Human Reproduction, Vol.27, No.12 pp.

3467–3473, 2012
Advanced Access publication on September 27, 2012 doi:10.1093/humrep/des341

ORIGINAL ARTICLE Infertility

A prospective randomized multicentre

study comparing vaginal progesterone
gel and vaginal micronized progesterone
tablets for luteal support after in vitro
fertilization/intracytoplasmic sperm
injection
Christina Bergh 1,* and Svend Lindenberg 2 on behalf of the Nordic
Crinone study group
1
Department of Obstetrics and Gynaecology, Institute of Clinical Sciences, Sahlgrenska Academy, Gothenburg University, Reproductive
Medicine, Sahlgrenska University Hospital, Gothenburg, Sweden 2Copenhagen Fertility Center, Copenhagen, Denmark

*Correspondence address. Tel: +46-31-3422992; Fax: +46-31-418717; E-mail: christina.bergh@vgregion.se

Submitted on May 23, 2012; resubmitted on August 12, 2012; accepted on August 22, 2012

summary question: Is vaginal progesterone gel equivalent to vaginal micronized progesterone tablets concerning ongoing pregnancy
rate and superior concerning patient convenience when used for luteal support after IVF/ICSI?
summary answer: Equivalence of treatments in terms of ongoing live intrauterine pregnancy rate has not been demonstrated; the
95% confidence interval (CI) for the difference in ongoing pregnancy rate (28.2 to 0.1%) did not lie entirely within the pre-specified equiva-
lence interval 27 to 7%.
what is known already: No significant differences in clinical pregnancy rates have been observed between vaginal progesterone
gel and other vaginal progesterone products in earlier studies. However, all previous studies included a limited number of patients.
study design, size and duration: This was a randomized, multicentre, controlled, assessor-blinded equivalence trial in 18
fertility centres in Denmark and Sweden between March 2006 and January 2010. A web-based randomization program was used with con-
cealed allocation of patients. Patients were randomized to one of two groups: vaginal progesterone gel or vaginal micronized progesterone
tablets. There was no blinding of patients.
participants and setting: A total of 2057 women ≤40 years of age were included and down-regulated, using the long
agonist protocol and rFSH for stimulation. Luteal support was given for 19 days after embryo transfer or until a negative pregnancy
test Day 14 after embryo transfer. Patient convenience was assessed using questionnaires to be filled in 14 days after embryo transfer,
before pregnancy test.
main results and the role of chance: Ongoing intrauterine pregnancy rates were 299/991 (30.2%) (95% CI 27.3 –
33.0%) in the progesterone gel group and 324/992 (32.7%) (29.7– 35.6%) in the micronized progesterone tablet group. The difference
in ongoing pregnancy rates between the groups was 24.1% (28.2 to 0.1%) and the difference in live birth rates was 23.4% (27.4 to
0.7%), both calculated after correction for significant confounders. Patient convenience and ease of use (1 ¼ very convenient, 10 ¼ very
inconvenient) was in favour of progesterone gel, as the overall score was 2.9 (2.7–3.0) for progesterone gel and 4.8 (4.7– 5.0; P ,
0.0001) for micronized progesterone tablets. This large equivalence trial shows that, even though equality could not be demonstrated,
there is no substantial difference in ongoing pregnancy rate between vaginal progesterone gel and vaginal micronized progesterone
tablets. It also shows that progesterone gel is considered more convenient by the patients.
bias, confounding and other reasons for caution: Blinding of patients was not possible in this study, but since
the outcome (pregnancy) is robust, blinding would have been unlikely to affect the results. Unfortunately, owing to an error in the

