In recent years, the epidemiology of IE has shifted, with an increasing proportion of HAIE cases, driven by aging populations and the rising use of invasive procedures, highlighting the need for further investigation of this specific patient group [6, 9]. This study is the first to systematically compare HAIE and CAIE patients in Turkey and provides epidemiological data specific to our country. It highlights HAIE patients, a group we expect to encounter more frequently in the coming years, and provides insights into their epidemiological characteristics. It also offers valuable information for the early identification of high-risk patients and the development of strategies to reduce short-term mortality.
Our study found a high proportion of HAIE cases (48.1%), with 40.5% being non-nosocomial HAIE. Previous studies have reported varying figures [14,15,16,17]. Toyoda et al. reported 51% HAIE cases, with a rise in non-nosocomial HAIE from 32.1 to 35.9% over 15 years, while nosocomial HAIE declined slightly [11]. A similar rate has been reported in a study from Tukey; however, our study found a higher proportion of non-nosocomial HAIE patients [18]. Consistent with global trends, our findings reflect a shift in patient profiles and highlight the need for a better understanding of HAIE in current clinical practice.
Another striking finding in our study is the high proportion of HD patients, accounting for 29% of all patients and 60.3% of the HAIE group. HD increases the risk of IE due to factors such as bacteremia, valve calcification, and immunosuppression [7, 19]. Previous studies have shown that HD patients face a more than 70-fold increased risk of IE compared to the general population [20]. In our cohort, all HD patients were classified under HAIE, with 84% utilizing central venous catheters for vascular access. This finding aligns with evidence that central venous catheters carry a higher IE risk than arteriovenous fistulas [20]. The prominence of dialysis as a predisposing factor in our study is likely influenced by the tertiary care hospital, which treats a high volume of dialysis patients. These findings highlight the importance of optimizing vascular access strategies and infection prevention measures in dialysis patients to mitigate the risk of IE.
In terms of demographic characteristics, the average age of IE patients in our cohort was higher than in previous studies conducted in our country [5, 7, 21,22,23]. However, recent studies in Turkey have also shown a steady increase in the average age of IE patients, similar to trends observed in developed countries [8, 11, 18, 24, 25]. In our study, when analyzed separately, the average age was similar in both CAIE and HAIE groups. Although previous studies have reported varying results [17, 26]Alonso-Menchén et al. found that patients with nosocomial HAIE tend to be older compared to those with CAIE or non-nosocomial HAIE, likely reflecting the higher comorbidity burden in this patient group [16]. The epidemiological characteristics of IE have shifted towards older age groups due to the impact of emerging risk factors. Historically, RHD was the main cause of IE, but recent studies show that degenerative valve disease is now the leading risk factor [27, 28]. In our study, degenerative valve disease emerged as the most common predisposing factor in both groups, while RHD was rarely observed, particularly in the HAIE group. Recent studies in our country have also reported a decline in the prevalence of RHD compared to older data [18]. Similarly, in our study, the proportion of RHD was low in both groups, particularly in the HAIE group, reflecting the evolving epidemiology of IE and supporting national data.
Comorbidity rates were higher in HAIE patients than in CAIE patients, though not statistically significant, likely due to the tertiary care nature of our hospital. CKD was particularly notable in HAIE patients (65.1%), reflecting the high proportion of dialysis patients. Other comorbidities showed similar distributions between the groups in our study. The CCI, a measure of long-term mortality risk [29]was higher in HAIE patients compared to CAIE patients in our study (6.0 vs. 4.0, respectively). This aligns with previous research which reported higher CCI scores in HAIE patients [14, 17, 26]. The elevated CCI scores in our cohort reflect the high proportion of HAIE cases and the prevalence of comorbidities, as nearly all patients had at least one underlying condition. These findings highlight the greater burden of comorbidities in HAIE patients, emphasizing the need for tailored management strategies in this high-risk group.
In our study, the blood culture positivity rate was 64.6%, which is lower, likely due to the high rate of prior antibiotic use. However, the pathogen detection rate was higher in HAIE patients compared to CAIE patients. Although there was no statistically significant difference between HAIE and CAIE patients in terms of the identified microorganisms, CoNS was found to be relatively higher in HAIE patients (38%) compared to CAIE patients (19%). In previous studies, culture negativity rates were found to be similar between the two groups, while CoNS species were reported to be more common in the HAIE group [16, 26, 30]. The proportion of CoNS in IE patients was reported to be lower in earlier studies [8, 21, 24, 31, 32]. In recent years, a growing number of studies have reported an increase in the occurrence of CoNS IE, reflecting the changing microbiology of IE as its epidemiology evolves [16]. In our study, methicillin resistance was a significant concern. In a study from Turkey, the proportion of methicillin-resistant CoNS among IE patients was reported as 63% [21]. In a European study, the methicillin resistance rate among CoNS isolates from IE patients was reported as 48% [8]. Another recent study found a methicillin resistance rate of 66.5% in CoNS-related IE [33]. The other frequently identified pathogen, S. aureus, showed methicillin resistance (MRSA) rates of 11.1% in CAIE patients and 16.7% in HAIE patients in our study. In most previous studies, MRSA rates have been reported to be higher in HAIE patients [16, 26, 30]. Given these findings, it is crucial to consider local resistance patterns when planning empirical therapy in patients. The high resistance rate in patients highlights the importance of considering methicillin resistance when selecting empirical therapy.
