The main thing is what Ehrlich would have said was the 'magic bullet' idea. You want as much action on the bug and as little on the host. That's obvious, isn't it? But you've got to balance that as an equation, maybe subconsciously with every patient really. Are they allergic to the drug? Have they got kidney failure? What's their age? What's their size? The, the stuff which we do kind of semi-automatically, but maybe not quite consciously enough in many cases. And then, when you know the bug, relatively easy you know which drugs or you... can look them up in a book, which drugs are appropriate, then you put the rest of the stuff together and say, do that in that dose. I think that all sounds so incredibly elementary, but most of the terrible mistakes I've seen have been because people don't do that quite at that level. They say, 'I want a drug for septicaemia or for pneumonia'. There's no such thing. I mean, antibiotics don't resolve inflammatory exulation of the lung. They don't stop the inflammatory cascade, which is killing somebody with septicaemia. All they do, not quite all, but for practical purposes all they do, is to inhibit or kill bugs and unless you go to the actual nub of the thing to say which bugs do I want to kill because I know or I think it's this one, but it could be that one and, then again, it could be that one, it doesn't matter because the patient will get better anyway, but it could be that one, but although it's only 1% it will kill him. So, again, there's that equation of the importance of the possible pathogens. I mean, it sounds so incredibly banal, but I really think that most mistakes are because people don't go through... go through that elementary ground of saying which bug does the patient... has the patient got an infection? What's the cause of it? Do I know? Do I think? Do I... how should I go about that thing? And then what's this equation between the drugs and the bug? I suppose actually the truth is we should be doing this with every, every medicine, shouldn't we, but we don't all the time.

[Q] Yes, yes, yes.

And I think a big part of my life was trying to teach people about that.

[Q] It's almost about you were trying to teach a... process of clinical reasoning. Is that, is that...

Yes, actually, that's right. I when I started reading a bit, which I didn't really understand, about, you know, rational diagnosis and decision theory of medicine, which was a bit beyond me, but I realised in a dim sort of way that actually that's what you're doing. It's decision theory, and I don't mean intuitively in the sense that it's by some mystical, magical insight which I got. What I mean is that you're subconsciously or consciously balancing up these factors as well as you can, in the light of what you know, exactly that.

[Q] Yes, yes. And, and do you feel that we... that, that you've been able to get that message across, or do you think it's, it's slipping away from us?

Well, no actually. No, no, I think the mistakes are just as crass as they always were, but I probably got it across to some people some of the time, including myself. No, actually, I, I don't know really. There's all the pharmacokinetics of the drug. I mean, without joking about it, we still do see terrible things about using drugs, which people don't need. I used to have a slide when I was teaching to... at the London School of Hygiene, and the end of it said, about antibiotics, 'Most people who get them don't need them, and most people who need them don't get them', which is, I'm afraid, the case in the world. But, no, no I wouldn't say it was completely futile. No, no, I wouldn't say that, but I'd say it was still a pretty hard thing to follow.

Probably the biggest influence has been the influence mainly in general practise to say surely not everyone needs drugs all the time, and the fact that that's spread to patients with their more anti-doctor, anti-medicine, anti this horrible poison because in my early years of this specialty I would give lectures to, forgive me, GPs. I'd make some very modest remark about maybe you could be a bit more restrictive about, in this particular situation, this mild sore throat and they, they practically attacked me. They said, 'What? Not give an antibiotic?' And practically drive me out of the room. Years later it was the other way around, and patients, too, not quite so... but let's face it, most people want some drug, don't they, when they come to see us.