1 Pediatric and Reproductive Endocrinology Branch, NICHD, Building 10, Room 9D42, NIH, Bethesda, Maryland 20892, 2 Department of Obstetrics and Gynecology, Walter Reed Army Medical Center, Washington, D.C., 3 Department of Obstetrics and Gynecology and 4 Department of Biometrics and Preventive Medicine, Uniformed Services University of the Health Sciences, 4301 Jones Bridge Road, Bethesda, Maryland 208144799, USA.
![]() |
Abstract |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Key words: assisted reproduction/embryo transfer/Fellows training/learning-curve/pregnancy rate
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Since embryo transfer is a provider-dependent skill, assessment of skill acquisition becomes an important consideration for training programmes. Two prior reports address training of practitioners in the technique of embryo transfer from an institutional perspective (i.e. groups of trainees). Barber et al. reported a 36% initial pregnancy rate per transfer for nurses, compared with a 29% pregnancy rate per transfer for experienced physicians (Barber et al., 1996). In support of these data, Sinclair et al. reported a pregnancy rate per nurse transfer of 40.2% and per physician transfer of 41% after instruction of nurses in the method of embryo transfer (Sinclair et al., 1998
). Despite the acknowledged importance of the provider at embryo transfer, there are no published accounts of an individual's training experience. The absence of data regarding training on an individual level results in uncertainty regarding the number of transfers that would be considered sufficient for training purposes. Specifically, how many embryo transfers should training programmes target for individuals in training? In addition, it is not clear whether training needs to be adapted to specific individual needs for performance and/or skill development. The relevance of embryo transfer skills to physician training was emphasized by the explicit requirement of this experience for Fellowship training in reproductive endocrinology by the American Board of Obstetrics and Gynecology (American Board of Obstetrics and Gynecology, 2000).
The principal objective of this study was to identify the minimum number of embryo transfers performed by Fellow trainees required to become proficient in the technique. A second objective was to assess each individual trainee's performance as measured by clinical pregnancy rates per transfer to determine whether a learning curve for embryo transfer existed. Since the number of embryos transferred and embryo quality greatly influence pregnancy rate, we controlled for these confounding variables. The results of this study suggest that for some Fellows pregnancy rates were lower for the first 25 embryo transfers, but that for all trainees by 4050 embryo transfers, pregnancy rates were indistinguishable from experienced staff.
![]() |
Materials and methods |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Patients
In general, patients >43 years of age and those with FSH concentrations >15 IU/ml on cycle day 3 (or on day 3 or 10 during a clomiphene citrate challenge test) were excluded from the study. Controlled ovarian stimulation was performed by using a combination of long-term gonadotrophin-releasing hormone (GnRH) agonist (Lupron, 1.0 mg/day; TAP Pharmaceuticals, Deerfield, IL, USA) or microdose flare GnRH agonist (Lupron, 20 µg or 50 µg) and gonadotrophins (Fertinex; Serono Laboratories, Norwell, MA, USA), or a combination of Fertinex and Humegon (Organon, West Orange, NJ, USA) as described (Leondires et al., 1999). Cycles were monitored using serial follicular ultrasonography and measurement of serum oestradiol concentrations. Administration of human chorionic gonadotrophin (HCG) and retrieval of oocytes was performed when follicular measurements and oestradiol concentrations were appropriate. Embryology laboratory protocols complied with Society for Assisted Reproductive Technologies guidelines. High-grade embryos were defined as grade I and/or grade II embryos, and all blastocysts transferred were scored as high-grade embryos (Scott et al., 1999
). Patient selection was not random, and in general patients with a history of a difficult transfer or failure at a prior ART attempt were not a candidate for transfer by a Fellow.
Transfer method
Five trainee Fellows performed embryo transfers under the direct supervision of an attending physician. Trainees performed intrauterine inseminations or mock uterine transfers using the Wallace catheter (Simcare Ltd, West Sussex, UK) to become familiar with handling of the catheter. Fellows in the Reproductive Endocrinology and Infertility Training Program were required to document experience with the Wallace catheter (~30 mock or intrauterine insemination procedures) and observe transfers before performing the embryo transfers. Patients were requested to arrive for transfer with a full bladder. All embryo transfers were performed using a Wallace catheter in the lithotomy position. Unless the mock transfer was difficult, or the transfer was expected to be difficult, no mock transfer was performed immediately prior to embryo transfer. Transabdominal ultrasonography (Ultramark 4 with 5 MHz transducer; ATL, Bothell, WA, USA) was performed before and during the transfer to monitor catheter placement. Catheter insertion in the uterine cavity was monitored and intrauterine transfer was visualized in most cases. The catheter remained in the uterine cavity for ~510 s after the embryos were deposited, was slowly rotated 180°, removed, and passed to the embryologist. The plunger of the catheter remained depressed from the point at which the embryos were deposited until the catheter was passed to the embryologist. Individual results were prospectively tracked, but there was no apparent alteration or change in individual techniques over the duration of the study (Hearns et al., 1999).
