1% vs 9.6%, p = 0.7). Permanent pacemaker implantation was more common after surgery (6.6% vs 2.2%, p < 0.01). Length of hospital stay and hospitalization costs were higher in the surgical group. On multivariable analysis, older age, heart failure, stroke, and extracorporeal membrane oxygenation (ECMO) were associated with in-hospital mortality. IE in children with CHD remains associated with substantial morbidity and mortality. In-hospital mortality remains high despite cardiac surgery. Older age, heart failure, stroke, and ECMO requirement are independent factors of in-hospital mortality.
Supplementary Information
The online version contains supplementary material available at 10.1007/s00246-025-04037-7.
Keywords: Infective endocarditis, Congenital heart disease, Morbidity, Mortality
Background
Infective endocarditis (IE) is a rare condition in children with congenital heart disease (CHD), with substantial morbidity and mortality [1, 2]. Children with CHD are at higher risk for developing IE, with a risk of 15–140 times higher than the general population [3, 4]. The cumulative incidence of IE in CHD patients varies considerably in the literature, ranging from 4.1 to 11.1 per 10,000 person-years [5, 6]. IE accounts for 0.1 to 0.8 per 1000 admissions for pediatric patients in other reports [7, 8], with increasing incidence due to the improved survival in patients with CHD [9]. IE in patients with CHD usually results from the presence of foreign cardiac prosthetic material, implantable cardiac devices, residual shunts, endothelial shear stress, and exposure to healthcare procedures during hospitalizations [1].
Cardiac surgery is often necessary in patients admitted with IE [10–12]. Surgery for IE in children with CHD is challenging due to the complexity of their heart anatomy and hemodynamics, previous cardiac surgeries and procedures, the presence of prosthetic materials (patches, valves or conduits) and the small body size for a prosthetic valve if indicated [10]. In contrast to CHD-associated IE in adults, data regarding outcomes of surgical treatment for IE in pediatric patients with underlying CHD is lacking. We therefore utilized the Pediatric Health Information System (PHIS) database to assess clinical characteristics and in-hospital outcomes in these patients from a national database perspective.
Methods
Dataset
The Pediatric Health Information System (PHIS) database was utilized to obtain our target population (https://www.childrenshospitals.org/). The PHIS database is an administrative dataset that contains clinical and resource utilization data from approximately 50 tertiary care children’s hospitals in the USA. Participating hospitals, including our institution (Monroe Carell Jr. Children’s Hospital at Vanderbilt) are affiliated with the Children’s Hospital Association (CHA) (Lenexa, KS, USA). The PHIS data provides patient-level information such as demographics, admission data, diagnoses, procedures, length of hospital stay, and cost of hospitalization. These data are tested for quality control measures before inclusion.
Our study was exempted by the Vanderbilt University Institutional Review Board (IRB) in agreement with the common rule (45 CFR 46.104 (d) category (4) (ii)).
Patient Population
The PHIS database was queried from January 2016 to November 2024 for inpatient hospitalizations for pediatric patients ≤ 21 years of age with an International Classification of Disease, 10th Revision (ICD-10) code of infective endocarditis (I33.0, I33, I38, I33.9, I38X, I39.0, I39.1, I39.2, I39.3, I39.4, I39.8, B37.6, T82.6). Patients with diagnosis codes indicating the presence of CHD were included (Supplementary file, Table 1). PHIS database is encounter-based, so we limited our analysis to the index hospitalization for IE during which cardiac surgery was performed. This ensured that each individual patient was counted once for the purposes of surgical outcomes, avoiding duplications from readmissions.
Hospitalization for scheduled/elective cardiac surgery was only included in the analysis if preceded by an initial hospitalization for IE at the same hospital, thereby minimizing double-counting (n = 43 out of 632). Our study followed a similar approach to data extraction and cohort selection to ensure comparability with a similar prior study [13].
Patients with chronic medical conditions including oncologic, significant gastrointestinal, end-stage renal disease, and prematurity (< 32-week gestation) were excluded. These excluded conditions were identified based on ICD-10 codes (Supplementary Data).
Baseline demographics included year of index hospitalization, age, sex, ethnicity, and geographic region. Comorbidities were identified based on ICD-10 diagnosis codes. Microbiology data were also obtained. ICD-10 and Current Procedural Terminology (CPT) codes were utilized to identify cardiac surgeries performed during hospitalization.
