Predicting the likelihood of a live birth for women with endometriosis-related infertility

Endometriosis affects 10% of women in reproductive age and alters fertility. Its management is still debated notably the timing of surgery and ART in infertility. Several tools have been created to guide the practitioner and the couple yet many limitations persist. The objective is to create a nomogram to predict the likelihood of a live birth after surgery followed by assisted reproductive technology (ART) for patients with endometriosis-related infertility. All women in a public university hospital who attempted to conceive by ART after surgery for endometriosis-related infertility from 2004 to 2016 were included. We created a model using multivariable linear regression based on a retrospective database. Of the 297 women included, 171 (57.6%) obtained a live birth. Age, duration of infertility, number of ICSI-IVF cycles, ovarian reserve and the revised American Fertility Society (rAFS) score were included in the nomogram. The predictive model had an area under the curve (AUC) of 0.77 (95% CI, 0.75–0.79) and was well calibrated. The external validation of the model was achieved with an AUC of 0.71 (95% CI, 0.69–0.73) and calibration was good. The staging accuracy according to AUC criteria for the nomogram compared to the currently used Endometriosis Infertility Index to predict live births were 0.77 (95% CI, 0.75–0.79) and 0.60 (95% CI: 0.57–0.63), respectively. This simple tool appears to accurately predict the likelihood of a live birth for a patient undergoing ART after surgery for endometriosis-related infertility. It could be used to counsel patients in their choice between spontaneous versus ART conception, or oocyte donation.