Response to Letter by Kaplan Regarding Article, “Bacteremia Associated With Tooth Brushing and Dental Extraction”
We appreciate Dr Kaplan’s comments and concerns. With regard to the nonsignificant but higher percentage of complex extractions in the placebo group, we devote a paragraph to this limitation in the Discussion section.1 We point out that the real significance of this outcome is that “… there is less of an impact from amoxicillin than Figure 3 suggests.”
With regard to culturing the tooth socket, we did acquire plaque samples from the deepest periodontal pocket from subjects in the extraction arms. These specimens are currently frozen for later analysis and matching with the blood culture results, and we agree that these samples may provide interesting results. Cultures of tooth sockets, however, would likely have limited value and would have added to the overall cost of the study. For example, the socket would not provide a good indication of the bacteria forced into the bloodstream by an invasive procedure, which would most likely have come from the tooth surface and gingival crevice. We were also concerned that the fresh extraction socket would be contaminated with bacteria from other areas of the oral cavity immediately after the tooth was removed.
On the issue of susceptibility testing, we agree that it would have provided interesting data in light of increasing antibiotic resistance in general. This was not a goal of this study, largely for reasons of cost, and we felt that sufficient evidence of an increasing resistance to oral streptococci was already available. This is another reason why antibiotic prophylaxis may be of limited value in the prevention of infective endocarditis.
With regard to the issue of 2 doses of antibiotic, we are quite familiar with the pivotal research by investigators in the 1970s and early 1980s that focused on the pathobiology of infective endocarditis. The McGowan animal study2 involved artificially induced cardiac valve lesions and massive injections of streptococci, which bears little relevance to the clinical arena today. We chose a methodology that reflected the American Heart Association recommendations for antibiotic prophylaxis. We agree that it would be interesting to know the impact of a second dose of antibiotic, but this is outweighed by Institutional Review Board considerations and the increased number of subjects (and cost) necessary for statistical significance.
Finally, with regard to the issue of antibiotic blood levels, we agree that such data would be interesting. Our last blood draw was at 60 minutes (not “several hours”), and no sufficiently large studies have looked at time periods longer than this. Although it makes sense that the antibiotic blood level and duration would be higher with 2 doses of amoxicillin, we do not know whether 2 doses would have an impact on our 4 surrogate measures of risk for infective endocarditis (incidence, duration, nature, and magnitude). This was not among the goals of this study, but a future study could address this question.