To investigate if self-inserted vaginal misoprostol prior to outpatient hysteroscopy will lead to satisfactory cervical ripening, compared to placebo.

In gynaecological practice, diagnostic and therapeutic hysteroscopy is one of the most common methods for diagnosing intrauterine pathology. The complications encountered during hysteroscopy, such as cervical tears, creation of false passages, and uterine perforation, are mainly related to the difficulty of cervical dilatation in nulliparous and postmenopausal patients. GnRH agonist use, previous cone biopsy, and markedly retroverted uterus are additional risk factors for complications.

Cervical ripening may be achieved either by mechanical means, such as with osmotic dilators, or biochemically with prostaglandins. Solid evidence supports the efficacy of misoprostol for cervical ripening before first-trimester suction curettage abortion. A consensus has emerged that the optimal treatment regimen to use is 400-μg misoprostol 3-4 hours pre-operatively, with vaginal administration being superior to the oral administration. However, two studies using 400-μg misoprostol via the oral and vaginal routes have previously shown that this does not result in satisfactory pre-operative dilatation in the majority of Norwegian women, and we have recommended that the minimum dosage be greater than 400-μg vaginal misoprostol fours hours pre-operatively.

The first trials describing prostaglandin priming of the cervix to reduce the frequency of traumatic cervical and uterine lesions in non-pregnant women prior to hysteroscopy were published in 1985.

Misoprostol for cervical priming in non-pregnant women is not well established. 16 published randomised controlled trials have shown different cervical response and outcomes. Most of the studies have separately, but not systematically compared the effect on different patient groups, such as nulliparous women and postmenopausal women. Seven of the studies included less than 50 patients, and almost all the trials were underpowered as regards to evaluating primary outcome measures. In the study with the largest number of patients included (Thomas et al), the authors have not explained why or how they converted the primary outcome measure from "the largest-number Hegar dilator that could be inserted without resistance" to a binary variable ("Hegar > 8/Hegar < 8) after they had analysed their results - an apparent protocol violation. It appears that none of the trials have been designed and conducted in accordance with the CONSORT statement. The dosages used in the studies have varied between 200 and 1000-μg misoprostol given between 2 and 24 hours before hysteroscopy, via the oral, sublingual and vaginal routes. Trials that have used higher dosages of misoprostol, via the vaginal route, using the longest interval between administration of misoprostol and hysteroscopy have tended to show more favourable outcomes as regards to cervical priming.

The hospital pharmacist will manufacture both active misoprostol ground up as a whitish powder in capsules (500-μg per capsule), as well as an inactive ingredient with an identical appearance (lactose) in capsules as placebo. The hospital pharmacist will prepare numbered, opaque, sealed envelopes. The envelopes will be numbered according to a randomisation list. The hospital pharmacy will then insert the prepared capsules into the envelopes. Each envelope will contain two capsules. Half of the envelopes will contain two capsules with 500-μg misoprostol each, while the other half will contain two placebo capsules.

All patients referred to outpatient hysteroscopy at Ullevål University hospital will be sent an invitation to be included in the study, together with dates for a prior out patient consultation and the hysteroscopy. The study invitation will include detailed information regarding the study, as well as an informed consent form. The out patient consultation will take place a few days prior to their scheduled hysteroscopy, where they will be given the option to participate in the study. The study authors will primarily be responsible for recruiting patients. During the study period, participant flow will be recorded.

If the patients are recruited, the gynaecologist at the pre-operative consultation will record the pre-operative variables on the case report form, and the patients will be given an envelope containing the capsules before leaving the hospital. Hence those involved in administering the intervention and assessing the outcomes will be blinded to the dosage, as well as the patients. Each participant will be randomised by opening a numbered, sealed, opaque envelope at home, containing either misoprostol, or an inactive ingredient with an identical appearance, in capsules. The design method of randomization and blinding should obviate potential subjectivity bias. The patients will be instructed to insert the capsules at 9 p.m. the evening before the operation.

Before the hysteroscopy, the gynaecologist will measure the pre-operative degree of cervical dilatation by passing Hegar dilators through the cervix in ascending order starting with a size two millimetres. The size of the largest dilator passed into the cervical os without subjective resistance felt by the operator will be recorded as the preoperative degree of dilatation. If there is resistance with Hegar dilator size two millimetres, the result will be recorded as < 2. Adverse events such as superficial cervical lacerations, false passage of the cervix during cervical dilatation, and perforation of the uterus will be recorded.

The patient records will be reviewed after 60 days to see if the women have contacted the hospital for post-operative complications.

A two sample sequential Wilcoxon test will be used to analyze the data. We performed a small pilot study in our department on 20 patients prior to hysteroscopy in January 2006 to investigate the pre-operative variability (SD) in cervical dilatation. The standard deviation was 1.3 mm in postmenopausal patients (n=5), and 1.4 mm in premenopausal patients (n=15). The mean cervical dilatation was 3.4 mm in postmenopausal patients and 5.4 mm in premenopausal patients. The design of the main trial is based on a significance level of 5% and a power of 95% against a 1 mm difference in the cervical dilatation caused by misoprostol and placebo. As it seems unlikely that use of misoprostol could cause a constriction of the cervix, a one-sided test is chosen. The trial will continue until either the null hypothesis or the alternative hypothesis has been refuted.

Inclusion Criteria:

• All patients who are referred to outpatient hysteroscopy, and who have given informed consent, will be eligible for study recruitment.

Exclusion Criteria:

• Women who are unable to communicate in Norwegian
• Women with a known allergy to misoprostol