A prospective randomized control trial comparing medical and surgical treatment for early pregnancy failure

Constantinos Demetroulis, Ertan Saridogan,1, Dattakumar Kunde and Alan A. Naftalin,2

Department of Obstetrics and Gynaecology, Newham General Hospital, London, UK


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
A prospective randomized control trial was designed to assess the effectiveness of single dose, 800 µg misoprostol administered p.v. compared with surgical evacuation for the treatment of early pregnancy failure. A total of 80 women with a diagnosis of early pregnancy failure were randomized to study (vaginal misoprostol) and control (surgical curettage) groups. Success of treatment, side-effects as assessed during, immediately after and 10 days after treatment, and patient satisfaction were compared. Intravaginal misoprostol was successful in 82.5% (33 out of 40) of the patients. None of the control group patients required a repeat evacuation. The number of patients who experienced significant abdominal pain following treatment did not differ between the groups. The duration of pain was shorter in the control group; however, they required more analgesics during this short period. The number of patients with significant vaginal bleeding, the duration or severity of bleeding did not show any significant difference between the groups. All 33 patients in the study group who had successful treatment expressed satisfaction, whereas only 58% of the control group did so. In conclusion this randomized control study demonstrated the efficacy and safety of the administration of 800 µg of misoprostol p.v. for the management of early pregnancy failure.

Key words: medical treatment/miscarriage/misoprostol/surgical curettage


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Miscarriage is the most common gynaecological emergency. Of recognized pregnancies, ~15% miscarry spontaneously in the first trimester (Warburton and Fraser, 1959Go). The use of sensitive pregnancy tests and the accessibility of ultrasound have led to an increase in the number of diagnosed miscarriages. It accounts for ~50 000 inpatient admissions in the UK each year (El-Refaey et al., 1992).

Surgical evacuation is still the most commonly used treatment for miscarriage. Surgical curettage under anaesthesia accounts for almost three-quarters of emergency gynaecological operations performed in the UK (McKee et al., 1992Go). However, dilatation and suction evacuation of the uterus under anaesthesia has certain morbidity, such as the risk of anaesthesia, uterine perforation, intrauterine adhesions, cervical trauma, and infections leading to infertility, pelvic pain and increased chance of ectopic pregnancy. In addition, sometimes, the patients occupy beds while waiting for a convenient space on theatre lists, usually out of hours, remain nil by mouth for an uncertain period of time and are usually operated on by less well-trained staff. The alternatives to surgical treatment are either expectant management, or medical treatment. Expectant management of selected cases of miscarriage has been shown (Nielsen and Hahlin, 1995Go) to have a similar outcome to dilatation and curettage, whereas the results of another study (Jurkovic et al., 1998Go) showed that the success of expectant management of missed miscarriage was too low to justify its use in routine clinical practice. With expectant management there is great uncertainty as to when complete evacuation is to occur and the need for a prolonged follow-up is needed in order to identify this. This uncertainty can be quite stressful for most of the patients.

Medical treatment includes the use of prostaglandin analogues and mifepristone, an antiprogesterone. A prostaglandin analogue, misoprostol (15-deoxy,16-hydroxy,16-methyl analogue of prostaglandin E1) has particularly gained popularity for the management of miscarriage. However, some of the studies (Henshaw et al., 1993Go; de Jonge et al., 1995Go) included small numbers of patients, or the success rates were low in the others (de Jonge et al., 1995Go; Chung et al., 1995Go, 1999Go). Furthermore, the majority of the studies were not randomized control trials. In most studies misoprostol was used by the oral route (de Jonge et al., 1995Go; Chung et al., 1995Go, 1999Go) and there are indications (El-Rafaey et al., 1995Go; Zalanyi, 1998Go) that the vaginal use may be more effective. For this reason in the current study vaginal misoprostol was used as a single dose regime for the management of miscarriage in a prospective randomized control trial in order to compare it with surgical treatment. An 800 µg dosage was chosen as it was shown to be highly effective in inducing abortion in the first trimester (El-Rafaey et al., 1995Go).


