Infective endocarditis (IE) is a rare and life-threatening disease caused primarily by a bacterial infection affecting heart structures, mainly the valves [1]. Microbiological identification of the etiological factors of IE is crucial because it allows for the administration of targeted antibiotic therapy. The main bacteria causing IE in the human population in Europe is Staphylococcus aureus (31%) [2–3], although some studies have shown an increasing incidence of Enterococcus faecalis over the last few years [4]. Nevertheless, in many cases, we encounter blood culture-negative infective endocarditis (BCNIE) [1]. This condition is mostly caused by antibiotic administration prior to performing blood cultures, fungi, or fastidious bacteria. The aim of this study was to analyze the etiological factors of IE based on blood culture results performed during hospitalization in the cardiology department and their impact on in-hospital mortality in our tertiary facility.
We retrospectively analyzed all consecutive cases of IE admitted to our tertiary center, which covers a population of 2.5 million people, from 2015 to 2023. Analyzed parameters included gender, age, number of cases, number of patients qualified for cardiac surgery or conservative treatment, in-hospital mortality (divided into 3 categories: overall, preoperative, and perioperative), and the etiological factor (with 8 categories, as listed in Table 1). Preoperative in-hospital mortality was defined as death occurring before cardiac surgery or in patients qualified for conservative treatment, while perioperative in-hospital mortality was defined as death occurring during cardiac surgery or after the procedure. Unless known intolerance or allergies were present, all patients received similar guideline-directed antibiotic therapy: empirical antibiotic therapy in all patients before blood culture results, continued in patients with BCNIE, and targeted antibiotic therapy in patients with identified IE pathogens.
Table 1.
Results of blood culture analysis and the impact of etiologic factor on overall, pre-, and perioperative in-hospital mortality (logistic regression analysis results)
| Etiologic Factor overall (2015–2023) | N | % | ||||||
|---|---|---|---|---|---|---|---|---|
| Unidentified etiology | 88 | 40.7% | ||||||
| Staphylococcus spp. | 64 | 29.6% | ||||||
| Enterococcus spp. | 28 | 13.0% | ||||||
| Streptococcus spp. | 20 | 9.3% | ||||||
| Mixed (more than one type of bacteria confirmed) | 10 | 4.6% | ||||||
| E. Coli | 3 | 1.4% | ||||||
| Pseudomonas aeruginosa | 1 | 0.5% | ||||||
| Other | 2 | 0.9% | ||||||
|
| ||||||||
| Most common etiologic factor by year | ||||||||
|
| ||||||||
| Unidentified etiology | Staphylococcus spp. | Enterococcus spp. | Streptococcus spp. | |||||
|
| ||||||||
| N | % | N | % | N | % | N | % | |
|
| ||||||||
| 2015 | 16 | 52% | 6 | 19% | 3 | 10% | 2 | 6% |
| 2016 | 9 | 30% | 12 | 40% | 4 | 13% | 4 | 13% |
| 2017 | 12 | 60% | 5 | 25% | 1 | 5% | 2 | 10% |
| 2018 | 11 | 41% | 8 | 30% | 3 | 11% | 4 | 15% |
| 2019 | 6 | 29% | 8 | 38% | 2 | 10% | 2 | 10% |
| 2020 | 8 | 40% | 7 | 35% | 3 | 15% | 2 | 10% |
| 2021 | 11 | 42% | 6 | 23% | 5 | 19% | 1 | 4% |
| 2022 | 4 | 21% | 8 | 42% | 5 | 26% | 1 | 5% |
| 2023 | 11 | 50% | 4 | 18% | 2 | 9% | 2 | 9% |
|
| ||||||||
| Univariate logistic regression analysis results | ||||||||
|
| ||||||||
| Overall deaths | OR (odds ratio) | Lower 95% confidence interval | Upper 95% confidence interval | P | ||||
|
| ||||||||
| Staphylococcus spp. | 0.65 | 0.32 | 1.28 | 0.211 | ||||
| Streptococcus spp. | 0.63 | 0.2 | 1.95 | 0.418 | ||||
| Enterococcus spp. | 0.68 | 0.26 | 1.76 | 0.424 | ||||
| E. Coli | 1.31 | 0.12 | 14.66 | 0.829 | ||||
| Unidentified etiology | 4.22 | 1.15 | 15.53 | 0.03 | ||||
|
| ||||||||
| Preoperative deaths | OR (odds ratio) | Lower 95% confidence interval | Upper 95% confidence interval | P | ||||
|
| ||||||||
| Staphylococcus spp. | 1.72 | 0.77 | 3.81 | 0.184 | ||||
| Streptococcus spp. | 0.67 | 0.15 | 3.03 | 0.6 | ||||
