Sir, The use of antibiotics to prevent infective endocarditis (IE) has generated extensive debate between dentists, microbiolo- gists, cardiologists and cardiothoracic surgeons. Up to 16 million of the UK population 1 may potentially be required to take antibiotic prophylaxis based on past guidelines produced by various specialist societies. In addition, several differing recommendations exist which perpetuates confusion among clinicians who are dealing directly with those at risk. However, more recently, national and international guidelines have recommended significant changes in the use of antibiotic prophylaxis for the prevention of IE which may facilitate rationalization. 2 patients at risk of IE 3 and recommended that antibiotic prophy- laxis should not be given to any patients at risk of IE undergoing dental, upper and lower gastrointestinal (GI) tract, genitourinary (GU) or respiratory tract procedures. NICE concluded that for dental procedures 'the evidence does not show a causal relation- ship between having an interventional procedure and the development of IE' and 'it is biologically implausible that a dental procedure would lead to a greater risk of IE than regular tooth brushing'. This was based on increasing evidence of cumulative bacteraemia from every day oral activities. 4 Included was a de novo UK relevant analysis of cost-effectiveness of those at risk of IE undergoing dental procedures. The economic evaluation was based upon a Markov model and replicated a pre- viously published American analysis but applied costs from the perspective of the NHS. The outcome, based on prophylactic options set out in BNF 54, suggested that even applying the most optimistic assumptions with regard to efficacy and anaphylaxis risk, a strategy of no prophylaxis, leads to fewer deaths, especially when using a penicillin-based regimen. The NICE guideline development group reasoned that pro- cedures involving the GI or GU tract are being increasingly undertaken in the UK with only a small number of associated IE cases being reported; therefore, logically there can be no causal relationship with IE. In contrast to dental procedures, there is a lack of data regarding non-dental intervention and subsequent IE risk, and no published studies show a conclusive link between procedures of the GU and GI tract and the development of IE. Consequently, NICE did not recommend antibiotic prophylaxis for any patient at risk of IE undergoing GU or GI procedures in a non-infected site. NICE defined the patients who should receive antibiotic pro- phylaxis to prevent IE as those undergoing GI or GU procedures at 'infected' or 'potentially infected' sites. A potentially infected site was not clearly defined and is therefore open to clinical interpretation. Specific indications could include endoscopic retrograde chologio-pancreaticography in the presence of biliary disorders or urethral instrumentation associated with urinary sepsis. These details need to be clarified by specialist advisory groups to prevent misinterpretation. Similarly, NICE does not give specific recommendation regarding those patients who are already being treated with anti- biotics for ongoing infection and who are due to have a pro- cedure local to that site, for example, a dental abscess or cholecystitis, or those who are already on long-term antibiotic prophylaxis for other reasons including asplenia and recurrent urinary tract infections. NICE recommended, based on the expert opinion of the guideline development group, that anti- biotics should be given to cover microorganisms which have been known to cause IE under these circumstances; again specific recommendations for these patients need to be produced by the relevant advisory groups. In view of the publication from NICE regarding antibiotic prophylaxis in IE, the BSAC Working Party considered that the current BSAC guidelines should be updated to bring them in line with the recommendations from NICE and also the BNF. We considered that the guidance from NICE should be accepted and endorsed by all interested UK groups so that a single national guideline can be achieved. The BSAC Working Party are now drawing up specific recommendations for those patients who are infected or potentially infected at the operative site and who are at risk of IE.