We need studies of the mortality effect of vitamin A supplementation, not surveys of vitamin A deficiency

It is usually acknowledged that high‐dose vitamin A supplementation (VAS) provides no sustained improvement in vitamin A status, and that the effect of VAS on mortality is more likely linked to its immunomodulating effects. Nonetheless, it is widely assumed that we can deduce something about the need for continuing or stopping VAS programs based on studies of the biochemical prevalence of vitamin A deficiency (VAD). This is no longer a tenable assumption. The justification for using VAS is to reduce child mortality, but there is now doubt that VAS has any effect on overall child mortality. What we need now are not surveys of VAD, but proper randomized trials to evaluate whether VAS has beneficial effects on overall child survival.