ABSTRACT
In epidemiologic and clinical risk assessment studies of adverse pregnancy outcomes, gestational age is clearly one of, if not the most important, factor in assessing perinatal risk. Estimation of gestational age in large, population-based studies relies heavily on the timing of the last normal menstrual period as reported by the patient. Unfortunately, the men-strual estimate of gestational dating is fraught with errors-both systematic and random. Inaccurate recall of the date of the last normal menstrual period is one of the chief biases in the process of over-and underestimation of gestational age1. In contrast, bleeding very early in pregnancy-often mistaken for menstrual bleeding2-frequently also leads to underestimation of gestational duration. Irregular or delayed menstrual cycles lead to an overestimation of gestational duration, as these women may not exhibit the presumed 15-day interval between ovulation and menstrual bleeding3-5.
