Sir, neither the British Congenital Cardiac Association (BCCA) nor the British Cardiovascular Society (BCS) share your editorial author's1 enthusiasm for the new guidelines for the prevention of endocarditis published by the British Society for Antimicrobial Chemotherapy (BSAC).2 In our view the changes to the guidelines are neither based upon science nor common sense and we do not commend these proposed changes to our members or the public.

If there was sound evidence to suggest that the risk of anaphylaxis outweighs the possible benefit of antibiotic prophylaxis, then the correct recommendation would be to abandon antibiotic prophylaxis. The BSAC do not recommend that, so logic suggests that they must be of the opinion that invasive procedures known to induce bacteraemia do indeed pose a risk for certain patients. To recommend prophylaxis only in patients with previous endocarditis, prosthetic valves or implanted conduits, excluding those with high velocity intracardiac jets known to be at high risk of endocarditis (such as mitral regurgitation or ventricular septal defect) defies logical explanation. To go on to recommend prophylaxis in a very restricted group of patients even for dental procedures which do not involve gingival damage and are therefore unlikely to induce significant bacteraemia similarly defies logic.

We recognise that current guidelines from the UK, Europe and North America are based upon broad consensus rather than hard evidence. To change these recommendations on the basis of views of one small group (the BSAC) rather than science is likely to simply repeat the mistakes of the past. The BCCA and the BCS wrote to the BSAC setting out the reasons for our disagreements with their new recommendations well in advance of publication but our views were dismissed. To publish new national guidelines which fail to take into account the consensus of the UK cardiologists' national professional body seems most unwise and is likely to cause much confusion in clinical and medicolegal practice.

We are pleased that the National Institute for Clinical Excellence (NICE) has agreed to report on prevention of endocarditis. We strongly recommend that no change in current practice should take place until NICE have published on the matter.

Dr Mike Martin responds to the above two letters: I would like to thank Drs Gibbs, Ramsdale and colleagues for their replies to my editorial. 3

There is one fact that all parties can agree upon and that is infective endocarditis (IE) is a devastating disease for those who contract it. It certainly does wreck lives and can cause serious problems for the carers of people who have the disease. The gravity of this disease is therefore not lost on anyone who writes, or makes any kind of recommendations about antibiotic prophylaxis.

I am sure that the British Society of Antimicrobial Prophylaxis (BSAC) working party on IE have examined all the evidence available to them, including that presented by Ramsdale, Gibbs and colleagues.4 All branches of medicine move forward based on evidence and unfortunately I see no new evidence presented by your correspondents. Well respected clinicians in this field have thoroughly reviewed the clinical and scientific evidence and have consistently provided evidence that the link between dentistry and IE is tenuous.5,6 Even in the civil courts, where the standard of proof is on 'the balance of probabilities' (ie 51%), it is very difficult to prove an association between dental treatment and IE.7 In addition, very many patients who are unaware of their predisposition to IE receive dental treatment every day and do not contract this disease; thankfully it is very rare.

The evidence that a single dose of amoxicillin works prophylactically is also problematical. A Cochrane review of the literature also failed to demonstrate any efficacy.8 The original paper on antimicrobial prophylaxis and IE by Northrop and Crowley reduced dentally-induced bacteraemias from 12.8% to 4%, however the methodology and the statistical tests used in this paper are questionable and the conclusions not justified. Bacteraemias generated from dentistry probably last for circa 20 minutes. Amoxicillin primarily works on dividing bacteria, a situation that does not apply to bacteraemias. The evidence that has accumulated from animal models of infective endocarditis show that initially bacteria present on the heart divide slowly with mean generation times of 12-24 hours, when there would be little amoxicillin present in the blood stream. Thus the mode of action of amoxicillin could be to hinder or prevent attachment of bacteria to platelets or fibrin, or perhaps to aid phagocytosis of the bacteria. Even in animal models prophylaxis does not completely prevent IE and there are well documented failures of amoxicillin prophylaxis in patients.9 Thus there is considerable room for doubt that a single dose of prophylactic antibiotic works reliably in preventing IE.

If the advisory bodies quoted by your correspondents wish to challenge the BSAC working party on IE they must do so on the basis of scientific evidence. The BSAC working party did consider completely abolishing antibiotic prophylaxis for dental procedures, I personally wish they had taken this step.4 I am pleased to hear that NICE is to examine the relationship between antibiotic prophylaxis and dentistry, I am confident that they will agree with the BSAC recommendations.

Perhaps there are active steps that cardiologists and dentists could take to negate any risk of IE from dental operations. Cardiologists could advise persons at risk of IE to get dentally fit and thus prevent constant low grade bacteraemias. 10 Dentists could also reduce bacteraemias by judicial use of topically applied antiseptics which are known to reduce bacteraemias. 11 These steps would prevent the high use of antibiotics in dentistry, where almost one third of all prescriptions are for prophylactic purposes. 12