Endometriosis: studies of a method for the design of a surgical staging system.

This study develops a method of identifying factors that predict pregnancy rates in patients with endometriosis. In 123 women, 60% had adhesions, 89% had implants, and 29% had endometriomas. The most common lesions were cul-de-sac implants (60%), ovarian implants (58%), and ovarian adhesions (55%). With no treatment, the life-table-estimated 3-year pregnancy rate was 40%; with oral contraceptives, 33%; and with surgery, 53%. Clustering techniques suggested new combinations of variables to be tested. We identified structures commonly involved simultaneously in a given patient, defined subgroups based on type of lesion, and developed a method to determine which factors were important in predicting outcome. In our patients, neither of two current methods of endometriosis staging predicted outcome. Further testing of the model systems suggested by the data may enable us to develop a staging system more predictive of pregnancy rates in endometriosis patients.