a few weeks back somebody asked about how to get NNTs from trials or meta-analyses reporting continuous data such as BP or depression scores. hope this isn't too late but i came across this reference to a letter: Furukawa TA (Lancet 1999; 353: 1680-1) Furukawa gives a table for converting Effect Sizes into NNTs.(The effect size is one way of presenting differences between two samples; it equals the difference between the means of the two samples divided by the pooled standard deviation. It has been spotted in meta-analyses.) I am not aware of this table being used. anyone any experience of it? -- Kev Hopayian, Seahills, Leiston Rd, Aldeburgh, Suffolk IP15 5PL, England %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%