Accessible online at: www.karger.com/journals/eur Floor: Dr Rischmann recommends urine culture if fever persists following BCG installation. Would not blood culture for mycobacteria also be appropriate if you are concerned about systemic BCG septicaemia? P. Rischmann: If fever and side-effects such as cystitis are moderate, then you should proceed with urine culture. While waiting for the results, fluoroquinolones and paracetamol can be given. In the case of severe systemic effects, one should not wait for bacteriological results before beginning treatment. Blood cultures are seldom positive; reverse-transcriptase-PCR may be more useful. A. Bohle: From our own experience, and from the experience in the literature, blood cultures are not very successful at revealing mycobacterial infection. Blood cultures are often negative even in the presence of systemic BCG infection. Floor: Is any of the various BCG strains superior? Furthermore, when referring to reductions in the dose of BCG, does it make more sense to talk about colony-forming units (CFUs) rather than milligrams? Equal doses (in milligrams) of different strains of BCG may have different numbers of CFUs. P.F. Bassi: Which parameters we use to measure BCG is an important topic, but one that is still largely unexplored. As a matter of fact, we do not know what the active component of the BCG is or how BCG works. As a consequence, we do not know what to measure: weight is only a manufacturer’s product information, which does not express the absolute therapeutic ability of the preparation. It is, therefore, not possible to make a comparison by weight between, for example, 150 mg of Pasteur BCG, 81 mg of Connaught BCG or 50 mg of OncoTice. Consequently, it is impossible to say that 150 mg Pasteur BCG is threefold ‘better’ than 50 mg Tice strain. Also the number of viable colonies (CFUs) has been supposed, in the past, to make a difference in terms of efficacy, but in the long run, this highly imprecise parameter has failed to be a reliable indicator. D.L. Lamm: The variability not only between strain and strain, but also between one ampoule and another ampoule of the same preparation is very high. For this reason, Mike O’Donnell’s recommendation for dose reduction makes a lot of sense, and in our experience is very useful. When patients start to have increased side-effects we reduce the dose to one third; if side-effects persist or increase, the dose is reduced to one tenth, and even to one hundredth if necessary. Basically, the patient does not have to suffer at all to receive effective immune stimulation with BCG, and I think that dose reduction therefore really makes more sense than prophylactic isoniazid. With isoniazid we are adding another variable, and we don’t know what percentage kill isoniazid prophylaxis results in, unlike log reductions in BCG dose, which do give us a rough idea of the effective decrease in dose. Floor: Do you think there is any correlation between the pre-treatment tuberculin test and the subsequent development of side-effects?