Missed abortion is a common complication of early pregnancy occurring in up to 15% of all clinically recognized pregnancies. 1 The majority of cases are currently treated by dilation and curettage (D&C). The rationale that all spontaneous abortions should be treated with D&C to prevent infection and hemorrhage has been questioned, 2 and several studies have reported success with expectant management of incomplete abortion. 3,4 Treatment of incomplete abortion with misoprostol has also been reported with varying degrees of success. 5–8 With the rising use of early ultrasound, an increasing number of miscarriages present as missed abortions before the onset of cramping and bleeding. A small case series reported seven of eight women with missed abortion had a complete abortion after treatment with vaginal misoprostol compared with three of 12 treated by the oral route. 9 We decided to investigate this further in a randomized trial in subjects with missed abortion to determine what the rate of complete abortion and D&C would be compared with expectant management.
MATERIALS AND METHODS
Subjects were approached for enrollment if they had an ultrasound diagnosis of a nonviable pregnancy and were not experiencing uterine cramping or bleeding. Standard ultrasound criteria for missed abortion were used. One of the following was necessary for a diagnosis of a nonviable pregnancy: embryo greater than 7 mm with no embryonic cardiac activity, irregular gestational sac with mean sac diameter greater than 16 mm2, or a gestational sac greater than 15 mm with no visible fetal pole. Subjects with active vaginal bleeding, cramping, dilatation of the internal os, or with a nonviable embryo that measured greater than a 12-week size were excluded. However, women who were experiencing light spotting, without cramping, and had a closed internal os were eligible. Subjects experiencing a steady flow of blood were excluded.
After obtaining informed consent, subjects were randomly allocated to treatment with misoprostol 800 mg per vagina or to placebo. Randomization was accomplished by a computer-generated random number list with subjects randomized into varying blocks of sizes 4 to 8. According to the randomization schedule, pharmacy staff placed the misoprostol or placebo tablets into numbered envelopes. The investigators were not aware of the randomization schedule, and the envelopes were not handled or opened until after randomization. As the misoprostol and placebo tablets were not identical, additional precautions were taken to maintain allocation concealment. After the clinical assessment, the study nurse placed the pills in an opaque vaginal introducer, which the physician then used to insert the tablets. The subjects were provided with acetaminophen and combined acetaminophen/codeine tablets for analgesia. They were instructed to use one or two of these tablets every 4 hours as needed. All Rhogam-negative women received Rhogam 300 μg intramuscularly. Subjects were also asked to complete a patient satisfaction questionnaire and a symptom log. An information sheet was provided with instructions to return to the hospital if heavy bleeding occurred. Heavy bleeding was defined as saturating more than one heavy pad every hour for more than 2 hours or more than one heavy pad per 30 minutes for more than 1 hour.
A container was also provided for any products of conception that the subjects were able to retrieve. Baseline hemoglobin and serum β-human chorionic gonadotropin (βhCG) levels were obtained. The serum βhCG was repeated at 48 hours and the hemoglobin at 1 week. Follow-up was arranged at 24 hours, 48 hours, and at 1 week. Speculum and bimanual examinations were performed at each visit, and any tissue passed by the subjects was examined. If complete abortion was not suspected after 24 hours, the medication was repeated. At 48 hours, if there had been no response to the medication or an incomplete abortion was suspected, the subjects were offered D&C. Transvaginal ultrasound was used in cases of suspected incomplete abortion as clinically appropriate. An ultrasound finding of a focal hyperechoic intrauterine mass was considered sufficient for diagnosis of incomplete abortion. All subjects with a clinically suspected complete abortion were instructed to have urine βhCG test after 4 weeks. Pathology reports for all tissue submitted for examinations were reviewed.
