To the Editor:
I read with great interest the paper by Madani et al. (1). The authors report a randomized controlled study of 110 women in which pushing 0.2 mL of air into the uterus after embryo transfer (ET) was associated with a significant improvement of the clinical pregnancy rate from 9/55 (16.4%) in the control group to 22/55 (40%) in the study group.
Although this study is reporting original and interesting data, several points in the methods, results and conclusions are questionable.
First, this report of a randomized controlled trial is not in accordance with the CONSORT statement (2) and no registration number is mentioned. The statistical methods are not described with enough precision. The dates defining the period of recruitment are not précised. Was there a power calculation when the study was planned? Did the authors expect such a huge difference, for example from preliminary unpublished data, to choose the small but perfectly adequate sample size? Indeed, using a bilateral test with α=0.05 and 1-β=0.8, the sample size to show a 23.6% difference between the groups should be 54. In our opinion, the 55 sample size is therefore statistically suspicious. Did they expect smaller difference, plan a larger study and stop it after interim analysis? Then what were the rules for stopping?
Second, the results are limited by the 16.4% pregnancy rate in the control group. The readers of Fertility and Sterility would have been more interested by an improvement of the pregnancy rate from 40% to 45% (testing this hypothesis would require 1531 women in each group) than from 16.4% to 40%. The authors did not specify if 16.4% is a normal pregnancy rate in their department. If not, why was the pregnancy rate so low in the control group? Recent data show that even trainees without experience perform better than that during their initial experience in ET (3). There might be a bias because the gynecologist performing all the ET could not be blinded to the technique he performed. Similarly, were the patients blinded to the transfer technique?
Third, the external validity of the results has not been discussed by the authors. Can we expect such a spectacular improvement in settings where the mean pregnancy rate is 40% or higher? Therefore, the conclusion of the study goes beyond its findings.
Lionel Dessolle, M.D., M.Sc.
Service de gynécologie-obstétrique et médecine de la reproduction
Centre Hospitalier Universitaire de Nantes
Nantes Cedex 1, France
1. Madani T, Ashrafi M, Jahangiri N, Abadi AB, Lankarani N. Improvement of pregnancy rate by modification of embryo transfer technique: a randomized clinical trial. Fert Ster. DOI: 10.1016/j.fertnstert.2010.03.046.
2. Schulz KF, Altman DG, Moher D, for the CONSORT group. CONSORT 2010 Statement: updated guidelines for reporting parallel group randomized trials. BMJ 2010; 340:c332.
3. Dessolle L, Fréour T, Barrière P, Jean M, Ravel C, Daraï E, Biau DJ. How soon can I be proficient in embryo transfer? Lessons from the cumulative summation test for learning curve (LC-CUSUM). Hum Reprod 2010; 25(2): 380-6.
Published online in Fertility and Sterility doi:10.1016/j.fertnstert.2010.06.085
The Authors Respond:
We would like to thank Dr Lionel Dessolle for his constructive comments.
We express our regret for not stating the registration number in the manuscript. The identifier is NCT00905788, which can be found on ClinicalTrials.gov
The period of recruitment was from February 2009 to July 2009.
According to the clinical pregnancy rates achieved in our pilot study in two groups of case and controls (38.3% versus 16%, respectively; α = 0.05 and 1-β = 0.8), the sample size was calculated as 60 in each group and by adding sample loss of 10%, the estimated sample size was calculated to be approximately 67. However, at the end of the study, the rate of sample loss was higher than anticipated due to cycle discontinuation (such as patients who were at risk for ovarian hyperstimulation syndrome (OHSS), those with no efficient follicle growth, the women who had no oocytes or those with no good-quality embryos). Therefore our final sample size reached 55, which achieved power equal to the first estimated one (1-β = 0.8).
The overall pregnancy rate of our center is more than the one reported in this study. Because our center is a referral center, the patients who come to us are those who were not successful in their previous trials. Naturally, the number of these couples varies in different periods of time. This factor might affect the pregnancy rate.
In fact, we would be happy if we could consider blinding in our study. But blinding is not always feasible or relevant. In our study, we used an insulin syringe to push air into the catheter and no intervention was performed in controls, so we were not able to blind our study due to our procedure . According to studies (1-2), knowledge of the intervention would not greatly affect measurement of objective outcome such as pregnancy rate, which is measured by checking the serum β-human chorionic gonadotropin levels by a laboratory technicians who were not aware of the study group allocation.
“Clearly, the results of RCTs or systematic reviews will never be relevant to all patients and all settings, but they should be designed and reported in a way that allows clinicians to judge to whom the results can reasonably be applied.” (3) The setting and our patients or the characteristics of randomized patients are some issues that can affect external validity and are particular (4); therefore we recommend further RCT studies as stated in the conclusion of the paper.
Tahereh Madani, M.D.
Mahnaz Ashrafi, M.D.
Nadia Jahangiri, M.Sc.
Akram Bahman Abadi, B.Sc.
Narges Lankarani, Ph.D.
Endocrinology and Female Infertility Department
Reproductive Medicine Research Centre
Royan Institute, ACECR
1. Schulz KF, Grimes DA. Blinding in randomised trials: hiding who got what. Lancet 2002; 359:696-700.
2. Wood L, Egger M, Gluud LL, Schulz KF, Jüni P, Altman DG, et al. Empirical evidence of bias in treatment effect estimates in controlled trials with different interventions and outcomes: meta-epidemiological study. BMJ 2008; 336: 601-5.
3. Rothwell PM. External validity of randomised controlled trials: “to whom do the results of this trial apply?. Lancet 2005; 365:82-93.
4. Rothwell PM. Factors that can affect the external validity of randomised controlled trials. PLoS Clin Trials. 2006; 1:e9.
Published online in Fertility and Sterility doi:10.1016/j.fertnstert.2010.06.086