Statistical Analysis
All statistical analyses were performed using IBM SPSS Statistics version 28 (IBM Corporation, Armonk, NY). Continuous variables were presented as medians with interquartile ranges (IQR) based on the data distribution, and comparisons were performed using the non-parametric Wilcoxon rank-sum test. Categorical parameters were reported as frequencies and percentages, and variables were compared using the Chi-square test or Fisher’s exact test (if the number of the variable is less than 5). A multivariable logistic regression model was created to identify factors significantly associated with in-hospital mortality. The Risk Adjustment for Congenital Heart Surgery (RACHS-2) category was included as a fixed covariate to adjust for procedural complexity, allowing estimation of associations between other clinical factors and in-hospital mortality independent of baseline surgical risk. Covariates chosen for inclusion in the regression model were selected according to clinical relevance and prior studies. No formal stepwise, forward, or backward selection algorithms were applied. These variables included age at index surgery, surgical complexity, staphylococcal infection, heart failure, prosthetic IE, stroke, type of CHD, multivalve surgery, requirement of ECMO, and pacemaker implantation. Odds ratios (OR) with 95% confidence intervals (CI) were reported. p < 0.05 was considered statistically significant.
Results
Baseline Characteristics
Patients’ characteristics and CHD diagnoses are summarized in Table 1.
Table 1.
Demographics and baseline characteristics in children with congenital heart disease undergoing cardiac surgery for infective endocarditis
| Non-surgical treatment (n = 1476) | Surgical group (n = 632) | p value | |
|---|---|---|---|
| Age in years (median [IQR] | 2.7 [0.04–11.8] | 9 [0.7–14.8] | < 0.01 |
| Age groups | |||
| 0–1 | 513 (34.8%) | 163 (25.8%) | < 0.01 |
| 1–3 | 133 (9%) | 35 (5.5%) | 0.007 |
| 3–6 | 114 (7.7%) | 29 (4.6%) | 0.009 |
| 6–9 | 89 (6%) | 63 (10%) | 0.001 |
| 9–21 | 419 (28.4%) | 285 (45.1%) | < 0.01 |
| Sex | 0.051 | ||
| Male | 825 (55.9%) | 383 (60.6%) | |
| Female | 651 (44.1%) | 249 (39.4%) | |
| Ethnicity | |||
| White | 900 (61%) | 397 (60%) | 0.66 |
| African American | 253 (17.1%) | 100 (15.8%) | 0.56 |
| Hispanic (missing, n = 76) | 161 (11.3%) | 52 (8.6%) | 0.07 |
| Other | 122 (8.6%) | 74 (12.3%) | 0.009 |
| Geographic region | < 0.01 | ||
| Midwest | 345 (23.4%) | 163 (25.8%) | |
| Northeast | 202 (13.7%) | 105 (16.6%) | |
| South | 616 (41.7%) | 218 (34.5%) | |
| West | 313 (21.2%) | 146 (23.1%) | |
| Diagnoses | |||
| Right-sided heart lesions | 454 (30.8%) | 201 (31.8%) | 0.63 |
| Left-sided heart lesions | 549 (37.2%) | 278 (44%) | 0.003 |
| Septal defects | 916 (62.1%) | 371 (58.7%) | 0.14 |
| Conotruncal malformation | 451 (30.6%) | 227 (35.9%) | 0.01 |
| Urgent or emergent admission status | 1173 (79.5%) | 439 (69.5%) | < 0.01 |
| Comorbidities | |||
| Heart failure | 342 (23.2%) | 263 (41.6%) | < 0.01 |
| Prosthetic valve endocarditis | 90 (6.1%) | 108 (17.1%) | < 0.01 |
| Embolic Stroke | 137 (9.3%) | 89 (13.6%) | 0.003 |
| Brain mycotic aneurysm | 3 (0.2%) | 6 (0.9%) | 0.02 |
| Pulmonary embolism | 116 (7.9%) | 102 (16.1%) | < 0.01 |
| Splenic infract | 14 (0.9%) | 11 (1.7%) | 0.12 |
| Renal infract | 25 (1.7%) | 12 (1.9%) | 0.74 |
| Musculoskeletal infections | 62 (4.2%) | 29 (4.6%) | 0.69 |
| IV drug user | 2 (0.1%) | 1 (0.2%) | 1.0 |
IQR interquartile range
A total of 357031 individual pediatric patients (≤ 21 years of age) with a diagnosis of CHD (Supplementary file, Table 2) were hospitalized during the study timeframe. Of them, 2108 patients (0.5%) were hospitalized for infective endocarditis. The included patients were distributed evenly across three distinct age groups: 32.1% in infants (0–12 months), 34.5% were aged between 1 and 12 years, and 33.4% were adolescents. Male sex accounted for 57% of the entire cohort. The incidence of IE in patients with CHD has increased from 0.6% in 2016 to 0.8% in 2024. The estimated risk of IE in patients with CHD is 5.6 per 1000 cases.
