Highlights
-
•
Infective endocarditis during pregnancy is rare but carries high maternal and fetal morbidity and mortality.
-
•
Our study quantifies the contemporary burden of infective endocarditis in pregnancy and demonstrates its strong association with increased in-hospital mortality and a broad spectrum of cardiovascular, thromboembolic, neurologic, and renal complications at delivery.
-
•
Our findings support the heightened surveillance and multidisciplinary cardio-obstetric care, while informing future research and policy efforts focused on early diagnosis, prevention, and standardized management of infective endocarditis in pregnancy.
Infective endocarditis (IE) in pregnancy is a rare but serious infection that increases morbidity and mortality. We aimed to assess the risk factors and outcomes in pregnant patients with IE as compared to pregnancy without IE. The National Inpatient Sample was used to identify hospitalizations for delivery from 2011 to 2020. Propensity score matching was performed to study the association of IE with the primary outcomes of in-hospital medical and obstetric complications. A total of 37,482,206 weighted delivery hospitalizations in women ≥18 years were identified, of which 2020 patients had IE. Pregnant patients with IE had a higher incidence of complications during delivery, including in-hospital mortality, pre-eclampsia/eclampsia, peripartum cardiomyopathy, acute coronary syndrome, stroke, acute kidney injury, pulmonary edema, cardiac arrhythmias, and venous thromboembolism compared to those without IE. In propensity-matched analysis, IE was significantly associated with a higher risk of in-hospital mortality (odds ratio 16.50 [95% confidence interval 3.65 to 74.60], p <0.001), acute kidney injury (30.70 [11.00 to 86.00], p <0.001), stroke (18.00 [4.02 to 80.60], p <0.001), cardiac arrhythmias (4.67 [3.19 to 6.83], p <0.001), and venous thromboembolism (7.50 [2.35 to 23.90], p <0.001) compared to patients without IE. In conclusion, delivery hospitalizations in patients with IE are associated with a high risk of mortality and complications.
Central illustration: Central figure depicting the key findings of our study.
Infective endocarditis (IE) is an infection of the endocardium that involves both native and prosthetic valves. While rare, the United States has seen an increase in the incidence of IE from 10.2 per 100,000 population in 1990 to 14.4 per 100,000 in 2019. This is a reflection of several factors, including an increase in intravenous drug use (IVDU), hemodialysis dependence, an increase in the use of cardiac devices (e.g., pacemakers and implantable cardioverter-defibrillators [ICDs]), and an aging population with age-related cardiovascular disease. ,, Notably, the incidence has increased in women. , Pregnancy is a state of relative immunosuppression with an increased susceptibility to infection. The incidence of IE in pregnancy has historically been reported as 0.006%. Increasing rates have been seen in recent years, paralleling the opioid pandemic. , IE in pregnancy is known to significantly increase maternal and fetal mortality and morbidity. ,, Cardiovascular disease, in general, is estimated to cause 26.5% of pregnancy-related deaths. Cardiovascular adaptations to pregnancy in the setting of valvular disease and infection confer substantial hemodynamic compromise during pregnancy and in the postpartum period. Management is individualized and exceptionally difficult, as antibiotic use is entangled with potential teratogenicity, and valvular surgery involving cardiopulmonary bypass is associated with poor fetal outcomes. , This underscores the importance of exploring complications that arise in those with IE during pregnancy. We hypothesized that delivery hospitalizations in patients with IE would be linked to higher in-hospital mortality and complications such as peripartum cardiomyopathy, preeclampsia/eclampsia, cardiac arrhythmias, acute coronary syndrome, acute kidney injury, stroke, pulmonary edema, and venous thromboembolism. To date, there are only a few studies exploring the complications of IE in pregnancy. ,, This current nationwide study aimed to evaluate patient characteristics of pregnant patients with IE undergoing delivery, the prevalence of maternal complications at delivery, and the association of IE with these complications.
Methods
Data source
The National Inpatient Sample (NIS) database from 2011 to 2020 was used to conduct this study. The NIS is the largest inpatient healthcare database managed by the Agency for Healthcare Research and Quality via the Healthcare Cost and Utilization Project (HCUP). About 20% of hospitals in all the states participating in HCUP contribute to NIS, which contains data of >7 million hospitalizations annually and is representative of >97% of the United States population. State, hospital, and patient identifiers are completely absent from the database to ensure patient confidentiality. Institutional Review Board approval is not required for this study as NIS uses publicly available, anonymous, and deidentified data. Patients and the public were not involved in the design, conduct, reporting, or dissemination of this research.