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Protocol
Following appropriate ethical committee approval, women admitted to Newham General Hospital with a diagnosis of spontaneous incomplete or missed miscarriage or with an anembryonic sac at up to 13 weeks gestation were included. Initial diagnosis was made by clinical examination, urine pregnancy test and pelvic ultrasound examination.

Diagnosis of incomplete miscarriage was made when there was a history of passage of tissue and/or heterogeneous echogenic material in the uterine cavity with a thickness of >15 mm was identified at ultrasound examination. A missed miscarriage was diagnosed when an intrauterine gestation with a fetal pole measuring >6 mm was demonstrated and no heart movements were observed; whereas an anembryonic pregnancy was diagnosed when the diameter of the gestational sac was >20 mm and no fetal pole was visible. Full blood count and blood group were checked in all patients.

Women with complete miscarriage, high temperature (>37.5°C), low haemoglobin (<10.0 g/dl), history of serious medical or surgical condition, history of medical condition which is a contraindication to prostaglandin treatment (asthma, hypertension, glaucoma, sickle cell disease, mitral stenosis), and heavy bleeding requiring emergency evacuation of the uterus were excluded. Women under the age of 16 years and who were unable to give informed consent were also excluded.

Following the initial selection, patients who agreed to participate were randomized to study (vaginal misoprostol) or control (surgical curettage) groups. Study group patients were administered 800 µg of misoprostol vaginally. The vital signs and vaginal bleeding were monitored closely, and any side-effects were recorded. After 8–10 h the patients were assessed clinically and an ultrasound examination was performed. The time of passage of products of conception was not available since a high proportion of the patients noticed they had miscarried while asleep. Patients were discharged if their uterus was fully evacuated. The patients in whom the treatment failed were booked for surgical evacuation which was performed without ultrasound guidance using conventional methods to ensure complete evacuation. The control group patients were booked for surgical evacuation of the uterus.

All patients were seen 10 days later. Signs of bleeding, pain, infection or any other possible complications were checked. A repeat full blood count was carried out at this visit.

Sample size calculation and statistics
In the largest randomized control trial (Chung et al., 1999Go) the use of oral misoprostol showed a success rate of ~50%. It is accepted that the success rate of surgical evacuation would be ~95%. Therefore, a 20% improvement over oral misoprostol (70%) will give a 25% difference from surgical evacuation. Based on these numbers a trial with a power of 85% and an {alpha} of 0.05 would require a total sample size of 80, 40 patients in each arm (Altman, 1991Go).

All the variables were statistically compared using the {chi}2, Student's t- or Fisher's exact test. P < 0.05 was considered to be statistically significant.

Assignment
Individual patients were randomized to the study or control groups using sealed opaque envelopes containing numbers generated by computer, following informed consent. No attempt was made to conceal the intervention assignment schedule from the patients or clinicians as the treatment methods for the study and control were obviously different.

Masking
No attempt was made to mask the interventions as the study compared a medical treatment with a surgical procedure.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Participant flow and follow-up
A total of 249 patients were seen at Newham General Hospital for early pregnancy problems during the study period; 94 (37.7%) patients were found to be eligible for the study and 14 (14.9%) declined to participate in the trial and chose surgical evacuation (Figure 1Go). In all, 80 patients were enrolled in the study, 40 women in the study group and 40 women in the control group. Patients' age, gestational age, obstetric history and haemoglobin concentrations between the control and the study group patients were comparable (Table IGo). The distribution of incomplete and missed miscarriage and anembryonic pregnancies was similar in the study and control groups (Table IIGo).



View larger version (31K):
[in this window]
[in a new window]
 
Figure 1. Flow diagram.

