Refine the approach--don't abandon the principle

Editor,--The problem with interpreting both the Family Heart Study Group's report of its one year trial1 and the result of the less interventionist OXCHECK study2 is that both studies tripped at the second hurdle for any clinical trial: after randomisation they failed to “screen” their controls in exactly the same way as they screened the intervention group. The Family Heart Study Group devotes much discussion to the ifs and buts of what the control results might have been and halve the observed reduction in coronary risk score to compensate. In the OXCHECK study, the figure seems to indicate, …