• Lionel Dessolle, Thomas Fréour, Paul Barrière, Miguel Jean, Célia Ravel, Emile Daraï, and David J Biau.
• Service de Biologie et Médecine de la Reproduction, Centre Hospitalier Universitaire de Nantes, 38, Bd Jean Monnet, Nantes Cedex 1 44093, France. lionel.dessolle@tnn.aphp.fr
• Hum. Reprod. 2010 Feb 1; 25 (2): 380-6.

BackgroundEmbryo transfer, a crucial step for achieving pregnancy after in vitro fertilization, is an operator-dependent technique but the number of procedures required for a trainee to reach proficiency is unknown. We set out to evaluate the learning curve (LC) of embryo transfer using a specifically designed statistical tool.MethodsThe first embryo transfers performed by five trainees were monitored by the cumulative summation test for learning curve (LC-CUSUM), a statistical tool designed to indicate when a process has reached a predefined level of performance. The main outcome measure was a positive hCG test. A 40% pregnancy rate (PR) per transfer was chosen to define adequate performance and a PR of 20% was considered inadequate. After the learning phase, standard CUSUM were implemented to ensure that performance was maintained. The same CUSUM parameters were also applied to monitor 241 consecutive embryo transfers performed by a senior gynaecologist.ResultsBetween 11 and 99 embryo transfers were necessary for the trainees to reach the predefined level of performance. Simple and intuitive graphical representations of the LCs were generated. CUSUM tests confirmed that performance was maintained after the learning phase. The PR of the senior gynaecologist was 42.7% and the CUSUM showed that performance remained adequate throughout the 241 procedures.ConclusionsThis study provides an exportable model for a quantitative monitoring of the LC of embryo transfer as well as a reference curve for continuous monitoring of performance in embryo transfer. The length of the LC of embryo transfer is highly variable, justifying a tailored training to learn this procedure.

13 keywords...

hide…