This is my go to book, whenever I need to know if a test is good or bad. McGraw has made what must be a huge effort to collects the evidence on clinical examinations and thereby challenging the tradionalistic way of thinking.
I still think actual calculation in the thresholdmodel is not yet (if ever) fit for primetime, as studies are lacking the information below AND because of the insecurity of this model, the number we get at the end will always be with CI95% (sort of speak).
What could be better is: - Consistent mentioning of population and clinical spectrum that the LR is taken from - Mentioning of gaps of knowledge - Including more paraclinical LR's (bloodtests, scans etc) - Including more CDR's - Include CI 95% as this is essential to how much we believe in this particular LR
In summary: essential book for anyone intrested in bayeresian thinking and general decision making in medicine
Want to know the difference between a good clinician and a great clinician... this book. During clinical training you will learn how to do various things and think that you are devising a waterproof diagnosis. Well, think again. By applying likelihood ratios and probability to your examination, you get to refine your differentials with even more (or less) precision.