External validation of a prediction model for an ongoing pregnancy after intrauterine insemination

ObjectiveTo assess the accuracy of our recently developed prediction model in a prospective validation study to predict the outcome of intrauterine insemination (IUI).DesignDescriptive prospective validation study.SettingSeven fertility centers in the Netherlands.Patient(s)Couples treated with IUI of whom the female partner had a regular cycle.Intervention(s)Intrauterine insemination with or without controlled ovarian hyperstimulation.Main Outcome Measure(s)Ongoing pregnancy after intrauterine insemination. Performance of the prediction model was assessed with calibration and discriminative capacity. Calibration was assessed by comparing the predicted ongoing pregnancy rate with the observed ongoing pregnancy rate. Discriminative capacity was assessed with receiver operation characteristic (ROC) analysis. For daily practice, a score worksheet of the validated model was developed to estimate the chance of an ongoing pregnancy after IUI per treatment cycle.Result(s)We included 1,079 subfertile couples who underwent 4,244 cycles of IUI. There were 278 ongoing pregnancies, that is, an ongoing pregnancy rate of 6.6% per cycle. External validation of the model showed good calibration. The predicted probability never differed by more than 1.5% of the mean observed probability. The area under the ROC curve was 0.56 (95% confidence interval, 0.53–0.59) at external validation.Conclusion(s)The prediction model was able to make a good distinction between couples with a good pregnancy chance and those with a poor pregnancy chance after IUI. This model can help in deciding which couples will benefit from IUI and which couples will not.