A total of 632 patients (30%) underwent cardiac surgery during the index hospitalization for IE. Patients who underwent surgical treatment were older (9 years [IQR (interquartile range), 0.7–14.8] vs 2.7 years [IQR, 0.04–11.8], p < 0.01). The presence of heart failure, stroke, brain mycotic aneurysm, and pulmonary embolism was more common in the surgical group (p < 0.05). Prosthetic valve infective endocarditis (PIE) was more frequently observed in the surgical group (108 patients (17.1%) vs 90 patients (6.1%), p < 0.01). Among those who required surgical intervention, the existing prosthetic valve location was pulmonary valve or conduit in 64 patients, aortic valve in 14 patients, mitral valve in 14 patients, tricuspid valve in 4 patients, and other non-valvar locations in 8 patients.
Microbiology Data
Microbiology information regarding the causative organism is depicted in Table 2. A code for any microorganism was not reported in 41.3% of patients. Staphylococcus was the most reported bacteria in 507 patients (24.1%) followed by streptococcus species in 471 patients (22.3%). Distribution was similar in the two groups.
Table 2.
Microbiology data
| Medical treatment (n = 1476) | Surgical group (n = 632) | p value | |
|---|---|---|---|
| Staphylococcus | 351 (23.8%) | 156 (24.7%) | 0.66 |
| Streptococcus | 328 (22.2%) | 143 (22.6%) | 0.84 |
| H. influenza | 14 (0.9%) | 10 (1.6%) | 0.21 |
| E. coli | 57 (3.9%) | 15 (2.4%) | 0.09 |
| Klebsiella | 54 (3.7%) | 22 (3.5%) | 0.84 |
| Pseudomonas | 42 (2.8%) | 15 (2.4%) | 0.54 |
| Candida | 125 (8.5%) | 47 (7.4%) | 0.43 |
Performed Surgeries
Previous cardiac surgery was present in 371 patients (58.7%). Detailed surgical interventions are reported in Table 3. Single valve surgery was required in the majority of patients (517 patients, 81.8%). Isolated pulmonary valve surgery (repair or replacement) or right ventricle to pulmonary trunk conduit replacement was the most common surgery performed (231 patients, 36.6%). Isolated tricuspid valve surgery was the second most common intervention in 109 patients (16.7%). Type of valve intervention, whether it is repair or replacement, was provided in the Supplementary file Table 3.
Table 3.
Surgical data of children with IE undergoing cardiac surgery
| Surgery performed | N = 632 |
|---|---|
| Single valve surgery | 517 (81.8%) |
| Isolated mitral valve surgery (repair or replacement) | 99 (15.7%) |
| Isolated aortic valve surgery | 73 (11.6%) |
| Isolated pulmonary valve or conduit surgery | 231 (36.6%) |
| Isolated tricuspid valve surgery | 109 (16.7%) |
| Multi-valve surgery | 84 (13.3%) |
| Concomitant aortic and mitral valves surgery | 7 (1.1%) |
| Concomitant mitral and tricuspid valves surgery | 26 (4.1%) |
| Concomitant tricuspid and pulmonary valve or conduit surgery | 25 (4.5%) |
| Unspecified valve surgery | 62 (9.8%) |
| Other non-valvar cardiac surgery | 31 (4.9%) |
| Redo heart surgery | 371 (58.7%) |
Classifications of cardiac lesions with corresponding surgical interventions are depicted in Table 4. Mitral and aortic valve procedures were performed in approximately 50% of patients with underlying left-sided lesions; interestingly, 37.8% of patients required pulmonary valve/conduit interventions. Pulmonary valve/conduit interventions were the predominant operation in right-sided and conotruncal heart lesions. Tricuspid valve procedures were more common in septal lesions.
Table 4.