& The Author 2012. Published by Oxford University Press on behalf of the European Society of Human Reproduction and Embryology. All rights reserved.
For Permissions, please email: journals.permissions@oup.com
3468
randomization, the intended age distribution allocated older women to the micronized progesterone tablet group. In the analysis of
results, adjustments were made for age and number of embryos transferred.
generalizability to other populations: The results can be generalized to other women ≥18 and ≤40 years of age under-
going IVF/ICSI who have regular menstrual cycles (25–35 days), both ovaries present and no more than two previous failed IVF attempts.
study funding/competing interest: Merck Serono supported the study but had no influence on the design of the study and was
not involved in the analysis of the results or preparation of the manuscript.
trial registration number: The trial was issued with the EudraCT number 2005-001248-22 with the Protocol code number 95576471.
Key words: IVF / luteal phase support / vaginal progesterone / randomized controlled trial
Introduction
Luteal phase support is an essential aspect of assisted reproduction
techniques (ARTs). Previous studies have demonstrated that proges-
terone is superior to placebo and equally efficient as hCG for luteal
support (van der Linden et al., 2011). Although earlier studies found
significantly higher pregnancy rates for i.m. compared with vaginal ad-
ministration (Propst et al., 2001), later studies (Yanushpolsky et al.,
2011) and meta-analyses (Zarutskie et al., 2009; van der Linden
et al., 2011) found comparable results. In one prospective study, al-
though not randomized (Silverberg et al., 2012), a higher live birth
rate was found in the group receiving vaginal progesterone compared
with the group receiving i.m. progesterone.
Moreover, vaginal progesterone was better tolerated than i.m. pro-
gesterone administration (Yanushpolsky et al., 2010). Several other
studies, including a recent meta-analysis (Polyzos et al., 2010), have
compared vaginal progesterone gel to other vaginal progesterone pro-
ducts and have found no significant differences in clinical pregnancy
rates. However, all included studies were of limited size and none
had sufficient power to determine whether there was any difference
in clinical outcome between different compounds.
The aim of this large trial was to compare vaginal progesterone gel
with vaginal micronized progesterone tablets in terms of ongoing preg-
nancy rates and patient convenience in patients undergoing IVF/ICSI.
Materials and Methods
Study design
This was a prospective, randomized, controlled multicentre trial performed
in 18 fertility centres in Denmark and Sweden. Patients were recruited
between March 2006 and January 2010 and randomized to one of two
treatment groups: Vaginal progesterone gel (Crinone 8%, 90 mg, Merck
Serono, Geneva, Switzerland) or vaginal micronized progesterone tablets
[in Denmark Progestan, 200 mg, MSD, Ballerup, Denmark and in Sweden
Progesteron MIC, 400 mg, Apoteksbolaget Produktion och Laboratorier
(APL) Sweden]. The study was assessor blinded (blinded for treating phys-
ician and statistician). The primary objective of the study was to show
equivalence in ongoing pregnancy rate at gestational week 7 between the
two treatment groups. Ongoing pregnancy was defined as a sonographically
verified intrauterine pregnancy, with a fetus with a heartbeat, 5 weeks after
embryo transfer (gestational week 7). Secondary objectives were to show
whether vaginal progesterone gel was superior to vaginal micronized pro-
gesterone tablets in terms of convenience and ease of use. Further second-
ary objectives were to compare positive pregnancy rates between groups,
bleeding pattern before pregnancy test, implantation rate, live birth rate
Bergh and Lindenbergon
and multiple birth rates. The inclusion criteria were women ≥18 and
≤40 years of age, having regular menstrual cycles (25–35 days), presence
of both ovaries and no more than two previous failed IVF attempts. The
exclusion criteria were known drug abuse, allergies to the study medication,
cancelled embryo transfer performed in the study cycle, embryo transfer
not performed on Day 2 and previous participation in the study. The
ethical committees of the participating centres approved the study and
the patients signed informed consent. The study was performed in accord-
ance with the Declaration of Helsinki and the ICH (International Conference
on Harmonization) Harmonized Tripartite Guidelines for GCP.
Patients were randomized directly after embryo transfer, in a propor-
tion of 1:1 into one of the two treatment groups. The patients had
been down-regulated according to the long protocol with GnRH agonist
(Synarela nasal spray, Pfizer, New York, NY, USA), 600 mg daily until
the start of FSH stimulation, followed by 400 mg daily. Stimulation was
performed with recombinant FSH (Gonal-f, Merck Serono, Geneva, Switz-