| Enterococcus spp. | 0.2 | 0.03 | 1.55 | 0.125 | ||||
| E. Coli | 3.17 | 0.28 | 36.11 | 0.352 | ||||
| Unidentified etiology | 1.59 | 0.32 | 7.87 | 0.57 | ||||
|
| ||||||||
| Perioperative deaths | OR (odds ratio) | Lower 95% confidence interval | Upper 95% confidence interval | P | ||||
|
|
||||||||
| Staphylococcus spp. | 0.14 | 0.03 | 0.62 | 0.009 | ||||
| Streptococcus spp. | 0.67 | 0.15 | 3.03 | 0.6 | ||||
| Enterococcus spp. | 1.42 | 0.49 | 4.07 | 0.52 | ||||
| Unidentified etiology | 4.62 | 1.22 | 17.46 | 0.024 | ||||
|
| ||||||||
| Multivariate logistic regression analysis results | ||||||||
|
| ||||||||
| Perioperative deaths | OR (odds ratio) | Lower 95% confidence interval | Upper 95% confidence interval | P | ||||
|
| ||||||||
| Staphylococcus | 0.17 | 0.04 | 0.73 | 0.017 | ||||
| Unidentified etiology | 2.56 | 0.61 | 10.77 | 0.199 | ||||
Statistical analysis was performed using Statistica 13.1 software (Tibco, Palo Alto, CA, USA) and R version 4.2.1. Nominal values are presented as both absolute values (with mean or median depending on whether the distribution is normal or not) and percentages. To assess the impact of particular pathogens on in-hospital mortality (including overall, pre-, and perioperative deaths), we used logistic regression. P-values < 0.05 were considered statistically significant.
A total of 216 consecutive IE cases were included in the analysis. Most of the patients were men (75.5%), and the median age was 61.5 years (interquartile range, IQR: 50–68.5). From the whole study group, 64.8% of patients were qualified for surgical treatment (n = 140). The overall inhospital mortality was 27.8% (n = 60), of which half of the patients died during or after cardiac surgery (perioperative deaths, n = 30). The most common IE localization was left-sided IE (91.2%, n = 197), with the aortic valve affected in 62.5% (n = 135) and the mitral valve in 41.2% (n = 89). The results of the blood culture analysis are presented in Table 1. In most cases, the etiological factor was unidentified (negative blood culture results). The univariate logistic regression identified an unidentified etiological factor as a risk factor for overall in-hospital death and perioperative deaths, in both cases increasing the risk of in-hospital death more than 4-fold. Moreover, patients with Staphylococcus spp. etiology had a better prognosis.
Our results are consistent with the current data on IE epidemiology with Staphylococcus spp. as the most common etiological factor. Similar results were presented by other authors on larger populations, such as Ambrosioni et al. (Staphylococcus spp. — 39.9%, Streptococcus spp. — 32%, Enterococcus spp. — 10.5%) [6]. On the other hand, current ESC guidelines for the management of IE [1] and other studies also highlight the increasing number of Enterococcus spp. cases [4].
Most importantly, our study showed the scale of the problem of the unidentified etiological factor in IE patients (over 40%). However, data on BCNIE frequency are inconsistent. In the EURO-ENDO Registry, only 16.8% of cases were BCNIE [5], while another study by Pecoraro et al. [7] showed as much as 55.6% of negative blood culture results in IE patients. Unfortunately, the recently published results of the Polish IE registry (POL-ENDO) did not provide information on the IE etiological factor in the Polish population [8]. The aforementioned EURO-ENDO registry showed that mortality in BCNIE was significantly higher, and negative blood culture results were a predictor of one-year mortality (HR1.28, 95%CI 1.04–1.56, p = 0.02) [5]. It is crucial to collect samples for blood culture analysis before the administration of antibiotics because the first dose of an antibiotic may interfere with the results, especially in septic/infective patients with any signs or symptoms of heart failure, which may suggest infection affecting heart structures. Early identification of the etiological factor allows the administration of targeted antibiotic treatment and may potentially improve prognosis.