Classifications of cardiac lesions and corresponding surgical interventions
| Left-sided lesions | |
| Mitral valve procedures | 69 (24.8%) |
| Aortic valve procedures | 62 (22.3%) |
| Tricuspid valve procedures | 69 (24.8%) |
| Pulmonary valve/conduit procedures | 105 (37.8%) |
| Right-sided lesions | |
| Mitral valve procedures | 28 (13.9%) |
| Aortic valve procedures | 16 (8%) |
| Tricuspid valve procedures | 66 (32.8%) |
| Pulmonary valve/conduit procedures | 109 (54.2%) |
| Conotruncal lesions | |
| Mitral valve procedures | 19 (8.4%) |
| Aortic valve procedures | 35 (15.4%) |
| Tricuspid valve procedures | 35 (15/4%) |
| Pulmonary valve/conduit procedures | 137 (60.4%) |
| Septal lesions | |
| Mitral valve procedures | 102 (27.5%) |
| Aortic valve procedures | 48 (12.9%) |
| Tricuspid valve procedures | 138 (37.2%) |
| Pulmonary valve/conduit procedures | 117 (31.5%) |
Classifications of RACHS scores according to type of cardiac lesion and type of surgical intervention are detailed in Tables 5 and 6.
Table 5.
Classification of RACHS-2 score according to CHD Type
| Left-sided lesions (n = 278) | Right-sided lesions (n = 201) | Conotruncal lesions (n = 227) | Septal lesions (n = 371) | |
|---|---|---|---|---|
| Missing | 3 (1.1%) | 2 (1%) | 3 (1.3%) | 4 (1.1%) |
| Category 1 | 33 (11.9%) | 26 (12.9%) | 26 (11.5%) | 64 (17.3%) |
| Category 2 | 60 (21.6%) | 58 (28.9%) | 84 (37%) | 105 (28.3%) |
| Category 3 | 74 (26.6%) | 37 (18.4%) | 34 (25%) | 80 (21.6%) |
| Category 4 | 28 (10.1%) | 27 (13.4%) | 42 (18.5%) | 47 (12.7%) |
| Category 5 | 52 (18.7%) | 30 (14.9%) | 15 (6.6%) | 36 (9.7%) |
Table 6.
Classification of RACHS-2 score according to type of surgical intervention
| Mitral valve procedures (n = 136) | Tricuspid valve procedures (n = 170) | Aortic valve procedures (n = 100) | Pulmonary valve/conduit procedures (n = 263) | |
|---|---|---|---|---|
| Missing | 0 | 1 (0.6%) | 1 (1%) | 2 (0.8%) |
| Category 1 | 2 (1.5%) | 44 (25.9%) | 24 (24%) | 40 (15.2%) |
| Category 2 | 53 (39%) | 43 (25.3%) | 16 (16%) | 91 (34.6%) |
| Category 3 | 54 (39.7%) | 33 (19.4%) | 39 (39%) | 39 (14.8%) |
| Category 4 | 13 (9.6%) | 14 (8.2%) | 10 (10%) | 31 (11.8%) |
| Category 5 | 3 (2.2%) | 19 (11.2%) | 6 (6%) | 28 (10.6%) |
In-hospital Outcomes
In-hospital mortality was comparable between the two groups. The requirement of extracorporeal membrane oxygenation (ECMO) was similar in both groups. Permanent pacemaker implantation was more common in the surgical group (6.6% vs 2.2%, p < 0.01). The length of hospital stay was longer in the surgical patients (33 days [IQR, 15–74] vs 26 days [8–76], p = 0.025). The estimated cost of hospitalization was significantly more expensive in the surgical group (244934 US$ [126094–609873] vs 133302 US$ [IQR, 37684–460001], p < 0.01). In-hospital outcomes are depicted in Table 7.
Table 7.
In-hospital outcomes of children undergoing cardiac surgery for IE
| Medical treatment (n = 1476) | Surgical group (n = 632) | p value | |
|---|---|---|---|
| In-hospital mortality | 142 (9.6%) | 64 (10.1%) | 0.72 |
| ECMO | 88 (6%) | 47 (7.4%) | 0.21 |
| Permanent pacemaker | 32 (2.2%) | 42 (6.6%) | < 0.01 |
| Length of hospital stay | 26 [8–76] | 33 [15–74] | 0.025 |
| Estimated cost of hospitalization | 133,302 [37684–460001] | 244,934 [126094–609873] | < 0.01 |
ECMO extracorporeal membrane oxygenation
Multivariable Analysis of Factors Associated with In-hospital Mortality
A multivariable logistic regression model was performed to identify factors associated with in-hospital mortality and thee results are presented in Table 8. Variables included in the model were selected based on clinical relevance and prior literature. Age was significantly associated with mortality, with younger patients demonstrating lower odds of death (OR = 0.9 per year decrease, 95% CI [0.88–0.98], p ≤ 0.01).
