Predicting the number of oocytes at collection and the fertilization rate for assisted reproductive technologies.

Abstract

Objective: Evaluating the failed IVF cycle often provides useful prognostic information. Before undergoing another attempt, patients experiencing an unsuccessful IVF cycle frequently request information about the probability of future success. To predict the number of oocytes at collection and their fertilization rate, we have introduced the concept of a repeatability coefficient and formulae to predict the next IVF cycle outcome. Design: Retrospective review of IVF cycle data and statistical analysis. Materials/Methods: Patient data with more than two IVF cycles between January 1998 and December 2000 were collected in our IVF center. During this period, all patients were treated by our standard stimulation protocol (LA/HMG, HCG 10,000 IU/ml) before oocyte retrieval. A repeatability coefficient (the intraclass correlation coefficient of repeated measurements on the same individual), (r) was estimated by the formula: r = (MSb − MSw)/(MSb + MSw (k − 1)). Where, MSb is the mean variance among patients, MSw is the mean variance of the different cycle times within a patient, and k indicates the weighted IVF cycle times. Based on this repeatability coefficient, a prediction may be calculated by formula: Px = kr/(1 + (k − 1) r × (Pk − Â) + Â. Px indicates the patient’s predicted value (for example, number of oocytes at collection or fertilization rate for the next cycle), Pk is the patient average of different cycle times, Â indicates the average level of all patients. k and r indicate the weighted IVF cycle times and repeatability coefficient, respectively. Results: A total of 66 patients undergoing 174 IVF retrieval cycles were analyzed. Each patient had an average of 2.63 retrieval cycles (range 2 to 7 cycles), an average retrieval of 11.3 ± 7.29 oocytes, and a repeatability coefficient of r = 0.52 (P < 0.01). In the fertilization rate group, a total of 65 patients with 168 IVF cycles were analyzed. Each patient had an average of 2.58 cycles (range 2 to 7), an average fertilization rate of 57.1% ± 23.5%, and repeatability coefficient of r = 0.58 (P < 0.01). These results indicate that the number of oocytes at collection and the fertilization rate are highly repeatable and allow the use of the formula: Px = 0.52k/(0.48 + 0.52k) × (Pk − 11.3) + 11.3 to predict the future number of oocytes per retrieval and the formula: Px = 0.58K/(0.42+0.58k) × (Pk − 57.1) + 57.1 to estimate the fertilization rate for a patient’s future IVF cycle. Conclusions: The introduction of a repeatability coefficient, commonly used to predict outcomes in animal genetics, to our human IVF program allows us to reliably predict future IVF cycle outcomes. The variables of oocyte number retrieved and fertilization rate demonstrate high repeatability, which provides a basis for the accurate prediction of these outcomes in future cycles. Accurate prediction of a subsequent IVF cycle attempt can provide prognostic information on which patients can make informed decisions and may allow for modification of the stimulation protocol to optimize cycle success.