A new paper in PLOS Medicine argues that climate change projections are often misused in health impact studies: they are best suited for shaping public health policies, not for triggering operation