Table 8.
Multivariable analysis for factors associated with in-hospital mortality
| Factors | OR | 95% CI | p value |
|---|---|---|---|
| Age in years | 0.93 | 0.88–0.98 | < 0.01 |
| Staphylococcus infection | 0.86 | 0.51–1.40 | 0.51 |
| RACHS score | 1.55 | 1.27–1.89 | < 0.01 |
| Requirement of cardiac surgery | 1.46 | 0.90–2.3 | 0.12 |
| Heart failure | 1.74 | 1.12–2.71 | 0.014 |
| Prosthetic IE | 1.64 | 0.18–1.85 | 0.21 |
| Embolic Stroke | 2.6 | 1.5–4.5 | < 0.01 |
| Underlying left-sided lesions | 1.46 | 0.90–2.4 | 0.14 |
| Underlying right-sided lesions | 0.53 | 0.33–0.86 | 0.009 |
| Underlying septal defects | 1.38 | 0.83–2.28 | 0.21 |
| Underlying conotruncal lesions | 0.88 | 055–1.42 | 0.61 |
| Multi-valve surgery | 0.57 | 0.18–1.85 | 0.35 |
| ECMO | 5.44 | 3.21–9.20 | < 0.01 |
| PPM | 1.77 | 0.85–3.7 | 0.13 |
OR odds ratio, RACHS_2 score risk adjustment for congenital heart surgery score, ECMO extracorporeal membrane oxygenation, PPM permanent pacemaker
ECMO requirement was significantly associated with in-hospital mortality after accounting for potential confounders (OR = 5.4, 95% CI [3.2–9.2], p < 0.01). Presence of heart failure was a strong risk factor for in-hospital death (OR = 1.7, 95% CI [1.1–2.7], p = 0.01). Lastly, the occurrence of stroke was also predictive of mortality (OR = 2.6, 95%CI [1.5–4.5], p < 0.01). Patients with underlying right-sided cardiac lesions were inversely related to in-hospital mortality (OR = 0.5, 95% CI [0.3–0.8], p = 0.009).
Discussion
The Incidence of IE in Pediatric Patients with CHD
Our data demonstrate that the estimated incidence of IE in hospitalized children with CHD is 5.6 per 1000 unique cases, substantially higher than in the general pediatric population [7]. Our finding aligns with a population-based study from the Quebec CHD database from 1988 to 2010 [14] which reported an incidence of 6.1 cases per 1000 children (birth—18 years of age). Another population-based analysis from Taiwan from 1997 to 2005 found that the overall incidence of IE in children with CHD (n = 24729) who were followed until 2010 was 11.13 per 10000 person-years [5]. The rate of IE in adults with CHD (ACHD) is greater than that in children with CHD. A nationwide study from the Netherlands [15] demonstrated that the overall incidence of IE in ACHD patients is 1.33 cases per 1000 person-years, approximately 3 times higher than the reported incidence in children from the Quebec CHD database (0.41 cases/1000 person-years). The authors explained the overall difference by cardiac lesion distribution and prevalence of prosthetic materials, with left-sided CHD lesions being more common in ACHD patients compared to the pediatric cohort. Additionally, adult patients are more vulnerable to developing IE due to the greater complexity of their cardiac disease and higher incidence of comorbidities and acquired diseases [16].
Age-Related Patterns and Potential Risk Factors
In accordance with previously published reports [9, 17, 18], in our cohort, the distribution of IE was relatively even across the three age groups: infants (0–1 year), middle childhood (1–9 years) and adolescents (9–21 years). Infants and young children are at higher risk for prolonged hospitalization in neonatal and pediatric intensive care units, high rates of indwelling venous catheters, and other invasive procedures, which can increase the incidence of healthcare-associated infective endocarditis. On the other hand, adolescents with CHD are more likely to experience IE for various reasons, including a higher prevalence of cardiac prosthesis, presence of residual shunts, and behavioral risk factors, including intravenous illicit drug use, body piercing or tattoos under non-sterile conditions, and undergoing dental procedures without prophylaxis. Prior literature corroborates these associations, emphasizing the importance of targeted education and preventive care in adolescents with CHD [19, 20].
