The inconsistency of results across studies as to the safer route of delivery (vaginal versus cesarean) may be the result of the small sample size, the inability to control for confounding characteristics, and selection biases induced mainly by differences in referral rates and patterns.