 

View this table:
[in this window]
[in a new window]
 
Table I. Patient characteristics in the study (misoprostol treatment) and control (surgical evacuation) groups
 

View this table:
[in this window]
[in a new window]
 
Table II. Type of miscarriage in the study and control groups. Values in parentheses are percentages
 
In the study group the treatment was successful in 33 patients (82.5%) with seven patients (17.5%) requiring surgical evacuation due to failure of treatment. Of the seven patients in whom the treatment was not successful, two had an anembryonic sac, four had missed miscarriages (one was a twin pregnancy), and one had an incomplete miscarriage. In other words the failure rate in the study group was 7.1% (1/14) for incomplete miscarriage and 23.1% (6/26) for missed miscarriage and anembryonic sac. None of the control patients required a repeat evacuation (P = 0.005, Fisher's exact test).

The side-effects, analgesic and antiemetic requirements during and immediately after the treatment are given in Table IIIGo. No patient from either group required emergency evacuation, antibiotics, oxytocin or blood transfusion. In the control group the mean time elapsed from the time of decision to surgical evacuation was 14.1 h (range 2–31 h). Four patients (10%) from the study group did not attend the 10 day follow-up examination—one of these patients was contacted by telephone and declared satisfaction with the treatment. It was not possible to trace the other three. From the control group five (12.5%) did not attend and the three of these patients who were contacted declared no problems.


View this table:
[in this window]
[in a new window]
 
Table III. Side-effects during and immediately after treatment. Values in parentheses are percentages
 
The duration of pain and vaginal bleeding and the analgesic requirements at follow-up are given in Table IVGo. The number of patients who experienced significant abdominal pain following treatment did not differ between the groups and the duration of pain was shorter in the control group; however, they required more analgesics during this short period. The number of patients with significant vaginal bleeding, the duration of bleeding or the severity of bleeding did not show any significant difference between the groups. Post-treatment haemoglobin concentrations were also comparable. There were no patients with haemoglobin <10 g/dl at follow-up in the study group, but there were two with haemoglobin concentrations <10 g/dl in the control group. From these two, one patient had a haemoglobin of 11 g/dl and the other 10.3 g/dl pre-operatively.


View this table:
[in this window]
[in a new window]
 
Table IV. Side-effects declared at follow-up
 
Two patients from the study group and five patients in the control group visited their general practitioner between the treatment and the follow-up period. The two study group patients had vaginal bleeding and antibiotics were prescribed to them. Two of the five patients from the control group had offensive discharge and were also given antibiotics. Two had vaginal bleeding and antibiotics were prescribed by their general practitioner. One patient had abdominal pains which were treated with analgesics only.

All 33 patients in the study group who had a successful treatment and attended for follow-up expressed satisfaction with their treatment. Therefore, out of the 40 patients randomized to the study group 82.5% (33 out of 40) were satisfied. In contrast, from the control group of the 38 patients in whom the information regarding satisfaction with the treatment was available, only 22 (58%) were satisfied (P = 0.000007, Fisher's exact test). The main reason for dissatisfaction was the long waiting time for operation.


    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Medical management of miscarriage is becoming increasingly popular. Furthermore, there are also reports (Nielsen and Hahlin, 1995Go; Jurkovic et al., 1998Go) on expectant management with conflicting results.

In the past, success rates of between 13% (de Jonge et al., 1995Go) and 96% (Henshaw et al., 1993Go) have been reported following medical treatment of miscarriage. This large variation may be due to several factors including the patient selection, the concomitant use of mifepristone, the type of prostaglandin analogue, the dosage and route of administration. While some studies included only patients with missed miscarriage (Herabutya and O-Prasertsawat, 1997Go; Zalanyi, 1998Go) or incomplete miscarriage (de Jonge et al., 1995Go) others included all types of early pregnancy failures (Hughes et al., 1996Go; Creinin et al., 1997Go). In some studies mifepristone was given prior to prostaglandin administration (Hughes et al., 1996Go; Nielsen et al., 1997Go). Another study (Henshaw et al., 1993Go) included treatment with both sulprostone, a synthetic prostaglandin E2, and misoprostol. The dosage and route of administration of misoprostol varied between the studies. The lowest success rate was reported with the use of a single dose of 400 µg of misoprostol orally. When 400 µg of misoprostol was administered orally to a total of three times, the success rate was 62% in another study (Chung et al., 1995Go). Studies (Creinin et al., 1997Go; Zalanyi, 1998Go) with the use of vaginal misoprostol reported high success rates, similar to the success rates of the current study. However these studies included either a small number of patients or did not include a proper control group. A recent randomized control trial (Chung et al., 1999Go) using oral misoprostol 400 µg every 4 h up to a total dose of 1200 µg demonstrated a reduction in the number of surgical evacuations; however, the failure rate was ~50%.