erland) according to the clinical routines of the different clinics. To trigger
ovulation, recombinant hCG (rhCG), (Ovitrelle, Merck Serono) was used.
Luteal support commenced on the day of embryo transfer, Day 2 after
oocyte retrieval, and was given for a total of 19 days (or until a negative
pregnancy test). Vaginal progesterone gel was administered once daily.
Vaginal micronized progesterone tablets were administered three times
daily. Pregnancy was detected using a urinary or serum hCG test on
Day 14 after embryo transfer. For patients having a positive pregnancy
test, a vaginal sonography was performed 5 weeks after embryo transfer
to establish whether the pregnancy was ongoing. A questionnaire regard-
ing convenience and comfort was filled in on Day 14 after embryo transfer,
before the pregnancy test. The questions included the following areas:
ease of use, hygiene, interference with coitus, itching, burning, pain,
leakage, time to administer the drug and overall impression.
Statistics
Patients were preliminarily recruited before starting the IVF cycle. If the
patient fulfilled all the inclusion criteria and none of the exclusion criteria,
the randomization took place on the day of embryo transfer after the
transfer procedure. The study was designed to show equivalence in
ongoing pregnancy rate between the two treatment groups. The true
ongoing pregnancy rates in the two groups were assumed to be 0.30
and 0.32. The probability was 80% that the 95% confidence interval (CI)
for difference in pregnancy rates between the groups was contained in
the interval (20.07, 0.07) if 1343 patients were included in each group.
The randomization protocol was a version of Pocock and Simon (1975),
often referred to as optimal allocation or minimization. The age of the
woman (≤35 or .35 years) and the number of embryos transferred
(1 or .1) were intended to be included as prognostic factors. A web-based
randomization program was used with concealed allocation of patients.
Descriptive statistics for continuous and categorical variables included
mean, standard deviation, median, minimum and maximum and
Luteal support after IVF/ICSI
frequencies. For the primary end-point, a 95% CI for the difference in
ongoing pregnancy rate between the groups was calculated. If the 95%
CI was entirely within 20.07 to +0.07, equivalence between the treat-
ments would be declared. The primary analysis was performed as a per
protocol analysis. For secondary continuous end-points, analysis of covari-
ance was used to detect differences between the groups. Covariates
included were: age of the woman, number of embryos transferred,
country and centre. For binary end-points, logistic regression or
Cochran –Mantel – Haenszel was used with adjustments for age, number
of embryos transferred, country and center. For each question in the
patient questionnaire regarding convenience data proportional odds
models were used. Adjustments were made for age group, number of
embryos transferred, country and centre. Point estimate and 95% CI
limits for the odds ratio were calculated. In case the assumptions of pro-
portionality were not fulfilled, the Cochran– Mantel– Haenszel models
were used. The variable overall impression was assessed as a score:
1 ¼ very convenient and 10 ¼ very inconvenient. Overall impression
was analysed with analysis of covariance models including age, number
of embryos transferred, country and centre. Least squares mean with a
95% CI were calculated.
All analyses on secondary end-points were performed on both the
intention-to-treat and the per protocol populations. Sub-analysis per
country was performed for all efficacy variables.
Error in randomization
Unfortunately, owing to data entry errors for the birth date of two patients
used to calculate the age for the randomization, the resulting distribution of
patients by age categories was very unbalanced. The age of the two patients
having an error of data entry was calculated to be 1809 years and 1810
years. Both patients were in the vaginal progesterone gel arm of the
study. Therefore, subsequent patients tended to be allocated to that arm
if they were younger than the average, and to the vaginal micronized proges-
terone tablet arm if they were older. Two well-recognized and independent
statisticians were contacted to give advice on how to manage this problem.
Both statisticians came to the same conclusion: the results would be
correct, provided that a stratified analysis with regard to age as a continuous
variable was performed. The investigators followed this advice and the
results are presented accordingly in this article.
Results
The trial was stopped after almost 4 years, owing to gradually falling
recruitment, before the total estimated number of patients had
been allocated. A total of 2662 patients fulfilling the initial inclusion cri-
teria were recruited; 2057 patients were randomized and 1998
patients completed the study (Fig. 1). Patient demographics are
