Guidelines for the prevention of endocarditis: report of the Working Party of the British Society for Antimicrobial Chemotherapy--authors' response

Sir, Working Party guidelines were produced to update infective endocarditis (IE) prophylaxis guidelines based on current best evidence.1 Three major studies2–4 have failed to demonstrate a causative link between dental treatment and IE. As detailed in the guidelines, the risk of IE due to normal daily activities such as brushing of teeth or chewing is considerably greater than from a single dental procedure,5 therefore even if prophylaxis were assumed to be 100% effective the overall benefit is negligible. This would also apply to moderate risk cardiac lesions. Of course, some cases of IE will occur following dental procedures6 but they may not be directly related. Indeed the use of prophylaxis has not decreased the incidence of IE7 and there are continued reports of failed prophylaxis.8 We agree with Dr Shanson's9 observation that oral streptococcal IE can have considerable associated morbidity and mortality, but we consider that the most effective approach to minimize IE is through meticulous oral hygiene to decrease the challenge of daily bacteraemia. This approach is emphasized in the suggested information for patients included in Appendix 1 of our article.1 The studies referred to by Dr Shanson10 are on small numbers of patients. The opinions of the Working Party take a balanced viewpoint of the available case-control data, some of which is conflicting. The Working Party has also noted that a single IE case misclassification in both studies quoted by Dr Shanson would have reversed the authors' conclusions. The study by Horstkotte11 referred only to prosthetic valves. The Working Party does recommend that prophylaxis be continued for this patient group. In the Netherlands, van der Meer and colleagues estimated that theoretically 6% of native valve IE may be prevented by using prophylaxis for certain dental procedures. This figure represents the theoretical maximum of preventable cases. However the number of cases of IE actually prevented may be considerably smaller despite extensive use of prophylaxis.6 Compliance with prophylaxis guidelines also remains limited, with low numbers of patients who require prophylaxis actually taking or receiving antibiotics12 and high numbers who do not require antibiotics receiving needless antibiotic treatment.13 The Working Party aimed to simplify current guidelines to also improve compliance. The development of IE remains unchallenged with a bacteraemia required to initiate infection. Streptococcal IE cases following dental procedures reported in the literature are historical14 and are accepted but, as with all reported cases, prone to selection bias. Dental procedures are clearly known to induce bacteraemia. The observation in Roberts's paper15 that pre-extraction blood cultures are largely negative compared with post-extraction blood culture is neither disputed nor surprising. The probability of a single blood culture in an individual being positive when taken at random is small. However, lysis filtration techniques, with higher sensitivity, have demonstrated that up to 86% of blood samples are positive for bacteria pre-procedure.16 The observation that dental procedures induce bacteraemia is valid, but on a background of daily bacteraemia does not necessarily increase the probability that an individual will develop IE as a direct result of dental intervention. The emphasis has changed to cumulative bacteraemia over longer timescales which are unrelated to dental intervention and are therefore much more likely to cause IE.5 It has been estimated that >50% of the population over 60 years of age fulfil antibiotic prophylaxis guidelines as published previously.17 With an ageing population the number of patients requiring prophylaxis in the UK alone is likely to be in the order of 16.5 million (Census 2001). The overall effect this has on microbial resistance is unknown. However with increasing numbers of resistant organisms both in community and hospital settings, more prudent use of antibiotics needs to be put in place as recommended by the House of Lords Select Committee on Science and Technology.18 In addition in a recent cost-effectiveness analysis on antibiotic prophylaxis in the USA it was found that the use of amoxicillin or ampicillin resulted in an overall net loss of life, based on a fatal anaphylaxis rate of 20 per million.19 Thus the safety of these antibiotics in this situation needs to be balanced against benefit. The cost-effectiveness of antibiotic prophylaxis for moderate cardiac lesions was also unfavourable in health economic terms despite the relative low cost of amoxicillin. The Working Party believes that there is sufficient doubt regarding dental prophylaxis that a prospective randomized trial to address the issue would be ethically acceptable. Unfortunately the necessary design and running of such a trial would be very difficult due to the multitude of different dental procedures and cardiac risk factors that would need to be considered. It has been estimated that over 6000 patients with cardiac disease would need to be recruited to provide sufficient statistical power to address this issue.6 The International Collaboration on Endocarditis, which has been recruiting since 1998 on a worldwide basis, has only recruited ∼2500 patients.20 This highlights the difficulty in recruiting in IE studies despite substantial efforts from leading IE researchers here and abroad. In response to the comments of Dr Gibbs and colleagues,21 as outlined above there is minimal evidence to suggest that dental prophylaxis prevents IE. The Working Party indeed concluded that dental prophylaxis could be abandoned for all cardiac risk factors based on the available evidence, but chose to adopt a stepwise approach to implementing this in the first instance. We are pleased that the British Congenital Cardiac Association and British Cardiovascular Society have considered our guidelines and we welcome their comments. The guidelines were subject to a formal process of national consultation including being published on the BSAC website and being issued to national organizations/bodies for comment, which the Working Party acknowledged. These views have been echoed in a recent Editorial.22 We accept that the use of prophylaxis for the prevention of IE is an emotive subject but believe that our guidelines rationalize the use of antibiotics making compliance more likely and minimizing the risk of antibiotic resistance emerging. These views have been echoed in a recent Editorial.22 We understand that the National Institute for Clinical Excellence (NICE) is also considering a review of antibiotic prophylaxis in IE and we look forward to contributing to this process. All the authors declare that during the preparation of this document they were not in the employment of, or receiving funding from, any pharmaceutical firm or other organization that may have resulted in a conflict of interest.