I was most impressed with the paper by Astin et al on the diagnostic value of symptoms for colorectal cancer in primary care.1 This is such important research, as each day we see patients with gastrointestinal symptoms and I find myself always concerned about missing the bowel cancer (where the prognosis is excellent when found early). My comment to Astin et al is that their abstract is confusing. Change in bowel habit (with rectal bleeding?) has a positive likelihood ratio but the bottom line of the results says the positive likelihood ration (PLR) is 1 or less for diarrhoea or constipation (change in bowel habit). I am not sure which ‘advice to follow’ and I wonder how many of my colleagues are fully conversant with likelihood ratios, and for the sake of good communication, perhaps these should be translated in to text. Otherwise, well done to their team.
- © British Journal of General Practice, January 2011