shown in Table I. Owing to the described error in the randomization
procedure (see Statistics) mean age and age group differed significantly
between the two groups. Otherwise, there were no significant differ-
ences in demographics between the groups.
Pregnancy and live birth
Ongoing intrauterine pregnancy, the primary end-point, was 30.2%
(95% CI 27.3 –33.0%) in the vaginal progesterone gel group and
32.7% (29.7– 35.6%) in the vaginal micronized progesterone tablet
group. After adjustment for the woman’s age and the number of
embryos transferred, the difference in ongoing pregnancy rate
between the groups was 24.1% (95% CI –8.2 to 0.1%). There was
a significant difference in the proportion of positive pregnancy tests,
3469
40.0% for vaginal progesterone gel versus 44.2% for vaginal micronized
progesterone tablets (P ¼ 0.01). Implantation rate was 25.8% in the
vaginal progesterone gel group and 26.8% in the vaginal micronized
progesterone tablets group. The rate of miscarriage, 12.2% in the pro-
gesterone gel group and 15.9% in the micronized progesterone tablets
group, did not differ significantly. The same applies to the live birth
rate, 28.4% in the progesterone gel group and 30.1% in the micronized
progesterone tablets group (Table III).
Considering secondary end-points, the FSH starting dose was signifi-
cantly lower in the progesterone gel group, [mean 167.2 standard devi-
ation (SD) 54.3 IU versus 179.6 (SD 63.5) IU (P , 0.001)]. The number
of embryos transferred did not differ between the groups, while the
number of good-quality embryos transferred differed significantly
[mean 0.8 (SD 0.6) in the progesterone gel group versus 0.9 (SD 0.6)
in the micronized progesterone tablets group (P , 0.01)] (Table II).
Patience convenience
The questionnaire showed that patient convenience and ease of use
was in favour of vaginal progesterone gel in six out of nine questions:
easy administration, hygiene, interference with coitus, leakage, time
consumption and overall impression, while no significant differences
were noted for itching, burning and pain (Fig. 2). The overall impres-
sion about vaginal progesterone gel was 2.9 (2.7 –3.0) and of vaginal
micronized progesterone tablets 4.8 (4.7 –5.0; P , 0.0001) (1 ¼
very convenient, 10 ¼ very inconvenient).
Bleeding pattern
A higher proportion of patients reported bleeding as well as more
abundant bleeding before the pregnancy test in the progesterone gel
group, 52.1 versus 38.0% (P , 0.0001). Among patients with bleeding,
menstrual bleeding was reported by 44.7% in the progesterone
gel group and 17.3% in the micronized progesterone tablets group
(P , 0.0001), while more patients in the micronized progesterone
tablets group reported spotting. However, when sonography was per-
formed, there was no significant difference between the groups in
terms of bleeding before pregnancy test among patients with an
ongoing pregnancy [progesterone gel 9.4% and micronized progester-
one tablets 11.5%, difference 22.1% (95% CI –6.9 to 2.7%)].
Adverse events
Most adverse and serious adverse events were unrelated to the study
drug. Six adverse events were reported to be possibly related to the
study medications; breast tenderness, rash, vulvo-vaginal pruritus,
headache, fungal infection and fatigue.
Discussion
The main finding of this large multicenter trial was that equivalence of
treatments in terms of ongoing live intrauterine pregnancy rate was
not demonstrated. The 95% CI for the difference in ongoing preg-
nancy rate did not lie entirely within the pre-specified equivalence
interval 27 to 7%. Although equivalence could not be declared, no
substantial difference existed in the primary end-point, ongoing preg-
nancy rate, between vaginal progesterone gel and vaginal micronized
progesterone tablets. After performing adjustment for age and
number of embryos transferred, the difference in ongoing pregnancy
3470
Figure 1 Disposition of patients.
rate between progesterone gel and micronized progesterone tablets
was 24.1% (95% CI 28.2 to 0.1%). Previous studies comparing dif-
ferent vaginal progesterone treatments have included significantly
fewer patients, resulting in less precise conclusions (Polyzos et al.,
2010). There is no doubt that luteal support following FSH stimulation
in ART is essential (Pabuccu and Akar, 2005). A successful pregnancy
depends mainly on the following factors: age of the woman, ovarian
Bergh and Lindenbergon
reserve, embryo quality and synchronization between the embryo de-
velopment and the endometrium. The supra-physiological steroid level
accomplished after controlled ovarian stimulation and co-treatment
with GnRH agonist or antagonist have both been discussed as deteri-
orating the luteal phase physiology (Macklon and Fauser, 2000), neces-
sitating luteal phase support with progesterone to secure the
reproductive outcome. Many clinicians have expressed concern that
Luteal support after IVF/ICSI 3471