Study population and covariates
Supplementary Table 1 in the supplementary file describes the International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM), and ICD-9-CM codes, which were used to identify delivery hospitalizations and IE. Individuals aged <18 years were excluded from our study. Variables including demographic characteristics (age, race, or ethnicity) and clinical history (hypertension, diabetes mellitus, hyperlipidemia, obesity, chronic heart failure, coronary artery disease, chronic kidney disease, history of valve replacement, history of ICD/permanent pacemaker placement, and substance abuse) were collected for analysis. Maternal complications at delivery were compared between pregnant patients with endocarditis and those without endocarditis. The primary outcomes of interest were in-hospital mortality, pre-eclampsia, eclampsia, peripartum cardiomyopathy, acute coronary syndrome, stroke, pulmonary edema, cardiac arrhythmias, acute kidney injury, and venous thromboembolism. Given the small number of eclampsia cases, these cases were combined with pre-eclampsia. Since NIS data are collected yearly, they can be utilized to analyze disease trends across duration using trend weights established by HCUP.
Statistical analysis
Descriptive statistics are presented as mean ± standard deviation for continuous variables and percentages for categorical variables. Categorical variables were compared using chi-squared or Fisher’s exact test, while Student’s t test was used for continuous variables (age) to assess statistical significance. An univariable model and a multivariable logistic regression model (adjusted for age, race or ethnicity, region, median income, chronic hypertension, diabetes, dyslipidemias, heart failure, coronary artery disease, chronic kidney disease, obesity, smoking, multiple gestation, and cesarean delivery) were performed to examine the association of endocarditis with maternal cardiovascular complications at delivery. For further confirmation, propensity score matching was also used to estimate the likelihood of the development of outcomes among patients with endocarditis undergoing delivery. The propensity score was estimated using logistic regression based on age, race or ethnicity, median income, diabetes mellitus, hypertension, hyperlipidemia, chronic heart failure, obesity, coronary artery disease, smoking, chronic kidney disease, hospital region, multiple gestation, and cesarean delivery. One-to-three nearest-neighbor matching was used. A good balance was achieved between patients with and without IE undergoing delivery, with all the standardized mean differences below 0.1 after matching.
A p-value of <0.05 was considered statistically significant, and all tests were 2-tailed. The Bonferroni correction method was employed to address the challenge of multiple comparisons and to control potential Type I error rates. This correction entails adjusting the significance levels (p-values) for multiple comparisons. The designated overall significance level, typically set at 0.05, was divided by the number of comparisons, yielding a corrected alpha error threshold of 0.005. The adjusted significance level was compared with the calculated p-value for each comparison. Statistical significance, post-Bonferroni correction, was established if the adjusted p-value was less than or equal to the new significance level (p <0.005). Most missing values were found in the variable for race and ethnicity (4.8%), and these patients were excluded from the analysis. If missing data from other variables occurred, these patients were also excluded before the analysis was conducted. The primary analysis examined the association of an overall diagnosis of endocarditis with pregnancy complications at delivery.
The analysis was conducted using StataCorp 2021 [Stata Statistical Software: Release 17. StataCorp LLC (College Station, TX)]. Appropriate weights provided by HCUP to generate national estimates were used in all estimations. Propensity matching analysis was conducted using the R software (version 4.3.1) and the MatchIt package (R Foundation for Statistical Computing, Vienna, Austria).
Results
Study population
A total of 37,482,206 weighted delivery hospitalizations in pregnant patients aged ≥18 years were identified from the NIS database from 2011 to 2020, out of which 2020 patients had IE. The study flow chart is shown in Figure 1 . Compared to delivery hospitalization in patients without endocarditis, those with endocarditis were almost of similar age (28.9 vs 28.7 years), more likely to have a history of smoking (21.7% vs 5.3%), cocaine abuse (8.6% vs 0.1%), and other substance abuse (cannabis, alcohol, sedative, and stimulants) and less likely to have a cesarean delivery (32.1% vs 47.2%). Among comorbidities, patients with endocarditis were also more likely to have prepregnancy hypertension (3.4% vs 0.4%), hyperlipidemia (0.99% vs 0.2%), chronic kidney disease (1.5% vs 0.08%), and coronary artery disease (2.2% vs 0.03%). The detailed baseline characteristics of the study population are described in Table 1 .
