Left Truncation, Susceptibility, and Bias in Occupational Cohort Studies