Tuesday, 28 August 2012

I hate that guy!

Well, I don't literally hate that guy; and I haven't thought about this in a while, so things are probably changed. But a couple of years back, after serious consideration, I had decided that John Cusack was, by far, my least favourite actor.

He had a string of movies which consistently sucked (of course this is just my very subjective opinion $-$ but seriously: they seriously...) and I couldn't really stand the guy.

I thought of this when I was finishing writing a new grant proposal (which I've submitted last Friday). Among other jems, James, one of the co-applicants, added in the text the word "serendipitously", which I thought was just amazing $-$ despite the sad memories it brought back of John Cusack.

Anyway, I really like James: he is a clinician and and enthusiast supporter of Bayesian methods, even in his applied research, which is very rare! But more importantly, he is one of those people with an incredible culture, which transpires in the way he writes, including the fancy, mostly unheard of, words. Other examples include: "dirigiste" and "egregious" (as in "the cost can be egregious").

The proposal is titled "Interpreting lower urinary tract symptoms using Bayesian methods: improved diagnostic inference, faster delivery, less investigation and lower costs". I think the idea is really cool: briefly, lower urinary tract symptoms (LUTS) are clinically a big deal because they may be indicative of serious pathologies of the urogenital tract, including cancers. But they often go mis-diagnosed, often because doctors look for presence/absence of signals in diagnostic tests, which are quite noisy.

Our proposal, on the other hand, is about using data and expert opinions to estimate the underlying probability of the potential causes, given the symptoms. Some preliminary evidence seems to suggest that this could improve effectiveness and even save money in the long run; thus we'll also include a cost-effectiveness component. The idea is to first validate the system in James' hospital and then trial it on a few NHS hospitals and, possibly, in a couple of different areas.