In terms of valve involvement, mitral and aortic valve involvement was more common in the CAIE group. There is variation in the findings of different studies on this topic [17, 26, 30] Another noteworthy finding in our study is catheter-related IE in HD patients, with no involvement of other heart valves. HD-associated endocarditis was only found in HAIE patients (%28.6), with no significant difference in native, prosthetic valve, mural, or CIED-associated endocarditis between the groups. With the increase in HD treatment practices today, there is a growing trend in HD-associated endocarditis cases. A review of the literature reveals that cases of catheter-associated endocarditis without valve involvement have been reported [34]. It is thought that catheter-related endocarditis without valve involvement requires separate evaluation, and there is a gap in the literature on this topic.
In our study, no significant differences were observed in complication rates between CAIE and HAIE patients, consistent with previous studies [17, 26, 30]. However, notably, relapse occurred exclusively in HAIE patients, 80% of whom were on hemodialysis, and 75% had not undergone catheter or fistula exchange. In a study, the relapse rate of IE in HD patients was significantly higher than in other IE patients (9.4% vs. 2.7%) [35]. This finding suggests that chronic HD is a persistent risk factor for subsequent IE episodes. These results underscore the critical importance of optimizing vascular access strategies and implementing catheter replacement protocols to reduce the risk of infection in dialysis patients.
The 28-day, in-hospital, and 6-month mortality rates for IE were 12%, 25%, and 34%, respectively, consistent with previous findings [6, 8, 25, 31]. In Turkey, studies indicate that in-hospital mortality rates range between 15% and 33% [4, 5, 22, 23, 36]. Previous studies on 30-day mortality in IE patients, such as those by Ahtela et al. and Baddour et al., found mortality rates ranging from 11.4–16.3% [31, 37]. Ahtela et al., showed that 30-day mortality increased with age and the CCI, although no gender-related differences in mortality risk were found [37].
In our study, elevated levels of CRP, PCT, troponin, and NT-proBNP were significantly associated with 28-day mortality in univariate analysis; however, none of these biomarkers retained statistical significance in the multivariate model. Thus, while these markers cannot be considered independent predictors of mortality in our cohort, their associations suggest a potential prognostic value that warrants further investigation. Previous studies have reported similar findings. For example, in a study by Valentina Scheggi and colleagues, PCT (> 0.22 mg/dL) and CRP (> 83 mg/dL) levels were identified as strong independent risk factors for mortality [38]. Similarly, another study reported that WBC, CRP, and PCT levels were significantly higher in IE patients who developed severe complications or mortality, with PCT being highlighted as the best predictor of clinical deterioration [39]. Notably, an admission PCT > 0.5 ng/mL was strongly associated with poor prognosis [39]. In our cohort, HAIE patients had higher admission PCT levels compared to CAIE patients, and a significant proportion had PCT levels above 0.5 ng/mL at the time of admission. Additionally, previous studies have shown that elevated troponin and NT-proBNP levels are also associated with poor outcomes in IE patients [40]. Some studies have even suggested that NT-proBNP may outperform CRP as a predictor of in-hospital mortality [40, 41]. Although these biomarkers did not emerge as independent predictors in our multivariate analysis, their association with mortality in univariate analysis—supported by the literature— indicates their exploratory prognostic potential. Future large-scale and multicenter studies are warranted to validate their role in clinical risk stratification for infective endocarditis.
In our study, complications such as valve insufficiency, neurological complications, and cerebral embolism were more frequent among patients who experienced 28-day mortality, consistent with previous research highlighting the role of cerebral embolism, paravalvular abscess, and heart failure in in-hospital mortality [8, 21]. Interestingly, heart failure was not identified as a significant risk factor for 28-day mortality in our cohort, possibly due to the lack of distinction between congestive and non-congestive forms. In our study, arrhythmia was found to be significantly associated with 28-day mortality in univariate analysis. Although myocardial infarction and arrhythmia are uncommon complications in infective endocarditis, their association with increased morbidity and mortality has been reported in previous studies [17, 42]. Supporting these observations, our findings underscore the critical importance of early recognition and management of complications to improve outcomes in patients with infective endocarditis. Our study did not find statistically significant difference in 28-day mortality between patients who underwent surgical treatment and those who did not, although mortality was relatively higher in non-surgical patients. Studies have shown that surgical treatment is a protective factor against mortality in IE patients [21, 32, 40]. The lack of statistical significance in our study may be attributed to the small sample size.
This study has several limitations. First, as a retrospective study, it may have led to the omission of some clinical variables. Second, it was conducted at a single tertiary care center where most patients had been referred from other hospitals for advanced diagnostic evaluation and treatment, often after prior antibiotic use. This may have contributed to the high rate of negative blood cultures observed in our study. Third, although the overall sample size appeared adequate at the cohort level, it was relatively small for subgroup comparisons such as between CAIE and HAIE, or among microbiological subtypes. Notably, the number of patients who experienced 28-day mortality was limited to only 15, which substantially reduced the statistical power of our multivariate analyses. This may have prevented potentially meaningful associations from reaching statistical significance, and resulted in wide confidence intervals that were clinically uninformative. Therefore, although no statistically significant associations were observed in the multivariate analysis, several variables were found to be significantly associated with 28-day mortality in univariate analysis. These findings should not be interpreted as conclusive but may still be considered clinically relevant and exploratory in nature. In this context, our results may help generate hypotheses and provide a basis for future large-scale studies with more robust methodological designs aimed at better understanding mortality-related factors in infective endocarditis.