Table I Demographics. Table II Stimulation results.

Crinone 8% Progestan/ Crinone Progestan/ P-value

(90 mg) progesterone 8% (90 mg) progesterone
micronized micronized
........................................................................................ ........................................................................................
Number of patients 991 992 Number of 991 992
(per protocol) patients (per
Age (years) mean (SD) 31.1 (3.5) 32.7 (4.1)a protocol)

Height (cm) mean (SD) 168.0 (6.6) 167.9 (6.3) FSH starting dose 167.2 (54.3) 179.6 (63.5) ,0.0001
(IU) mean (SD)
Weight (kg) mean (SD) 67.0 (11.9) 67.5 (11.5)
Number of days 11.4 (1.8) 11.4 (1.8) 0.9717
BMI (kg/m2) mean 23.7 (3.8) 23.9 (3.8)
with FSH mean
(SD) (SD)
Previous fresh cycles Number of 10.5 (5.3) 10.0 (5.1) 0.0526
0 773 741 oocytes retrieved
1 169 195 mean (SD)
2 49 56 Number of 6.0 (4.0) 5.8 (3.8) 0.2773
embryos on Day 2
Cycle length (days) 28.7 (1.9) 28.6 (1.9)
post OPU mean
mean (SD)
(SD)
Age-group
Number of GQE 2.1 (2.2) 2.2 (2.4) 0.1079
≤35 years 90.0% 68.0%a available mean
.35 years 10.0% 32.0%a (SD)
Causes of infertility Number of GQE 0.8 (0.6) 0.9 (0.6) 0.0175
transferred mean
Tubal factor 148 (14.9%) 156 (15.7%)
(SD)
Anovulation 25 (2.5%) 16 (1.6%)
Number of 1.3 (0.5) 1.3 (0.5) 0.3886
Male infertility 441 (44.5%) 425 (42.8%) embryos
Unexplained 316 (31.9%) 341 (34.4%) transferred mean
Endometriosis 40 (4.0%) 34 (3.4%) (SD)

Other 21 (2.1%) 20 (2.0%) Number of 2.0 (2.6) 2.0(2.4) 0.6288

embryos for
SD, standard deviation. freezer mean (SD)
a
Age distribution (mean age and age-groups) differed significantly between the
groups. See Material and Methods, paragraph Statistics, error in randomization. All OPU, ovum pick-up; GQE, good-quality embryos; SD, standard deviation.
other variables did not differ significantly between groups. Crinone versus vaginal progesterone.

Table III Pregnancy data.