Statistical analysis
Because data were analysed retrospectively, sample size was defined by the study interval or a maximum of 50 embryo transfers, and was not determined by power analysis. The number 50 was chosen based on other reports of training experiences (See et al., 1993; Moore et al., 1995
; Watson et al., 1996
). The main outcome variable was clinical pregnancy, defined as the presence of a gestational sac on ultrasonography at 68 weeks. Ectopic pregnancies were not counted as gestational sacs or clinical pregnancies. High-order multiple pregnancies (HOM) were defined as triplet or quadruplet gestations. The
2 test was used to analyse statistical differences in pregnancy rates among Fellow trainees. Differences in patient age were analysed by using the t-test after testing normality of the distribution. Number of embryos and embryo grade were analysed by using one-way ANOVA. An
error of < 0.05 was considered statistically significant.
Training experience was evaluated by two approaches reported for other procedures (See et al., 1993; Moore et al., 1995
; Watson et al., 1996
). Individual training experience was evaluated by plotting pregnancy outcome according to the chronological order of the embryo transfer. In addition, experience of each trainee was arbitrarily divided roughly at the midpoint (25 transfers) and outcome of the two subgroups (125 and 2650 transfers) was compared. Institutional experience was assessed by comparing the experience of all trainees with experienced staff (Hearns-Stokes et al., 2000
).
![]() |
Results |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
|
|
Likewise, with the exception of Fellow no. 1, patient age did not change significantly between the first 25 and second 25 embryo transfers. It should be noted that the average age of patients in the second 25 transfers of Fellow no. 1 increased significantly (P < 0.05), while the pregnancy rate fell from 60% for the first 25 transfers to 48% for the second 25 transfers. It is quite possible that the fall in pregnancy rate for the second 25 transfers was due in part to the older patients transferred in the second half of Fellow no. 1's training experience. Consistent with a fall in pregnancy rates, the average number of high grade embryos decreased from 3.0 to 2.3, also possibly contributing to the lower pregnancy rate for the second 25 embryo transfers for Fellow no. 1.
![]() |
Discussion |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
While our results clearly indicate that on an individual level, 50 procedures is sufficient, it is possible that some trainees require fewer, or more, procedures to attain proficiency. To address this issue, we tested whether a learning curve might exist. First, we evaluated individual improvement over time, and second, we divided each individual's experience into two subgroups. Both approaches have been successfully used to gauge skill acquisition (See et al., 1993; Moore et al., 1995
; Watson et al., 1996
) for a similar number of skill-dependent procedures. Using both methods, three of four Fellows for whom data were available showed improvement in clinical pregnancy rates from the first 25 to the second 25 transfers. In marked contrast, Fellow no. 1 showed an, albeit non-significant, decrease in pregnancy rate, but interpretation of this result was complicated by the fact that patient age increased significantly, and the average embryo quality decreased in conjunction with the decrease in pregnancy rate during the second period.
If Fellow no. 1 were to be excluded from the group because of the significant difference in patient age, the improvement in pregnancy rate between the first 25 and the second 25 embryo transfers for the remaining Fellows (n = 3) would be from 33 to 53% (P < 0.05). It is possible that Fellow no. 1 may represent an outlier, and a larger sample size may be required to determine whether a statistically significant difference would exist between the first 25 and the second 25 embryo transfers. Because embryo quality and the number of embryos transferred influence pregnancy rates, we controlled for these confounding variables but did not find statistically significant differences between the trainees. With the exception of Fellow no. 1 as noted above, differences in performance may not be attributed to differences in patient age.
Reports that individuals may influence outcome after embryo transfer (Karande et al., 1999; Hearns-Stokes et al., 2000
), are strong arguments for assessment of a learning curve, since differences among providers could be caused by differences in training or skill acquisition. The reason for the substantial influence of the provider upon pregnancy rates after embryo transfer (Karande et al., 1999
; Hearns-Stokes et al., 2000
) remains unclear, although provider-dependent characteristics such as manual dexterity or patience are possible explanations. If manual dexterity influenced pregnancy outcome, we reasoned that a learning curve might be present, but we wished to minimize the `learning curve effect' due to possible deleterious effects upon patient care. For this reason, before engaging trainees in actual embryo transfers, we required mock transfers to familiarize trainees with the features of the Wallace catheter. Accordingly, it should be noted that in this study the number of 50 embryo transfers followed a rather extensive training experience of 30 mock transfers. It is possible that increasing the number of mock transfers from the 30 used in this study might eliminate or reduce the lower pregnancy rate observed in the first 25 patients.
