The first Russian experience in external validation of the Templeton model in the center of assisted reproductive technologies

Objective: To evaluate the validity of the Templeton model in predicting IVF and ET outcome in current Russian clinical practice. Materials and methods. 348 couples undergoing IVF were included in the prospective analysis. The model discrimination was evaluated by C-statistics, calculated as the area under the ROC curve. Calibration was assessed by Cox-Miller method. Results. The Templeton model for predicting live birth after IVF showed the best predictive capacity. The area under the ROC-curve for the original Templeton model was 0.61 (95% CI 0,55-0,67, p = 0,001), the sensitivity and specificity at 0.13 cut-off level were 56% and 60%, respectively. Calibration showed acceptable predictive power in the low values of the predicted probability (less than 10%). Conclusion. The Templeton prediction model can be used in initial clinical decision-making, but due to the lack of it’s predictive power a new model, developed on a more recent data set is highly warranted.