Immobilisation versus immediate mobilisation after intrauterine insemination: randomised controlled trial

Methods

Subfertile women between 18 and 43 years of age with an indication for treatment with intrauterine insemination were eligible for the trial. Couples using donor semen (fresh or cryopreserved) could also be included in the trial. We made no restrictions with regard to the use and type of controlled ovarian hyperstimulation during treatment cycles.

All couples had been investigated for infertility according to the guidelines of the Dutch Society of Obstetrics and Gynaecology.6 This included a medical history, cycle monitoring, semen analysis, postcoital test, and assessment of tubal patency. The woman’s age, duration of subfertility, and whether subfertility was primary or secondary were documented. We defined duration of subfertility as the time from when the couple started actively trying to conceive to the time of start of treatment. If the couple had a previous pregnancy that had not resulted in a live birth, we defined duration of subfertility as the time from the first day of the pregnancy to the time of start of treatment. We defined primary subfertility as the absence of pregnancy in the current relationship.

If cryopreserved donor sperm was used, we defined subfertility as at least 12 cycles of unsuccessful intracervical insemination before intrauterine insemination. Ovulation was confirmed by basal body temperature curve, midluteal serum progesterone, or sonographic monitoring of the cycle. We included anovulatory women in the trial only after ovulation had been induced for at least six to 12 months without conception or if a male factor was also present, as in these instances an indication for intrauterine insemination existed.

At least one well timed postcoital test was done (except in couples using cryopreserved donor sperm) during the basic assessment of fertility. The test was planned according to the basal body temperature curve or findings of ultrasonography. A cervical factor was diagnosed if no progressive spermatozoa were seen in five high power fields at 400 times magnification and the total motile sperm count was less than 10×10⁶ spermatozoa/ml. Tubal pathology was assessed by a chlamydia antibody test, a hysterosalpingogram, or laparoscopy. In the case of a positive chlamydia antibody test, the tubal status was subsequently evaluated with a hysterosalpingogram or laparoscopy; in women with a negative chlamydia antibody test, tubal pathology was considered to be absent. Patients had to have at least one patent tube to be eligible for the study. We defined male subfertility as total motile sperm count less than 10×10⁶ spermatozoa/ml and unexplained subfertility as total motile sperm count more than 10×10⁶ spermatozoa/ml and exclusion of a cervical factor.

Controlled ovarian hyperstimulation, semen preparation, and insemination regimens were done according to hospital specific protocols. Controlled ovarian hyperstimulation was done with clomiphene citrate 50-150 mg on days five to nine of the cycle or subcutaneous injections of recombinant or urinary follicle stimulating hormone daily (Gonal F, Serono Benelux, The Hague, Netherlands; Puregon, Organon, Oss, Netherlands; or Menopur, Ferring, Hoofddorp, Netherlands). Controlled ovarian hyperstimulation was primarily done with recombinant follicle stimulating hormone in all clinics but one, where clomiphene citrate was used as a first line treatment. Ovulation was induced with 5000 IU or 10 000 IU of human chorionic gonadotrophin (Pregnyl, Organon), and women were inseminated 36-40 hours later. If more than three dominant follicles (>16 mm) were present, the treatment cycle was cancelled. Semen samples were processed within one hour of ejaculation by density gradient centrifugation followed by washing with culture medium. The volume of semen that was inseminated varied between 0.2 ml and 1.0 ml.

Patients were asked to participate before start of the first intrauterine insemination cycle. After giving written informed consent, the couples were randomly assigned to have three cycles of intrauterine insemination followed by 15 minutes of immobilisation (intervention group) or three cycles of intrauterine insemination with immediate mobilisation (control group). We randomised the couples before the first insemination, by using a web based computer program with a stratification procedure for age (18-34 years and 35-43 years) and centre. Women were inseminated in the lithotomy position in a Trendelenburg tilt. Depending on their allocation, women remained in the supine position for 15 minutes (timed by an alarm clock) or were mobilised immediately.

The primary outcome measure was the occurrence of an ongoing, viable intrauterine pregnancy (within four months after randomisation), defined as fetal heart beat seen by transvaginal ultrasonography at 12 weeks’ gestation. Secondary outcomes included live birth, biochemical pregnancy, ectopic pregnancy, and miscarriage. Pregnancy was determined by a qualitative urine test for β human chorionic gonadotrophin if no menstruation occurred 14 days after insemination.

Assuming an ongoing pregnancy rate of 10% per cycle in the mobilisation group, we believed that an increase in the ongoing pregnancy rate from 10% to 14% per cycle would be relevant. This corresponds to a 12% difference after three cycles. As expecting that 15 minutes of immobilisation would perform worse than immediate mobilisation would not be logical, we used one sided statistical tests. Using an α error of 0.05 and a β error of 0.20, and assuming a dropout rate of 10%, we needed 185 couples in each arm.

We calculated the rates of ongoing pregnancy per couple in each group and the corresponding relative risk with 95% confidence intervals. We used a two tailed Fisher’s exact test to test for significance. We did stratified analyses for different subgroups and used Kaplan-Meier analysis to calculate time to pregnancy. We initially analysed data according to the intention to treat principle and followed this with a per protocol analysis.