The subjects were also asked to complete a post-treatment questionnaire. Patient satisfaction was assessed by asking the subjects to rate their degree of agreement with two statements: 1) I would recommend the treatment with the vaginal tablets to a friend or family member who had a missed abortion. 2) I would try treatment with the vaginal tablets again if I had another missed abortion. The subjects indicated their degree of agreement with the statements on a five-point scale: Strongly Disagree, Disagree, Neutral, Agree, or Strongly Agree. The primary outcomes were rates of complete abortion and D&C. Complete abortion was defined as expulsion of the products of conception without D&C and a negative follow-up urine βhCG test at 4 weeks or as the absence of products of conception in the surgical specimen from D&C in subjects who had suspected incomplete abortion. The rates of complete abortion and D&C were compared and relative risks and confidence intervals calculated. Statistical significance was determined with a Fisher exact test, and analysis was by intention to treat. Also, t tests were used to compare continuous variables with log transformations where appropriate. A sample size calculation had estimated that 25 subjects in each group would be necessary to achieve 80% power to detect an estimated reduction in the rate of D&C from 50% in the placebo group to 10% in the treatment group. The study protocol was approved by the University of Calgary Ethics Review Board.
Between February 1999 and April 2000, 50 women were enrolled in the study. One half of the subjects received misoprostol and one half placebo. Their gestational age ranged from 7 weeks to 17 weeks (median 12 weeks). The diameter of the gestational sac on ultrasound ranged from 1.4 cm to 7.5 cm. The demographic and clinical characteristics were similar between the two groups (Table 1). All of the subjects completed the study protocol. Within 24 hours of the first dose, all of subjects in the misoprostol group (25 of 25) and 20% (five of 25) of the subjects in the placebo group reported uterine cramping and bleeding.
In the misoprostol group, 15 of 25 appeared to have aborted completely after the first dose. On examination at 48 hours after enrollment, 21 of 25 of the misoprostol group appeared to have aborted completely, three of 25 had aborted incompletely, and one subject was unchanged. Typically, patients reported the bleeding and cramping would reach a peak with the expulsion of a sac or tissue from the uterus and then rapidly diminish. One subject who appeared to have aborted completely passed some necrotic products of conception on day 14 after treatment. Two subjects had ongoing bleeding and had an ultrasound, which suggested retained products of conception. Both had D&Cs, but no products of conception were documented on histologic examination of the surgical specimen. Only one subject who appeared to have completely aborted had a positive pregnancy test 4 weeks after treatment. A D&C was performed, and a small amount of trophoblastic tissue was removed. One subject had heavy bleeding and had an emergency D&C. Her hemoglobin dropped from 141 g/L on day 1 to 129 g/L on day 7.
In the placebo group, four subjects aborted completely, one by 48 hours and three after 1 week. Two subjects had incomplete abortions, and the remainder reported no change with treatment.
Overall, the rate of D&C was 28% (seven of 25) in the misoprostol group and 84% (21 of 25) in the placebo group, relative risk 0.33 (0.17, 0.64), P < .001. The rate of complete abortion was 80% (20 of 25) in the misoprostol group and 16% (four of 25) in the placebo group, relative risk 0.20 (0.08, 0.50), P < .001.
The laboratory data were also analyzed and compared between the two treatment groups (Table 2). Baseline and post-treatment (day 7) hemoglobin levels were obtained in 19 of 25 subjects in the placebo group and 20 of 25 subjects in the misoprostol group. The change in hemoglobin between day 1 and day 7 was not significantly different between the two groups: misoprostol 3.2 g/L versus 4.3 g/L in the placebo group, P = .72. Nine subjects had a decrease in hemoglobin concentration of more than 10 g/L, five were in the misoprostol group, and 4 were in the placebo group. No subject required a blood transfusion. The mean serum βHCG levels, at presentation, were higher in the subjects who had a complete abortion, 24,518 IU/L versus 15,674.5 IU/L. However, after logarithmic transformation of the data, the difference was not statistically significant (P = .29). Likewise, the mean change in βHCG levels, from baseline to 48 hours after treatment, was greater in the group that completely aborted, 18,652 IU/L versus 7768 IU/L, but the difference was not statistically significant (P = .39). The gestational sac size was also similar between those who completely aborted and those who did not, 3.16 mm versus 3.48 mm (P = .60).
