ANALYSIS OF ECTOPIC PREGNANCY DATA USING MARGINAL AND CONDITIONAL MODELS

This work is motivated by a longitudinal study of women and their ecto pic pregnancy outcomes in Lund, Sweden. In this article, we review and apply the Liang-Zeger methodology to the Lund ectopic pregnancy data set. We further analyse the ectopic pregnancy data using conditional m odelling approaches suggested by Rosner and Bonney. From the Lund ecto pic pregnancy data, we learned that PID is the strongest predictor of subsequent development of ectopic pregnancy and that there is a monoto ne relationship between PID severity and ectopic pregnancy. We also le arned that the presence of mycoplasma from lower or upper genital trac t sites at index laparoscopy is also a strong predictor of ectopic pre gnancy. Other correlates of ectopic pregnancy include age at pregnancy and history of gynaecologic surgery. (C) 1997 by John Wiley & Sons, L td.