ABC of alcohol

for use in dentists' surgeries because we know that they are not given. Amoxycillin is known to provide the required bactericidal activity, and proof of efficacy in the rabbit model is not particularly relevant. The logic of unselective prophylaxis relates to the 60">, or so of patients with viridans endocarditis who had previously been unaware of heart disease. Selective prophylaxis, even if perfectly protective, could not do more than make a small dent in the incidence since the majority of cases would not have been given any protection. If prophylaxis is relevant and works, more people should receive it and provided the regimen is simple they probably will. C M OAKLEY J H DARRELL