The current study demonstrated highly successful results with the use of a single dose of vaginally administered misoprostol. Although the success rate was slightly lower than the 100% success rate of the surgical control group, patient satisfaction was significantly higher in the study group. Reasons for dissatisfaction in the control group were long waiting time for treatment and uncertainty of the time of operation. In contrast, the treatment was immediately provided in the study group, so that the patients were not left waiting for treatment. This may have been the main factor resulting in higher satisfaction rates in this group, together with reasonably high success rates.

A recent randomized trial comparing expectant with medical management for first trimester miscarriages showed that expectant management is almost as effective as the medical treatment and the success rates were similar to the rates found in the current study (Nielsen et al., 1999Go). This study suggested that the current medical treatment regimens do not improve the results over expectant management and future studies should concentrate on finding more effective alternatives. However, some patients may prefer to be actively treated rather then to wait for an unspecified period for a miscarriage to be completed, and this can be particularly unacceptable for a woman who knows she is carrying an `abnormal' pregnancy. For this reason there will always be a place for active management for miscarriage.

The duration between administration of misoprostol and intervention also seems to affect the outcome. A recent study (Chung et al., 1997Go), where up to 48 h were allowed before surgical intervention, showed a better success rate (69.6%) compared with their previous study (62.5%), where a 24h protocol was used. In the current study the intervention time was limited to 8–10 h in order to avoid patients becoming frustrated due to a delay in successful outcome. It is quite possible that success rates can be increased further by delaying the intervention time.

In this study the failure rate was higher for missed miscarriage/anembryonic pregnancy compared to incomplete miscarriage in the study group (23.1 versus 7.1%). However, the numbers are too small to reach a meaningful conclusion.

Previous studies as well as the current study have demonstrated the safety of administering misoprostol both by the oral or the intravaginal route. By the intravaginal route it is possible that the gastrointestinal side-effects are minimized. The side-effects are minimal and most of them are inconsequential. In the current study, the comparison of the short-term side-effects did not show any significant difference between the study and the control groups, except the presence of nausea which was significantly more common in the control group (P < 0.001).

High success rates as well as low side-effects of this approach seems to lead to a high satisfaction rate with this treatment. The lower satisfaction rate in the surgical control group was mainly due to the delay in their treatment. Medical treatment has the advantage of avoiding this delay and probably significantly contributes to the higher satisfaction rate.

Misoprostol is a cheap substance and easy to store. This allows a significant reduction in the cost of management of miscarriage by avoiding anaesthetic and theatre costs. It was demonstrated (Hughes et al., 1996Go) that the average cost of surgical treatment was £397 compared with £347 for medical treatment. The safety of misoprostol suggests that it can possibly be used as an outpatient treatment without admission to the hospital, which will reduce the costs further. However, larger studies will need to confirm its safety for this purpose.

In conclusion, this randomized control study demonstrated the efficacy and safety of the administration of 800 µg of misoprostol intravaginally for the management of missed, incomplete and anembryonic miscarriages. Larger studies are needed to compare this finding before this type of treatment without the hospitalization of patients can be considered as an outpatient treatment.


    Notes
 
1 Present address: Department of Obstetrics and Gynaecology, University College London Hospitals, London, UK Back

2 To whom correspondence should be addressed at: Newham General Hospital, Glen Road, Plaistow, London E13 8SL, UK.E-mail: anaftalin{at}zetnet.co.uk Back


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Altman, G.D. (1991) Practical Statistics for Medical Research. Chapman & Hall, London, pp. 455–457.