The single-center character of this study is its main limitation. We also did not single out S. aureus (the most common etiological factor) cases from the Staphylococcus spp. group in our analysis. In cases of negative blood cultures, tissue ribosomal ribonucleic acid and plasma microbial cell-free deoxyribonucleic acid sequencing were not performed — the results of these analyses might decrease the frequency of BCNIE. In a prospective study by Witten et al. on a group of 43 IE patients, postoperative tissue sampling yielded positive culture results in all blood culture-negative patients [9]. Caroselli et al., in their case report and review, emphasize the importance of fungi as an IE etiological factor, reporting that only 10% (out of 74) of the patients with Aspergillus spp. IE in the review had positive blood cultures [10]. On the other hand, most of the results on IE etiology in those studies were collected postoperatively. This would not change the preoperative management of IE patients because all BCNIE patients should receive guideline-directed empirical antibiotic therapy. Our study raises an important issue of the need to improve diagnostic pathways in septic patients in other hospitals to reduce the incidence of negative blood culture results in confirmed IE patients transferred from lower-level cardiology or internal medicine wards.
In summary, in most IE cases, the etiological factor was unidentified, which increased the risk of overall and perioperative death. Moreover, the number of Enterococcus spp. IE is rising. It is crucial to collect samples for blood cultures before the antibiotic treatment administration, especially in patients presenting symptoms suggesting infective endocarditis or heart failure.
Footnotes
Conflict of interest: None declared.
Funding: None declared.
References
- 1.Delgado V, Ajmone Marsan N, de Waha S, et al. ESC Scientific Document Group. 2023 ESC Guidelines for the management of endocarditis. Eur Heart J. 2023;44(39):3948–4042. doi: 10.1093/eurheartj/ehad193. [DOI] [PubMed] [Google Scholar]
- 2.Murdoch DR, Corey GR, Hoen B, et al. International Collaboration on Endocarditis-Prospective Cohort Study (ICE-PCS) Investigators. Clinical presentation, etiology, and outcome of infective endocarditis in the 21st century: the International collaboration on Endocarditis-Prospective Cohort Study. Arch Intern Med. 2009;169(5):463–473. doi: 10.1001/archinternmed.2008.603. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Habib G, Erba PA, Iung B, et al. EURO-ENDO Investigators. Clinical presentation, aetiology and outcome of infective endocarditis. Results of the ESC-EORP EURO-ENDO (European infective endocarditis) registry: a prospective cohort study. Eur Heart J. 2019;40(39):3222–3232. doi: 10.1093/eurheartj/ehz620. [DOI] [PubMed] [Google Scholar]
- 4.Escolà-Vergé L, Fernández-Hidalgo N, Larrosa MN, et al. Secular trends in the epidemiology and clinical characteristics of Enterococcus faecalis infective endocarditis at a referral center (2007–2018) Eur J Clin Microbiol Infect Dis. 2021;40(6):1137–1148. doi: 10.1007/s10096-020-04117-x. [DOI] [PubMed] [Google Scholar]
- 5.Kong WKF, Salsano A, Giacobbe DR, et al. The EURO-ENDO Investigators. Outcomes of culture-negative vs. culture-positive infective endocarditis: the ESC-EORP EURO-ENDO registry. Eur Heart J. 2022;43(29):2770–2780. doi: 10.1093/eurheartj/ehac307. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Ambrosioni J, Hernández-Meneses M, Durante-Mangoni E, et al. International Collaboration for Endocarditis (ICE) Investigators. epidemiological changes and improvement in outcomes of infective endocarditis in europe in the twenty-first century: An International Collaboration on Endocarditis (ICE) Prospective Cohort Study (2000–2012) Infect Dis Ther. 2023;12(4):1083–1101. doi: 10.1007/s40121-023-00763-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Pecoraro AJ, Herbst PG, Pienaar C, et al. Modified Duke/European Society of Cardiology 2015 clinical criteria for infective endocarditis: time for an update? Open Heart. 2022;9(1) doi: 10.1136/openhrt-2021-001856. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Orzech JW, Zatorska K, Grabowski M, et al. Preliminary results from the Polish Infective Endocarditis Registry (POL-ENDO): Time to change clinical practice? Kardiol Pol. 2024;82(6):609–616. doi: 10.33963/v.phj.100275. [DOI] [PubMed] [Google Scholar]
- 9.Witten JC, Tan CD, Rodriguez ER, et al. Invasive aortic valve endocarditis: clinical and tissue findings from a prospective investigation. Ann Thorac Surg. 2022;113(2):535–543. doi: 10.1016/j.athoracsur.2021.03.072. [DOI] [PubMed] [Google Scholar]
- 10.Caroselli C, Suardi LR, Besola L, et al. Native-valve endocarditis: Case report and literature review. Antibiotics (Basel) 2023;12(7) doi: 10.3390/antibiotics12071190. [DOI] [PMC free article] [PubMed] [Google Scholar]