Surgical Interventions and Clinical Complexity
Approximately, 30% with CHD and IE in our cohort underwent cardiac surgery. This surgery frequency is similar to the frequency of surgery in a prior cohort of IE in pediatric patients not limited to patients with pre-existing CHD [9] that reported an incidence of 33.8% of pediatric patients with pre-existing heart disease admitted for IE; however, it was unclear whether the cardiac surgery occurred before or after the IE diagnosis. Literature reported an incidence ranging between 25 and 50% in the general population [10, 11]. Pulmonary valve replacement or right ventricle to pulmonary artery conduit replacement was the most common performed surgery in our cohort. These conduits are usually homografts containing valves and are performed in cyanotic lesions including tetralogy of Fallot with pulmonary atresia. These conduits tend to degenerate and constrict over time, resulting in stenosis and regurgitation, which create a high-risk environment for seeding bacteria. A prior study suggests that different types and materials of RV-PA conduits may be associated with a higher risk of IE, such as bovine jugular grafts that were associated with a substantially higher risk of late IE compared to other valved conduits [21]. Surgery for pulmonary valve and/or right ventricular outflow tract (RVOT) IE ranges from 14% to 28.3% in single-center experiences [18, 22, 23]. In accordance with a large prospective cohort in the adult population [24], our data demonstrate that prosthetic valve endocarditis was present in 9.4% of patients. Interestingly, we did not observe any difference in in-hospital mortality between prosthetic valve endocarditis and those with native endocarditis, suggesting that modern perioperative care and the emergence of dedicated pediatric cardiac intensive care units may mitigate some excess risk. Prosthetic valve IE was significantly associated with death, demonstrated by a retrospective study from a tertiary center in Australia [23].
In-hospital Outcomes and Healthcare Resource Utilization
The overall in-hospital mortality in our investigation was 9.8% in the whole cohort, falling within the range reported in other pediatric IE studies (2.1%–15%) [10, 18, 23, 25]. Notably, our results reveal no difference in in-hospital mortality between those who required cardiac surgery and those who were managed medically. This is consistent with findings from large retrospective cohorts and supports the notion that surgical intervention in pediatric IE, when indicated, does not inherently confer increased perioperative risk [22, 25, 26].Our cohort also illustrates substantial utilization of healthcare resources in the surgical group. The surgical patients were sicker, with higher rates of heart failure, septic emboli, and postoperative permanent pacemaker implantation.
Predictors of In-hospital Mortality
In agreement with the literature, multivariable analysis identified that increasing age was associated with higher in-hospital mortality. This can be possibly explained by the higher complexity of long-standing CHD, presence of multiple prior surgeries, ventricular dysfunction, and pulmonary hypertension. Heart failure was common in our cohort with an incidence of 28.7% and was predictive of in-hospital mortality. Heart failure can result from valve regurgitation, particularly aortic valve, and is considered an indication for cardiac surgery according to the AHA guidelines [27]. Embolic stroke was significantly associated with in-hospital mortality on multivariable analysis. Embolic events are common complications of IE with an incidence between 30 and 40% of left-sided IE [28, 29]. Maraey et al. [30]queried the National Inpatient Database from 2016 to 2020 for IE patients with and without embolic stroke. They concluded that embolic stroke is significantly associated with lower survival rates, prolonged hospitalization, and healthcare resource utilization.
In conclusion, this study reveals several significant results regarding the characteristics and outcomes of IE in CHD patients. Despite still being a rare complication, IE occurs with an estimated incidence of 5.6 per 1000 hospitalized patients with CHD. IE most commonly occurs in two age groups—infants and adolescents—suggesting age-related risk factors.
Cardiac surgery is required in nearly one-third of IE patients, with replacement of the pulmonary valve or right ventricle to pulmonary artery conduit and tricuspid valve surgery being the most common procedures. Despite the substantial operative risk, in-hospital mortality was not significantly different between operated patients and nonoperated counterparts.
However, patients who underwent surgery had more complex clinical courses, including increased rates of postoperative permanent pacemaker implantation, increased duration of hospitalization, and increased health care utilization. Notably, several factors including older age at the diagnosis of IE, stroke, and ECMO requirement were significantly associated with in-hospital mortality and are useful benchmarks for risk stratification.
