Fig. 3From: Impact of sample size on the stability of risk scores from clinical prediction models: a case study in cardiovascular diseasePercentile ranges of risk for individuals, subsetted by C-statistic of the models (female cohort, N = 10,000). Each data point represents the 5–95th percentile range in risk for an individual across a group models defined by their C-statistic. Two thirds of models had C-statistic > 0.866%, and one third of models had C-statistic > 0.867%Back to article page