Antibiotic prophylaxis in dentistry for the prevention of infective endocarditis

Abstract

In many parts of the world, it is praxis to administer prophylactic antibiotics prior to invasive dental procedures to reduce the risk of postoperative infective endocarditis (IE) among at-risk individuals. Such endocarditis prophylaxis (EP) is based on tradition rather than evidence, and although it was proposed by the American Heart Association in 1955, a Cochrane review published in 2021 concluded that there is not enough evidence to prove its efficacy in reducing IE. Randomized clinical trials have not been conducted, and would be resource- heavy due to the rarity of IE.

Since October 2012, EP is no longer recommended in Swedish dentistry. This change created an opportunity to study the incidence of IE at times when EP was and was not recommended. Considering the lack of evidence, such epidemiological studies provide a pragmatic approach to finding evidence for future guidelines.

In the studies included in this thesis, nationwide Swedish register data was used to assess whether IE increased after the change, in Sweden, and particularly among the at-risk individuals affected by this change. Linkage of multiple registers enabled prospectively recorded follow-up of high standard by international measures, and consideration of several background characteristics.

Study I found no elevated incidence of IE among the general population during the five years following the change in recommendations compared to before. National data indicate that dentists’ prescriptions of amoxicillin decreased by 41% during the same period. The study was ecological and conducted on a national scale, and the Swedish national quality register for IE provided a more specific way of measuring IE than did studies from other countries where such a register was not available.

Study II focused on the risk groups previously covered by EP recommendations. Using a cohort design, study persons were identified using national health registers from 1964 onward. At-risk individuals living in Sweden were followed between 2008 and 2018, and incidence rates of IE were compared to a cohort without known risk factors. The study found no increase of IE after October 2012 among those previously covered by the recommendations. The direction of the results was unchanged after adjusting for educational attainment, age, and sex.

Study III found that cases of IE among at-risk individuals were rarely associated with dental procedures. Instead, >95% of cases were more likely associated with daily bacteremia, caused, for example, by mastication or tooth brushing. Moreover, dental procedures were not more common in the months before IE than during reference periods. A nested case- crossover design was used to account for intra-individual confounding factors that did not change over time. For completeness, a traditional nested case-control design was also employed. Both designs showed similar results. The study was conducted by linking data from study II to the Swedish national dental health register, and enabled the calculation of the amount of antibiotics that could be saved by recommending the cessation of EP in Sweden.

In conclusion, the studies in this thesis find no evidence that oral streptococcal IE increased in Sweden or among high-risk individuals after the recommended cessation of EP to these individuals. The results indicate that EP may do little to reduce IE among high-risk individuals.