Chung, T.K.H., Cheung, L.P., Leung, T.Y. et al. (1995) Misoprostol in the management of spontaneous abortion. Br. J. Obstet. Gynaecol., 102, 832–835.[ISI][Medline]

Chung, T., Leung, P., Cheung, L.P. et al. (1997) A medical approach to management of spontaneous abortion using misoprostol. Extending misoprostol treatment to a maximum of 48 hours can further improve evacuation of retained products of conception in spontaneous abortion. Acta Obstet. Gynecol. Scand., 76, 248–251.[ISI][Medline]

Chung, T., Lee, D.T.S., Cheung, L.P. et al. (1999) Spontaneous abortion: a randomized, controlled trial comparing surgical evacuation with conservative management using misoprostol. Fertil. Steril., 71, 1054–1059.[ISI][Medline]

Creinin, M.D., Moyer, R. and Guido, R. (1997) Misoprostol for medical evacuation of early pregnancy failure. Obstet. Gynecol., 89, 768–772.[Abstract/Free Full Text]

de Jonge, E.T.M., Makin, J.D., Manefeldt, E. et al. (1995) Randomised clinical trial of medical evacuation and surgical curettage for incomplete miscarriage. Br. Med. J., 311, 662.[Free Full Text]

El-Rafaey, H., Hinshaw, K., Henshaw, R. et al. (1992) Medical management of missed abortion and anembryonic pregnancy. Br. Med. J., 305, 1399.[ISI][Medline]

El-Rafaey, H., Dhamnasekar, R., Abdalla, M. et al. (1995) Induction of abortion with mifepristone (RU 486) and oral or vaginal misoprostol. N. Engl. J. Med., 332, 983–987.[Abstract/Free Full Text]

Henshaw, R.C., Cooper, K., El-Refaey, H. et al. (1993) Medical management of miscarriage: non-surgical uterine evacuation of incomplete and inevitable spontaneous abortion. Br. Med. J., 306, 894–895.[ISI][Medline]

Herabutya, Y. and O-Prasertsawat, P. (1997) Misoprostol in the management of missed abortion. Int. J. Gynecol. Obstet., 56, 263–266.[ISI][Medline]

Hughes, J., Ryan, M., Hinshaw, K. et al. (1996) The costs of treating miscarriage: a comparison of medical and surgical management. Br. J. Obstet. Gynaecol., 103, 1217–1221.[ISI][Medline]

Jurkovic, D., Ross, J.A. and Nicolaides, K.H. (1998) Expectant management of missed miscarriage. Br. J. Obstet. Gynaecol., 105, 670–671.[ISI][Medline]

McKee, M., Priest, P., Ginzlet, M. et al. (1992) Can out-of-hours operating in gynaecology be reduced? Arch. Emerg. Med., 9, 290–298.[ISI][Medline]

Nielsen, S. and Hahlin, M. (1995) Expectant management of first-trimester spontaneous abortion. Lancet, 345, 84–88.[ISI][Medline]

Nielsen, S., Hahlin, M. and Platz-Christensen, J.J. (1997) Unsuccessful treatment of missed abortion with a combination of an antiprogesterone and a prostaglandin E1 analogue. Br. J. Obstet. Gynaecol., 104, 1094–106.[ISI][Medline]

Nielsen, S., Hahlin, M. and Platz-Christensen, J. (1999) Randomized trial comparing expectant with medical management for first trimester miscarriages. Br. J. Obstet. Gynaecol., 106, 804–807.[ISI][Medline]

Warburton, D. and Fraser, F.C. (1959) Genetic aspects of abortion. Clin. Obstet. Gynecol., 2, 22–35.[ISI][Medline]

Zalanyi, S. (1998) Vaginal misoprostol alone is effective in the treatment of missed abortion. Br. J. Obstet. Gynaecol., 105, 1026–1035.[ISI][Medline]

Submitted on June 8, 2000; accepted on October 23, 2000.