Return to Article Details Bias in exposure-outcome associations when using left truncated birth datasets. Download Download PDF