Limitations
Our study is limited by the inherent factors in a large retrospective database. PHIS database is an administrative database and is subject to miscoding [31]. However, the CHA follows strict quality measures in order to minimize errors in coding and data entry from the participating hospitals. Our data is obtained from 50 tertiary children’s hospitals in the USA, which limited the generalizability of our results. Comorbidities, clinical diagnoses, and index procedures were identified using ICD-10 and CPT codes; hence, granular results regarding severity of disease, echocardiographic data, and more details about the index procedures are lacking. A key limitation of PHIS database is the inability to determine the temporal sequence of events relative to index surgery, as PHIS database is encounter based and does not include timestamps for discharge diagnoses. It is not possible to distinguish whether certain variables such as stroke occurred preoperatively or postoperatively. This limits our ability to accurately identify preoperative risk factors and may introduce misclassification bias. Our analysis also includes the hospitalization during which cardiac surgery was performed and does not capture the medical-only admissions for patients who may have undergone staged or delayed surgical intervention after discharge. As such, our cohort may limit generalizability to all children with IE, particularly those managed medically or with planned delayed surgery.
Conclusion
Infective endocarditis in children with CHD has substantial mortality and morbidity with higher incidence in infants and adolescents. Surgical patients demonstrate a high burden of disease with a greater need for healthcare resource utilization. While cardiac surgery is often required, it does not impact hospital mortality. Older age, embolic stroke, and heart failure were associated with worse outcomes. These findings highlight the importance of early recognition and prevention of IE in this vulnerable population.
Supplementary Information
Below is the link to the electronic supplementary material.
Author Contributions
Mohamed Elsisy: designed the idea, performed statistical analysis, drafted the manuscript. Tammy Ruth: critical revision of the manuscript. Garrett Coyan: critical revision of the manuscript. Daniel Dulek: critical revision of the manuscript. David Parra: designed the idea, revised the manuscript.
Funding
None.
Data Availability
Data supporting our results are available. Our cohort is extracted from a publicly available database (Pediatric Health Information System [PHIS]) database. Our spreadsheet and analysis are stored in Excel files and we are willing share our analysis as requested.
Declarations
Conflict of interest
The authors declare no competing interests.
Footnotes
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Cahill TJ et al (2019) Contemporary epidemiology of infective endocarditis in patients with congenital heart disease: a UK prospective study. Am Heart J 215:70–77 [DOI] [PubMed] [Google Scholar]
- 2.Cahill TJ et al (2017) Challenges in infective endocarditis. J Am Coll Cardiol 69(3):325–344 [DOI] [PubMed] [Google Scholar]
- 3.Eleyan L et al (2021) Infective endocarditis in paediatric population. Eur J Pediatr 180:3089–3100 [DOI] [PubMed] [Google Scholar]
- 4.Baltimore RS et al (2015) Infective endocarditis in childhood: 2015 update: a scientific statement from the American Heart Association. Circulation 132(15):1487–1515 [DOI] [PubMed] [Google Scholar]
- 5.Sun LC et al (2017) Risk factors for infective endocarditis in children with congenital heart diseases - a nationwide population-based case control study. Int J Cardiol 248:126–130 [DOI] [PubMed] [Google Scholar]
- 6.Moore B et al (2017) Incidence, predictors and outcomes of infective endocarditis in a contemporary adult congenital heart disease population. Int J Cardiol 249:161–165 [DOI] [PubMed] [Google Scholar]
- 7.Elder RW, Baltimore RS (2015) The changing epidemiology of pediatric endocarditis. Infect Dis Clin North Am 29(3):513–524 [DOI] [PubMed] [Google Scholar]
- 8.Snygg-Martin U et al (2021) Cumulative incidence of infective endocarditis in patients with congenital heart disease: a nationwide, case-control study over nine decades. Clin Infect Dis 73(8):1469–1475 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Day MD et al (2009) Characteristics of children hospitalized with infective endocarditis. Circulation 119(6):865–870 [DOI] [PubMed] [Google Scholar]
