Risk of Infective Endocarditis Due to Invasive Dental Procedures

A NICE Conclusion

Larry M. Baddour, MD; Bernard D. Prendergast, DM


Circulation. 2018;138(4):364-366. 

Who would argue against the prevention of a life-threatening infection? It can be achieved by a variety of methods, including vaccines, antimicrobials, and infection prevention and control practices. But what if the preventive measures had been in place for >6 decades and never been of proven efficacy because of the lack of randomized, placebo-controlled trials? Such is the case for antibiotic prophylaxis in the prevention of infective endocarditis (IE). In 1955, giants in the field of streptococcal infections advocated the use of penicillin in patients undergoing dental manipulations or operative procedures involving the oral cavity in the initial version of the American Heart Association guidelines on prevention of IE.[1] As 1 (oral) treatment option, penicillin was recommended to begin 24 hours before the procedure and extend for 5 days thereafter.

Since then, the number of editorials, letters to the editor, and case reports alleging an association between invasive dental procedures and IE far outnumber the relatively scant number of publications that include more advanced methodological evaluation of the possible association. Moreover, no placebo-controlled, randomized trials have been conducted to further evaluate the perennial questions as to whether invasive dental procedures have anything to do with the risk of IE and whether an intervention (in this case, antibiotic administration before the procedure) can ultimately prevent an infection whose morbidity and mortality rival that of some cancers.

Fast forward to the early 21st century, when guidelines committees and governing bodies took notice of several different investigations that included relatively large databases and, with a renewed interest in promoting antimicrobial stewardship (against a backdrop of alarming increases in drug resistance among a panoply of organisms), restricted (or totally eliminated) the practice of antibiotic administration to patients undergoing invasive dental procedures deemed at risk for the development of IE. By restricting the use of antibiotic prophylaxis to only high-risk individuals undergoing dental procedures as per the tenth and most recent iteration of the American Heart Association guidelines published in 2007;[2] for example, the expectation was that antibiotic prophylaxis use would plummet by ≈90% in the United States without precipitating an IE epidemic attributable to viridans group streptococci. The impact of these changing guidelines on the incidence of IE has been accessed in numerous before-and-after studies, which have provided mixed results, even when the same database was examined by different groups of investigators as 1 example.[3] The consistent Achilles' heel that has characterized these investigations (and that requires urgent resolution) is that no International Classification of Diseases code exists for viridans group streptococci and that coding for enterococcal infection is still grouped within the genus Streptococcus.

In the current issue of Circulation, Chen and colleagues[4] provide a risk analysis of IE complicating invasive dental procedures by using a Taiwanese database. On initial review, the declaration that "antibiotic prophylaxis for prevention of IE is not required for the Taiwanese population" indicates that this investigation is unique in its ability to clarify 2 age-old conundrums: (1) whether invasive dental procedures predispose to the subsequent development of IE and (2) whether administration of preprocedural antibiotics reduces this risk. Some may consider this a bold position given that others have only suggested a compromise by reserving antibiotic prophylaxis for those at highest risk of IE. In the absence of definitive evidence, the National Institute for Health and Care Excellence recommended in 2008 that the use of antibiotic prophylaxis should cease in the United Kingdom; antibiotic prescribing before invasive dental procedures plummeted and the incidence of IE increased in England.[5] However, microbiological data were unavailable to define responsible pathogens and whether there was a possible link to invasive dental procedures.

The use by Chen et al of 2 novel case-only designs, a case-crossover method and a self-controlled case series, is unique in the assessment of the dental connection. These designs were selected because confounding factors do not vary over time and are adjusted for implicitness. Patient-related data for these evaluations were obtained from the Health Insurance Database in Taiwan, which has 99.9% of Taiwan's population enrolled, and the International Classification of Diseases, Ninth Revision, Clinical Modification diagnosis codes were used along with other clinical information. All patients who have IE (International Classification of Diseases, Ninth Revision, Clinical Modification code 421) between 20 and 100 years of age were included during the study period with exposure to specific invasive dental procedures over a period of up to 16 weeks. Only the first IE episode per patient was included in analyses. Approximately 27 million individuals were enrolled in the national health insurance program by the last year (2013) of the study period with >16 000 patients who had IE identified. Over 8000 patients were included in each of the 2 study designs, and no increased IE risk in the ≤16-week periods following invasive dental procedures was observed. Moreover, no association was seen for a subset of patients at increased IE risk (rheumatic heart disease or prosthetic valve replacement/valvuloplasty).

Several other case-controlled investigations with varying methodology have been published over the past ≈30 years. Consistent with prior case-control analyses, one of these investigations[6] examined 1 million randomly selected members of the same Taiwanese database as used in the current Circulation article by using a case-crossover design and demonstrated no association between dental procedures and risk of IE.

Where do we go from here? A large, multicenter, randomized, placebo-controlled trial with several thousand patients should be conducted to address the questions of whether invasive dental procedures result in IE and whether antibiotic prophylaxis prevents it. Patients with a history of IE and prosthetic valves would be the highest-risk populations for enrollment. As things stand, the trial would have to be conducted in the United Kingdom, where National Institute for Health and Care Excellence guidelines are in place, because of ethical concerns about the inclusion of a placebo-controlled trial elsewhere. Unfortunately, funding for such a trial is almost prohibitive and, even if available, would likely lead to more questions regarding the nuances of host factors and types of invasive dental procedures. Guidelines committees should therefore evaluate the need for antibiotic prophylaxis before dental procedures in patients at risk of IE by using studies that involve other methodologies; in that sense, the present study by Chen et al[4] is an important contribution.