Vaginal misoprostol treatment appeared to be well tolerated. Only one patient reported significant gastrointestinal side effects. All but two subjects required only acetominophen with codeine for analgesia. One subject had a D&C because of excessive pain, and one presented to the emergency room and required an intramuscular injection of narcotic. Twenty-one subjects completed and returned the post-treatment questionnaire. Four subjects did not return the questionnaire; three of these subjects failed misoprostol treatment and had a D&C. Overall, patient satisfaction with the treatment appeared to be high. For the statement “I would try treatment with misoprostol again if I had another missed abortion,” 19 of 21 agreed or strongly agreed, one subject was neutral, and one disagreed. For the statement “I would recommend misoprostol treatment to friends or family with a missed abortion,” 18 of 21 agreed or strongly agreed, and three subjects were neutral.
Obstetrician/gynecologists have recently been challenged to rethink their approach to miscarriage. The doctrine of prompt surgical evacuation has been challenged by a small number of studies in women with incomplete abortions. Furthermore, the high success rates (greater than 90%) in the medical therapeutic abortion also suggest nonsurgical treatment should be considered. 10 However, in North America, medical termination requires treatment with methotrexate injections, which is not available in the average practitioner's office and may not be as acceptable to patients as misoprostol alone. Furthermore, medical termination studies typically limited enrollment to women with less than 49 days of amenorrhea because of concerns of excessive bleeding.
Our results suggest that two vaginal doses of 800 mg of misoprostol alone are effective in producing complete abortion in approximately 80% of subjects and significantly reduce the need for D&C. Although 28% of our subjects had a D&C, this may be reduced with further experience, as two of the subjects had no identifiable products of conception at the time of surgery. Also, one subject dilated to 2 cm but did not expel the pregnancy sac after two doses of misoprostol. In a more flexible setting, outside of a randomized clinical trial, further doses could have been given and it is likely the sac would have been expelled. There may be additional benefits even if the treatment is not completely successful. All of the subjects who had a D&C in the treatment group had a dilated cervix at the time of surgery, which intuitively should reduce the risk of perforation and cervical lacerations. The treatment also appeared to be well tolerated. The vast majority of subjects required only codeine and acetominophen for analgesia, and only one D&C was performed because of excessive pain. The initial data also suggest that medical management of missed abortion is safe. There was no occurrence of serious life-threatening bleeding, and no subject required a transfusion. Only one D&C in the study group was performed for significant bleeding. However, a sample of 25 subjects may be insufficient to detect rare serious side effects. Ultimately, we hope to perform a large prospective cohort study of women requesting medical management of missed abortion. We anticipate that medical management with vaginal misoprostol will prove to be a viable alternative to D&C for women with missed abortion.
1. Daya S. Habitual abortion. In: Copeland LJ, ed. Textbook of gynecology. Philadelphia, PA: WB Saunders; 2000: 227–71.
2. Aucott Ballagh S, Harris HA, Demasio K. Is curettage needed for uncomplicated incomplete spontaneous abortion? Am J Obstet Gynecol 1998;179:1279–82.
3. Chipchase J, James D. Randomised trial of expectant versus surgical management of spontaneous miscarriage. Br J Obstet Gynaecol 1997;104:840–1.
4. Nielsen S, Hahlin M. Expectant management of first-trimester spontaneous abortion. Lancet 1997;345:84–6.
5. Chung T, Leung P, Cheung LP, Haines C, Chang AM. 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 1997;76:248–51.
6. Chung TKH, Cheung LP, Leung TY, Haines CJ, Chang AMZ. Misoprostol in the management of spontaneous abortion. Br J Obstet Gynaecol 1995;102:832–5.
7. de Jong EM, Makin JD, Manefeldt E, De Wet GH, Pattinson RC. Randomised trial of medical evacuation and surgical curettage for incomplete miscarriage. Br Med J 1995;311:662.
8. Henshaw RC, Cooper K, El-Refaey H, Smith NC, Templeton AA. Medical management of miscarriage: Nonsurgical uterine evacuation of incomplete and inevitable spontaneous abortion. Br Med J 1993;306:894–5.
9. Creinin M, Moyer R, Guido R. Misoprostol for medical evacuation of early pregnancy failure. Obstet Gynecol 1997;89:768–72.
10. Grimes DA. Medical abortion in early pregnancy: A review of the evidence. Obstet Gynecol 1997;89:790–6.