Crinone 8% n and point estimate Progestan/progesterone Micronized n Difference between P-valuea

with 95% CI (n 5 991) and Point estimate with 95% CI treatments with a 95% CIa
(n 5 992)
.............................................................................................................................................................................................
Clinical 320, 32.3% (29.4 –35.3%) 359, 36.2% (33.2 –39.3%) 25.2% (29.4 to 20.9%) 0.0178
pregnancy
Ongoing 299, 30.2% (27.3 –33.1%) 324, 32.7% (29.7 –35.7%) 24.1% (28.2 to 20.1%) 0.0542
intrauterine
pregnancy
Positive hCG 396, 40.0% (36.9 –43.0%) 438, 44.2% (41.1 –47.2%) 25.8% (210.2 to 21.3%) 0.0128
Implantationb 331/1285, 25.8% (23.4 –28.2%) 350/1304, 26.8% (24.4 –29.3%) 21.1% (24.4 to 2.3%) 0.5617
Miscarriagec 39, 12.2% (8.6– 15.8%) 57, 15.9% (12.1 –19.7%) 21.3% (26.4 to 3.8%) 0.6239
Live birth 281, 28.4% (25.6 –31.2%) 299, 30.1% (27.3233.0%) 23.4% (27.4 to 0.7%) 0.1062
Multiple birth 31, 11.0% (7.6– 15.3%) 21, 7.0% (4.4– 10.5%) 2.3% (24.3 to 8.8%) 0.4941

Crinone versus vaginal progesterone.

