OBJECTIVE: Development of an endometriosis classification system based on empirically derived stages of the disease, to supplant the Acosta (1973), Kistner (1977), and American Fertility Society (1985) classifications, which are based on arbitrarily defined stages and often fail to predict pregnancy rates. DESIGN: Retrospective cohort analysis. SETTING: University infertility clinic. PATIENTS AND METHODS: Women with endometriosis and > or = 1 year of infertility. Diagnosis of endometriosis was made by direct visualization, with type of lesion (implant or adhesion) at multiple sites recorded; total of 202 patients. All diagnosed infertility problems were treated based on semen analysis, postcoital test, and endometrial biopsy. Pregnancy rates were analyzed by life-table and cluster analyses, and combinations of site and type were also analyzed by Cox's regression model. RESULTS: No individual anatomic site or type significantly affected prognosis, nor was any cluster useful for predicting outcome. CONCLUSION: Anatomic site and type of lesion are insufficient for predicting fertility when used as sole components of a clinical staging system for endometriosis.
OBJECTIVE: Development of an endometriosis classification system based on empirically derived stages of the disease, to supplant the Acosta (1973), Kistner (1977), and American Fertility Society (1985) classifications, which are based on arbitrarily defined stages and often fail to predict pregnancy rates. DESIGN: Retrospective cohort analysis. SETTING: University infertility clinic. PATIENTS AND METHODS: Women with endometriosis and > or = 1 year of infertility. Diagnosis of endometriosis was made by direct visualization, with type of lesion (implant or adhesion) at multiple sites recorded; total of 202 patients. All diagnosed infertility problems were treated based on semen analysis, postcoital test, and endometrial biopsy. Pregnancy rates were analyzed by life-table and cluster analyses, and combinations of site and type were also analyzed by Cox's regression model. RESULTS: No individual anatomic site or type significantly affected prognosis, nor was any cluster useful for predicting outcome. CONCLUSION: Anatomic site and type of lesion are insufficient for predicting fertility when used as sole components of a clinical staging system for endometriosis.