- 10.Murakami T et al (2012) Factors associated with surgery for active endocarditis in congenital heart disease. Int J Cardiol 157(1):59–62 [DOI] [PubMed] [Google Scholar]
- 11.Prendergast BD, Tornos P (2010) Surgery for infective endocarditis: who and when? Circulation 121(9):1141–1152 [DOI] [PubMed] [Google Scholar]
- 12.Hoen B, Duval X (2013) Infective endocarditis. N Engl J Med 368(15):1425–1433 [DOI] [PubMed] [Google Scholar]
- 13.Pasquali SK et al (2012) Trends in endocarditis hospitalizations at US children’s hospitals: impact of the 2007 American Heart Association antibiotic prophylaxis guidelines. Am Heart J 163(5):894–899 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Rushani D et al (2013) Infective endocarditis in children with congenital heart disease. Circulation 128(13):1412–1419 [DOI] [PubMed] [Google Scholar]
- 15.van Melle JP et al (2023) Infective endocarditis in adult patients with congenital heart disease. Int J Cardiol 370:178–185 [DOI] [PubMed] [Google Scholar]
- 16.Nakagawa N (2022) Infective endocarditis in congenital heart disease. Endocarditis-diagnosis and treatment. IntechOpen, London [Google Scholar]
- 17.Sélégny M et al (2021) Infective endocarditis in children: a 10-year multicentric study. Arch Cardiovasc Dis Suppl 13(4):300–301 [Google Scholar]
- 18.Russell HM et al (2013) Outcomes of surgical therapy for infective endocarditis in a pediatric population: a 21-year review. Ann Thorac Surg 96(1):171–175 [DOI] [PubMed] [Google Scholar]
- 19.Armstrong ML, DeBoer S, Cetta F (2008) Infective endocarditis after body art: a review of the literature and concerns. J Adolesc Health 43(3):217–225 [DOI] [PubMed] [Google Scholar]
- 20.Carvajal V et al (2024) Endocarditis in adult congenital heart disease patients: prevention, recognition, and management. Curr Cardiol Rep 26(9):1031–1045 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Mery CM et al (2016) Risk factors for development of endocarditis and reintervention in patients undergoing right ventricle to pulmonary artery valved conduit placement. J Thorac Cardiovasc Surg 151(2):432-9-441.e1-2 [DOI] [PubMed] [Google Scholar]
- 22.Carrillo SA et al (2023) Surgical outcomes of infective endocarditis in pediatrics: moving the needle to a contemporary, multidisciplinary approach. J Thorac Cardiovasc Surg 165(1):275–284 [DOI] [PubMed] [Google Scholar]
- 23.Khoo B et al (2019) Outcomes of surgery for infective endocarditis in children: a 30-year experience. J Thorac Cardiovasc Surg 158(5):1399–1409 [DOI] [PubMed] [Google Scholar]
- 24.Murdoch DR et al (2009) Clinical presentation, etiology, and outcome of infective endocarditis in the 21st century: the International Collaboration on Endocarditis-Prospective Cohort Study. Arch Intern Med 169(5):463–473 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Shamszad P et al (2013) Early surgical therapy of infective endocarditis in children: a 15-year experience. J Thorac Cardiovasc Surg 146(3):506–511 [DOI] [PubMed] [Google Scholar]
- 26.Nomura F et al (1995) Surgical intervention for infective endocarditis in infancy and childhood. Ann Thorac Surg 60(1):90–95 [PubMed] [Google Scholar]
- 27.Baddour LM et al (2015) Infective endocarditis in adults: diagnosis, antimicrobial therapy, and management of complications. Circulation 132(15):1435–1486 [DOI] [PubMed] [Google Scholar]
- 28.Steckelberg JM et al (1991) Emboli in infective endocarditis: the prognostic value of echocardiography. Ann Intern Med 114(8):635–640 [DOI] [PubMed] [Google Scholar]
- 29.Vilacosta I et al (2002) Risk of embolization after institution of antibiotic therapy for infective endocarditis. J Am Coll Cardiol 39(9):1489–1495 [DOI] [PubMed] [Google Scholar]
- 30.Maraey A et al (2023) Abstract 11693: comparison of in-hospital outcomes in infective endocarditis patients with and without embolic stroke. Circulation 148(Suppl_1):A11693–A11693 [Google Scholar]
- 31.Strickland MJ et al (2008) The importance of nomenclature for congenital cardiac disease: implications for research and evaluation. Cardiol Young 18(S2):92–100 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Data supporting our results are available. Our cohort is extracted from a publicly available database (Pediatric Health Information System [PHIS]) database. Our spreadsheet and analysis are stored in Excel files and we are willing share our analysis as requested.