a
Adjustment for age and number of embryos transferred.
b
No adjustments.
c
Calculated per clinical pregnancy, including both early and late miscarriages ,22 weeks.
3472
Figure 2 Patient satisfaction. Columns represent the percentage of n ¼ 991 patients treated with Crinone progesterone gel and n ¼ 992 patients
treated with Progestan micronized progesterone tablets.
there might be a significant difference in efficacy between the diverse
forms of vaginally applied progesterone. It is therefore reassuring that
the results, from this large trial, demonstrate no substantial difference
in the number of ongoing pregnancies and no significant differences in
miscarriage rate or live birth rate between the vaginal progesterone
gel group and the vaginal micronized progesterone tablets group.
A second finding of the present study was that vaginal progesterone
gel was better tolerated by the patients. Not only ease of administra-
tion, hygiene, interference with coitus, leakage and time consumption,
but particularly the overall impression was in favour of the vaginal pro-
gesterone gel. Better patient convenience for vaginal progesterone gel
compared with intramuscularly administered progesterone was also
reported in a recent study (Yanushpolsky et al., 2010). Moreover, pre-
vious studies, although including a limited number of patients, reported
vaginal progesterone gel to be superior to other vaginal progesterone
products in terms of patient tolerability and ease of use (Ludwig et al.,
2002; Ng et al., 2003; Simunic et al., 2007; Lan et al., 2008).
More patients noted early bleeding, before the pregnancy test,
in the progesterone gel group. However, of greater importance,
among the patients having an ongoing pregnancy, bleeding was
noted at a similar rate in the two groups, 9.4 and 11.5%, respectively.
More bleeding, when using vaginal progesterone gel, has previously
been reported in a study, comparing it with i.m. progesterone admin-
istration (Yanushpolsky et al., 2011). In accordance with our results,
early bleeding occurred at similar rates among patients with an
ongoing pregnancy in both groups. Only the non-pregnant patients
had a higher incidence of early bleeding when treated with vaginal
progesterone gel.
Bergh and Lindenbergon
This study highlights how a randomization error can affect the dis-
tribution of patients in a marked way. Computerized, web-based ran-
domization is regarded as the most valid method of randomization,
minimizing the risk of bias. In the present study, a web-based random-
ization program was used, with stratification for two variables: the age
of the woman (≤35 or .35 years) and the number of embryos trans-
ferred (1 or .1). Owing to an error in the program, it was possible to
enter the age of a patient incorrectly, which unfortunately occurred for
two patients. The age of these two patients was entered in such a way
that made it look much higher than their actual age. The randomiza-
tion program tried to balance for these two incorrect ages, resulting
in a skewed distribution of patients according to age. This error was
discovered at the end of the study, when performing the analysis of
the results. It seems natural, when performing clinical trials, to trust
a randomization program. However, errors can take place and the
lesson to be learnt from this study is to perform interim controls:
not necessarily by interim analyses, but to check that the randomiza-
tion has been performed correctly. If this had been done in the
present study, the error would probably have been noticed earlier
and the randomization program corrected.
In order to remedy the error as far as possible, we contacted two
well-recognized and independent statisticians for advice on how to
handle the problem. Both statisticians came to the same conclusion:
the results should be correct provided that a stratified analysis regard-
ing age as a continuous variable was performed. Another weakness
of this trial was that the recruitment of patients was stopped prema-
turely, before the total estimated number of patients had been allo-
cated. As such, the results should be viewed with some caution.
Luteal support after IVF/ICSI
In conclusion, equality of treatments in terms of ongoing pregnancy
rates between vaginal progesterone gel and vaginal micronized proges-
terone tablets could not be demonstrated even though the results do
not differ substantially. Vaginal progesterone gel provides a higher
patient convenience. Finally, when performing clinical trials, one
should be aware that errors in the randomization program may
occur, which underlines the importance of adequate control systems.
Acknowledgements
We thank Hans Wedel, Professor of Epidemiology and Biostatistics,
The Nordic School of Public Health, Göteborg, Sweden, and Bernhard
Huitfeldt, BH Statistical Consulting, Stockholm, Sweden, for valuable
statistical advice concerning the error in randomization, and Kerstin
Wiklund, statistician, Pharma Consulting group, Stockholm, Sweden,
for performing the statistical analysis. The randomization program,
however, was not made by the Pharma Consulting group, but by
another company. We also thank the participating patients, the
