There has been considerable interest in recommending that pregnant women take a combination of vitamins and minerals to improve the weights and outcomes of their babies. This recommendation had progressed to the planning phase for international roll-out of programmes. Using statistical methods and a rigorous experimental design, doubt was cast on the solidity of the evidence base for advocating this. The outcome has been the avoidance of expenditure on a global programme whose results might have been less impressive than imagined.
This is one of a series of notes on how the use of good statistics has made a difference to policy making.