Discussion

This retrospective observational study describes the clinical, microbiological, and echocardiographic characteristics of 133 unique confirmed clinical episodes IE managed at a UK teaching DGH. The principal finding was a significant association between documented invasive dental procedures and oral-flora IE, with patients with documented recent dental exposure demonstrating approximately 2.5-fold greater odds of infection with oral microorganisms. In addition, Staphylococcus aureus remained the predominant pathogen, native-valve IE accounted for most cases, the aortic valve was the most frequently affected cardiac structure, and overall in-hospital mortality was 11.3%. These findings provide contemporary data describing the epidemiology of IE in routine clinical practice while highlighting the importance of documenting dental history and oral health in patients presenting with suspected IE [4-6].

The observed association between documented dental exposure and oral-flora IE is biologically plausible and consistent with the recognised pathogenesis of viridans streptococcal endocarditis. Manipulation of gingival tissues during invasive dental procedures may result in transient bacteraemia, particularly in individuals with poor oral hygiene or underlying structural heart disease. However, transient bacteraemia may also occur during routine daily activities such as tooth brushing and chewing, reinforcing the importance of maintaining good oral health in patients at increased risk of IE. Current European Society of Cardiology (ESC) and American Heart Association (AHA) guidelines recommend antibiotic prophylaxis only for carefully selected high-risk patients undergoing invasive dental procedures, whereas National Institute for Health and Care Excellence (NICE) guidance does not recommend routine antibiotic prophylaxis for dental procedures. All three place greater emphasis on preventive dental care and oral hygiene than on widespread antibiotic use [1-3,12-15].

Although our findings demonstrate a statistically significant association between documented dental procedures and oral-flora IE, they should not be interpreted as evidence of causality. Approximately half of the patients with documented dental procedures had IE caused by non-oral organisms, while many patients with oral-flora IE had no recorded recent dental procedure. These findings suggest that incomplete documentation, unrecognised dental disease, spontaneous bacteraemia arising from routine daily activities, or alternative portals of microbial entry may all contribute to the development of IE. Similar observations have been reported in previous studies evaluating the relationship between dental procedures and IE [12-15].

Among the 26 patients with documented dental procedures, prophylaxis was documented in only two patients (7.7%), while 22 patients (84.6%) had no recorded information regarding antibiotic prophylaxis. This may reflect incomplete documentation rather than confirmed non-adherence to guideline recommendations, as this retrospective study cannot determine whether prophylaxis was administered but not recorded or was not administered. Nevertheless, this represents an important quality improvement opportunity. Introducing a standardised IE admission pro forma incorporating mandatory documentation of recent dental procedures, oral health status, and antibiotic prophylaxis may improve adherence to guideline-based practice and facilitate future clinical audits [1-3,14,15].

The microbiological findings observed in this study are consistent with those reported in contemporary international cohorts. Staphylococcus aureus remained the most frequently isolated organism, reflecting the increasing contribution of healthcare-associated infection and invasive medical procedures to the epidemiology of IE. Oral-flora organisms accounted for approximately one-third of cases, while enterococci also represented an important proportion of infections, emphasising the heterogeneous microbiological profile of modern IE. These findings are broadly consistent with reports from the International Collaboration on Endocarditis and other contemporary series [4-6,10,11].

Our echocardiographic findings similarly reflect current epidemiological trends. Native-valve IE predominated, with the aortic valve representing the most frequently involved cardiac structure. Device-related IE accounted for a substantial proportion of cases, illustrating the growing burden of infections involving cardiac implantable electronic devices. Approximately one in five patients required surgical intervention or device extraction, emphasising the complexity of managing IE and the importance of multidisciplinary collaboration involving cardiology, cardiothoracic surgery, microbiology, infectious diseases, and cardiac imaging specialists [1,4,5,16-18].

Overall in-hospital mortality was 11.3%, which is lower than that reported in many contemporary registries. This may reflect the use of a strictly defined in-hospital mortality endpoint rather than longer-term follow-up, as well as differences in patient populations and referral pathways. Although prosthetic valve IE had the highest crude mortality, no demographic, microbiological, or echocardiographic variable was significantly associated with in-hospital mortality in this cohort. The relatively small number of mortality events limited the statistical power for multivariable modelling and should be considered when interpreting these findings [4-6,19-21].

This study has several strengths. It includes a contemporary cohort of consecutive confirmed IE episodes identified among patients undergoing TOE for suspected IE at a UK teaching DGH and provides comprehensive clinical, microbiological, echocardiographic, and outcome data. Furthermore, to our knowledge, it is among the few UK studies to specifically evaluate the relationship between documented dental exposure and oral-flora IE while also assessing antibiotic prophylaxis documentation. These findings have direct relevance to clinical practice and quality improvement initiatives.

Several limitations should be acknowledged. First, this was a retrospective single-centre study of patients undergoing TOE for suspected IE, which may introduce selection bias and limit generalisability, as patients who did not undergo TOE, including those who were too unwell for the procedure or were diagnosed through other pathways, were not captured. Second, documentation of dental procedures, their timing and indication, and antibiotic prophylaxis was incomplete, introducing potential information bias. The six-month exposure window was selected to capture recent documented dental exposure, but the broad definition and limited documentation of the timing and nature of individual procedures may have resulted in exposure misclassification. Reverse causation is also possible, whereby patients with undiagnosed IE may develop dental or oral symptoms and seek dental care before the diagnosis of IE is established. Third, the relatively small number of oral-flora-associated IE cases limited the ability to perform reliable multivariable adjustment, and the confidence interval around the primary association was relatively wide (OR, 2.47; 95% CI, 1.03-5.94), indicating uncertainty regarding the precision of the estimate. Fourth, although prophylaxis was documented in only a small proportion of patients, the retrospective design does not allow distinction between prophylaxis that was not administered and prophylaxis that was administered but not documented. Finally, the single-centre DGH setting may limit generalisability to tertiary referral centres and other healthcare settings. Despite these limitations, the study provides contemporary observational data on documented dental exposure and oral-flora-associated IE in a UK teaching DGH [1,4,5,20,21].