Sublingual versus vaginal misoprostol for cervical dilatation 1 or 3 h prior to surgical abortion: a double-blinded RCT.

Can sublingual administration of misoprostol 1 h prior to vacuum aspiration be more effective than vaginal administration and as effective as either route three 3 h prior to surgery? Sublingually administered misoprostol is superior to vaginally administered misoprostol when given 1 h pre operatively, and it is as effective as after a three 3 h priming interval with either route of administration. Misoprostol reduces complications and morbidity when used for cervical priming prior to surgical dilatation and vacuum aspiration in first trimester pregnancy. Despite the widespread use and extensive studies, the optimal route of administration of misoprostol before surgical abortion remains to be defined. The optimal priming interval after vaginal and sublingual administration of 400 mcg misoprostol has been reported to be 3 h. A longer interval will not improve dilatation but will increase the risk for bleeding and expulsion of the uterine contents before surgical evacuation. The pharmacokinetic properties of misoprostol indicate that sublingual compared with vaginal administration of misoprostol may result in a more rapid cervical priming effect. Women were randomized to four treatment groups and received 400 mcg misoprostol sublingually, or vaginally, 1 or 3 h prior to surgery. The study was a double-blinded RCT with regard to route of misoprostol administration but not the timing interval. The primary outcome was baseline cervical dilatation after misoprostol priming. The study was conducted between June 2007 and March 2014 and 184 women aged 18 years or older were recruited. Women were recruited among nulliparous women undergoing elective surgical first trimester abortion. Exclusion criteria were any contraindication for misoprostol, untreated genital infection, previous history of surgery to the cervix, or abnormal pregnancy. Gestational age was established by endovaginal ultrasound examination. The trial was conducted in a university hospital outpatient clinic. The allocated medication (misoprostol and placebo) was self-administered 1 h or 3 h prior to surgery. All women received 2 tablets of 200 mcg misoprostol and 2 identical looking placebo tablets. Prophylactic pain medication, 100 mg oral diclofenac, was administered at the time of misoprostol. Side effects were recorded immediately before surgery and women were asked which administration route of administration they found most convenient and which they would have preferred. The exact priming time (from misoprostol administration to initiation of dilatation) and signs of bleeding prior to dilatation were recorded. Vacuum aspiration was performed under general anaesthesia according to clinical routine. Dilatation was performed using tapered Pratt-dilatators and the resistance of the cervix was assessed objectively using a tonometer. All surgery was performed by two investigators, experienced in using the tonometer. The cumulative force required to dilate the cervix was calculated by adding the peak force needed for each dilatator up to 9.7 mm. The time needed for surgery including cervical dilatation and vacuum aspiration, was recorded. Intra-operative blood loss was measured and any surgical complications noted. Six women were excluded retrospectively from the analysis. Multivariate analysis of the primary outcome baseline dilatation showed a significant influence on route of administration (P = 0.034, 95% confidence interval (CI) -2.202, -0.086) as well as the interaction variable between route of administration and total priming time (P = 0.042, 95% CI 0.00, 0.016), with the vaginal route becoming more effective with longer priming time. These factors also had a significant influence on the peak force (administration route P = 0.042, 95% CI 0.221, 12.427, interaction P = 0.049, 95% CI -0.089, 0.000) and cumulative force (administration route P = 0.023, 95% CI 3.142, 40.877, interaction P = 0.026, 95% CI -0.293, -0.019) used for dilatation. The total priming time had a significant influence on bleeding before surgery, with more women bleeding the longer the total priming time (P = 0.003, 95% CI 2.203, 49.706). For abdominal pain before surgery there was a significant influence of administration route (P = < 0.001 95% CI 0.028, 0.235) and the interaction variable between administration route and priming time (P = 0.003, 95% CI 2.005, 30.757) with more women in the sublingual group experiencing abdominal pain the longer the priming time. The groups did not differ regarding duration of surgery, amount of bleeding and rate of side effects, such as nausea and shivering. Women in our study preferred vaginal treatment, as they disliked the taste of the misoprostol tablets. Vaginal treatment was also perceived as quicker to administer (P = 0.0001). The cervical tissue has viscoelastic properties, i.e. tissue resistance to mechanical dilatation depends also on the rate at which dilatation is performed. The ideal measurement of dilatation force should therefore also record the rate and time of dilatation. To ensure comparability, only nulliparous women without prior cervical surgery were recruited. In addition, time of dilatation was recorded and did not differ between the groups, and it is therefore assumed that dilatation took place at approximately the same rate. A limitation is that the study was conducted over a long time period because there was only one tonometer, decreasing numbers of surgical abortions and the fact that the main author was on a rotation schedule. In addition, the study was not powered to detect differences in side effects. Priming with misoprostol is recommended prior to surgical abortion. The priming interval of misoprostol may be reduced to 1 h after sublingual administration but not after vaginal administration. The results of the present study will increase choice and flexibility in cervical priming. The Swedish research council (521-2009-2605), Swedish Council for Working Life and Social Research (1404/08), Stockholm County Council and Karolinska Institutet (ALF 2009-2012). All authors declare that they have no conflicts of interest. www.clincaltrials.gov, NCT 01933360.