Lastly, due to the fact that training programmes only enrol one or two Fellows per year, and we planned to compare rates between Fellows within one programme, the study necessarily spanned several years. Stated differently, the experience of Fellow no. 2 occurred in a different year than Fellow no. 5, but comparison of the average number of high grade embryos per transfer for these two Fellows was similar, and in both cases, the pregnancy rate rose after the first 25 embryo transfers. Thus, despite the fact that each trainee performed transfers in a different year, the outcomes were remarkably similar. Since the experiences spanned several years, cycle-to-cycle variations in pregnancy rate could have influenced the results. In general, cycle-to-cycle variation in pregnancy rate was not problematic, and these data have previously been published for our programme (Hearns-Stokes et al., 2000).
In conclusion, the pregnancy rate of the five Fellow trainees as a group was 45.5% per embryo transfer and closely approximated the overall programme pregnancy rate of 47.3% during the same period. The chronological plot of individual rates was arguably more revealing, since this analysis indicated that after 2530 transfers most trainees began to achieve acceptable rates of pregnancy.
![]() |
Acknowledgements |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
Notes |
---|
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Barber, D., Egan, D., Roas, C. et al. (1996) Nurses performing embryo transfer: successful outcome of in- vitro fertilization. Hum. Reprod., 11, 105108.[Abstract]
Fanchin, R., Righini, C., Olivennes, F. et al. (1998) Uterine contractions at the time of embryo transfer alter pregnancy rates after in-vitro fertilization. Hum. Reprod., 13, 19681974.[Abstract]
Gonen, Y, Dirnfeld, M., Goldman, S. et al. (1991) Does choice of catheter for embryo transfer influencethe success rate of in-vitro fertilization? Hum. Reprod., 6, 10921094.[Abstract]
Hearns, R.M., Hill, J., Scott, L. et al. (1999) The effect of embryo transfer technique on clinical pregnancy rates during in-vitro fertilization [abstract]. In ASRM/CFAS 1999 Annual Meeting Supplement, Toronto, CA: American Society for Reproductive Medicine. Fertil. Steril., 72, S34.
Hearns-Stokes, R.M., Miller, B.T., Scott, L. et al. (2000) Pregnancy rates after embryo transfer depend on the provider at embryo transfer. Fertil. Steril., 74, 8086.[ISI][Medline]
Karande, V.C., Morris, R., Chapman, C. et al. (1999) Impact of the `physician factor' on pregnancy rates in a large assisted reproductive technology program: too many cooks spoil the broth? Fertil. Steril., 71, 10011009.[ISI][Medline]
Leondires, M., Escalpes, M., Segars, J. et al. (1999) Microdose follicular phase GnRH-a is a suitable alternative to luteal phase GnRH-a for ovarian stimulation at IVF. Fertil. Steril., 72, 10181023.[ISI][Medline]
Moore, M.J., Bennett, C.L. and The Southern Surgeons Club (1995) The learning curve for laparoscopic cholecystectomy Am. J. Surg., 170, 5559.[ISI][Medline]
Roseboom, T.J., Vermeiden, J.P.W., Schoute, E. et al. (1995) The probability of pregnancy after embryo transfer is affected by the age of the patient, cause of the infertility, number of embryos transferred and the average morphology score, as revealed by multiple logistic regression analysis. Hum. Reprod., 10, 30353041.[Abstract]
Scott, L., Alvero, R., Broussard, J. et al. (1999) Development of positive morphological predictors for human blastocyst implantation [abstract]. In ASRM/CFAS 1999 Annual Meeting Supplement, Toronto, CA: American Society for Reproductive Medicine. Fertil. Steril., 72, S227.
See, W.A., Cooper, C.S. and Fischer, R.J. (1993) Predictors of laparoscopic complications after formal training in laparoscopic surgery. JAMA, 270, 26892692.[Abstract]
Sinclair, L., Morgan, C., Lashen, H. et al. (1998) Nurses performing embryo transfer: the development and results of the Birmingham experience. Hum. Reprod., 13, 699702.[Abstract]
Visser, D.S., Fourie, F. and Kruger, H.F. (1993) Multiple attempts at embryo transfer: effect on pregnancy outcome in an in vitro fertilization programme. J. Assist. Reprod. Genet., 10, 3743.[ISI][Medline]
Watson, D.I., Baigrie, R.J. and Jamieson, G.G. (1996) A learning curve for laparoscopic fundoplication. Ann. Surg., 224, 198203.[ISI][Medline]
Submitted on December 15, 2000; accepted on March 23, 2001.