ABSTRACT
The author feels that the Royal College of General Practitioners requires general practitioners to be able to identify the questions arising from a consultation, then to find and appraise the relevant scientific literature. The positive predictive value of a test is the prevalence of the disease in patients who test positive. The negative predictive value is the prevalence in those who test negative. The author suggests replacing the stumbling blocks with two building blocks: odds and likelihood ratios. Odds and likelihood ratios do three things: first, they give one an idea how useful or useless a test is; second, they help one make a diagnosis; and third, they indicate how certain or uncertain the diagnosis is. Clinicians need know nothing about where this ratio comes from: they just need to know how to use it. Once the clinician is sure about these two building blocks, and how they fit together, they will have a firm foundation on which to build.