Nordic Crinone study group and the clinical staff. Finally, we thank
Merck Serono for supporting the study. Support was given by a
grant and providing the study drug. Merck Serono had no influence
on the design of the study and was not involved in the analysis of
the results or preparation of the manuscript.
Authors’ roles
C.B.: design of the study, acquisition of data, analysis and interpret-
ation of data; drafting the article; final approval of the article. S.L.:
design of the study, acquisition of data, analysis and interpretation of
data; revising the article; final approval of the article.
Funding
Financial support was provided by Merck Serono, Denmark and Sweden.
Conflict of interest
C.B. has earlier received research grants from Merck Serono and
Ferring. S.L. has previously accepted unrestricted research grants
from Merck Serono and Organon.
References
Lan VT, Taun PH, Canh LT, Tuong HM, Howles C. Progesterone
supplementation during cryopreserved embryo transfer cycles: efficacy
and convenience of two vaginal formulations. Reprod Biomed Online
2008;17:318 – 323.
Ludwig M, Schwartz P, Babahan B, Katalinic A, Weiss JM, Felderbaum R,
Al-Hasani S, Diedrich K. Luteal phase support using either Crinone
8% or Utrogest: results of a prospective, randomized study. Eur
J Obstet Gynecol Reprod Biol 2002;103:48 – 52.
Macklon NS, Fauser BC. Impact of ovarian hyperstimulation on the luteal
phase. J Reprod Fertil Suppl 2000;55:101– 108.
Ng EH, Miao B, Cheung W, Ho PC. A randomized comparison of side
effects and patient convenience of two vaginal progesterone
formulations used for luteal support in in vitro fertilization cycles. Eur
J Obstet Gynecol Reprod Biol 2003;111:50 – 54.
3473
Pabuccua R, Akara ME. Luteal phase support in assisted reproductive
technology. Curr Opin Obstet Gynecol 2005;17:277– 281.
Pocock SJ, Simon R. Sequential treatment assignment with balancing for
prognostic factors in the controlled clinical trial. Biometrics 1975;31:103–115.
Polyzos NP, Messini CI, Papanikolaou EG, Mauri D, Tzioras S, Badawy A,
Messinis I. Vaginal progesterone gel for luteal phase support in IVF/ICSI
cycles: a meta-analysis. Fertil Steril 2010;94:2083– 2087.
Probst AM, Hill JA, Ginsburg ES, Hurwitz S, Politch J, Yanushpolsky EH. A
randomized study comparing Crinone 8% and intramuscular
progesterone supplementation in in vitro fertilization-embryo transfer
cycles. Fertil Steril 2001;76:1144 – 1149.
Silverberg KM, Vaughn TC, Hansard LJ, Burger NZ, Minter T. Vaginal
(Crinone 8%) gel versus intramuscular progesterone in oil for luteal
phase support in in vitro fertilization: a large prospective trial. Fertil
Steril 2012;97:344 –348.
Simunic V, Tomic V, Tomic J, Nizic D. Comparative study of the efficacy
and tolerability of two vaginal progesterone formulations, Crinone 8%
gel and Utrogestan capsules, used for luteal support. Fertil Steril 2007;
87:83– 87.
van der Linden M, Buckingham K, Farquhar C, Kremer JA, Metwally M.
Luteal phase support for assisted reproduction cycles. Cochrane
Database Syst Rev 2011; Oct 5:CD009154.
Yanushpolsky E, Hurwitz S, Greenberg L, Racowsky C, Hornstein M.
Crinone vaginal gel is equally effective and better tolerated than
intramuscular progesterone for luteal phase support in in vitro
fertilization-embryo transfer cycles: a prospective randomised study.
Fertil Steril 2010;94:2596– 2599.
Yanushpolsky E, Hurwitz S, Greenberg L, Racowsky C, Hornstein M.
Patterns of luteal phase bleeding in in vitro fertilization cycles
supplemented with Crinone vaginal gel and with intramuscular
progesterone—impact of luteal estrogen: prospective, randomized
study and post hoc analysis. Fertil Steril 2011;95:617 – 620.
Zarutskie PW, Phillips JA. A meta-analysis of the route of administration of
luteal phase support in assisted reproductive technology: vaginal versus
intramuscular progesterone. Fertil Steril 2009;92:163 – 169.
Appendix
Investigators and study centres
Denmark: Svend Lindenberg, Copenhagen Fertility Center (Principal
Investigator), Finn Hald, Regionshospitalet Braedstrup, Anders Nyboe
Andersen, Rigshospitalet, Copenhagen, Michael Aasted, Fertilitetsklinik-
ken Dronninglund, Jeanette Wullf Bogstad, Hvidovre Hospital, Sven
Skouby, Fertilitetskliniken Herlev Hospital, Anette Lindhard, Roskilde
Amtssygehus, Peter Humaidan, Skive Sygehus/Odense Universitetshos-
pital, Anders Faurskov, Faurskov Fertilitet og ultralyd, Sven Rex, Odense
IVF Klinik, Anne-Lis Mikkelsen, Holbaek Sygehus, Hans Krog, Dansk
Fertilitetsklinik, Per Emil Rasmussen, Odense Universitetshospital.
Sweden: Christina Bergh, Department of Obstetrics and Gynaecol-
ogy, Institute of Clinical Sciences, Gothenburg University, Reproduct-
ive Medicine, Sahlgrenska University Hospital, Gothenburg
(Co-ordinating Investigator), Ann Thurin-Kjellberg, Department of
Obstetrics and Gynaecology, Institute of Clinical Sciences, Gothenburg
University, Reproductive Medicine, Sahlgrenska University Hospital,
Gothenburg, Karin Rova, Karolinska University Hospital, Stockholm,
Leif Bungum, Skåne Univeristy Hospital, Malmö, Kjell Wånggren, De-
partment of Women’s and Children’s Health, Uppsala University,
Uppsala, Nils-Gunnar Solensten, Umeå University Hospital, Umeå.