The statistical significance of randomized controlled trial results is frequently fragile: a case for a Fragility Index
Distribution of the number of citations over years.