Study flowchart.
Table 1
Baseline variables in delivery hospitalization in patients with and without infective endocarditis
| Variables | Delivery without endocarditis (n = 37,480,186) | Delivery with endocarditis (n = 2020) | p-value |
|---|---|---|---|
| Mean age | 28.77 ± 5.74 | 28.98 ± 5.53 | |
| Hospital region | <0.001 | ||
| Northeast | 6,056,798 (16.16) | 305 (15.10) | |
| Midwest | 7,893,327 (21.06) | 368 (18.24) | |
| South | 14,493,587 (38.67) | 1012 (50.12) | |
| West | 9,033,211 (24.10) | 334 (16.53) | |
| Race or ethnicity | <0.001 | ||
| White | 18,848,785 (50.29) | 1348 (66.75) | |
| African American | 5,303,446 (14.15) | 221 (10.95) | |
| Hispanic | 7,428,572 (19.82) | 245 (12.13) | |
| Others | 5,899,381 (15.74) | 205 (10.16) | |
| Median income | 0.475 | ||
| Quartile 1 ($1-$49,000) | 10,295,314 | 873 | |
| Quartile 2 ($50,000-$64,999) | 9,302,136 | 544 | |
| Quartile 3 ($65,000-$85,999) | 9,243,958 | 349 | |
| Quartile 4 (≥ $86,000) | 8,189,695 | 204 | |
| Cesarean delivery | 17,698,143 (47.22) | 649 (32.15) | <0.001 |
| Multiple gestations | 745,855 (1.99) | 49 (2.44) | 0.517 |
| Hypertension | 164,912 (0.44) | 70 (3.49) | <0.001 |
| Diabetes mellitus | 404,786 (1.08) | 30 (1.49) | 0.435 |
| Hyperlipidemia | 79,064 (0.21) | 20 (0.99) | <0.001 |
| Chronic heart failure | 13,118 (0.035) | 155 (7.67) | <0.001 |
| Chronic kidney disease | 32,607 (0.087) | 30 (1.51) | <0.001 |
| Coronary artery disease | 11,618 (0.031) | 45 (2.23) | <0.001 |
| History of valve replacement | 4872 (0.013) | 50 (2.48) | <0.001 |
| History of ICD/PPM | 10,494 (0.028) | 35 (1.24) | <0.001 |
| Obesity | 3343187 (8.92) | 200 (9.9) | 0.487 |
| Smoking | 1,993,945 (5.32) | 440 (21.78) | <0.001 |
| Alcohol | 41,228 (0.11) | <11 | 0.017 |
| Cannabis | 288,597 (0.77) | 85 (4.21) | <0.001 |
| Opioids | 258,613 (0.69) | 583 (28.88) | <0.001 |
| Cocaine use | 67,464 (0.18) | 175 (8.66) | <0.001 |
| Sedative use | 13,492 (0.036) | 30 (1.49) | <0.001 |
| Stimulant use | 112,440 (0.30) | 140 (6.90) | <0.001 |
| Complications | |||
| In-hospital mortality | 2548 (0.0068) | 55 (2.72) | <0.001 |
| Preeclampsia/eclampsia | 1,311,806 (3.5) | 128 (6.35) | 0.001 |
| Peripartum cardiomyopathy | 712,123 (0.019) | 35 (1.73) | <0.001 |
| Acute coronary syndrome | 2661 (0.0071) | 25 (1.24) | <0.001 |
| Stroke | 3186 (0.0085) | 55 (2.72) | <0.001 |
| Pulmonary edema | 19,115 (0.051) | 464,754 (1.24) | <0.001 |
| Cardiac arrhythmia | 329,825 (0.88) | 233 (11.53) | <0.001 |
| Venous thromboembolism | 20,614 (0.055) | 49 (2.45) | <0.001 |
| Acute kidney injury | 31,108 (0.083) | 205 (10.13) | <0.001 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
