International Journal of Gynecology and Obstetrics 129 (2015) 98–103
Contents lists available at ScienceDirect
International Journal of Gynecology and Obstetrics journal homepage: www.elsevier.com/locate/ijgo
REVIEW ARTICLE
Second-trimester postabortion care for ruptured membranes, fetal demise, and incomplete abortion Alice G. Mark a,⁎, Alison Edelman a,b, Lynn Borgatta c a b c
Ipas, Chapel Hill, NC, USA Department of Obstetrics and Gynecology, Oregon Health and Science University, Portland, OR, USA Department of Obstetrics and Gynecology, Boston University, Boston, MA, USA
a r t i c l e
i n f o
Article history: Received 29 May 2014 Received in revised form 5 November 2014 Accepted 9 January 2015 Keywords: Abortion Intrauterine fetal death Mifepristone Misoprostol Postabortion care Premature rupture of membranes Second trimester Unsafe abortion
a b s t r a c t Background: Guidance for postabortion care (PAC) is established for the first trimester but limited in the second trimester. Objectives: To establish evidence-based recommendations for PAC in the second trimester. Search strategy: Medline, POPLINE, and the Cochrane Central Register of Controlled Trials were searched with terms related to second-trimester PAC, including fetal demise, ruptured membranes, and incomplete abortion. The reference lists of retrieved articles were also searched. Selection criteria: Clinical trials and comparative studies of women presenting in the second trimester (12–28 weeks) were included if more than 50% of participants met PAC criteria or if outcomes for PAC were analyzed separately. Data collection and analysis: Data were extracted from included studies. When interventions in at least two articles were comparable, a meta-analysis was performed. Main results: Overall, 17 studies of 1419 women met inclusion criteria. Misoprostol given vaginally, sublingually, or buccally was associated with shorter expulsion times than was oral misoprostol. Additionally, 200 μg of misoprostol was more effective than lower doses. Pretreatment with mifepristone decreased expulsion time. Misoprostol was more effective than oxytocin. Conclusion: Misoprostol with or without mifepristone is an effective treatment for second-trimester PAC. The minimum misoprostol dose is 200 μg vaginally, sublingually, or buccally every 6–12 hours. © 2015 International Federation of Gynecology and Obstetrics. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
1. Introduction Postabortion care (PAC) is the comprehensive treatment of women presenting after spontaneous or induced, safe or unsafe abortion. PAC includes several key components: community and service provider partnerships, counseling, treatment of incomplete abortion with uterine evacuation, management of complications, contraception, and reproductive health services [1]. Following unsafe abortion, PAC services reduce maternal morbidity and mortality. Evidence-based guidelines [2,3] exist to support medical management and vacuum aspiration for incomplete abortion in women with a uterine size less than 13 weeks of pregnancy. However, many women in countries with restricted abortion access present later in pregnancy with fetal demise, ruptured membranes, retained placenta, hemorrhage, a foreign body, or infectious complications resulting from unsafe or self-induced abortion. These women could receive inconsistent care from different providers in obstetrics and gynecology wards or emergency departments. Magnitude studies from Sub-Saharan ⁎ Corresponding author at: PO Box 9990, Chapel Hill, NC, 27515, USA. Tel.: +1 919 960 5581; fax: +1 919 929 0258. E-mail address: [email protected] (A.G. Mark).
Africa and Southeast Asia show high rates of hospital presentation for PAC in the second trimester: 17% of women presenting for PAC in Cambodia are in the second trimester [4], whereas this rate is 35% in Malawi [5] and 41% in Kenya [6]. As the pregnancy duration increases, so does the risk of complications [7]. Creating evidence-based guidance for PAC in the second trimester could reduce harm for women who need PAC at a later stage in pregnancy. The aim of the present review was to establish recommendations for women needing second-trimester PAC. 2. Materials and methods The review was prepared in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines [8] and included the following review protocol. 2.1. Search strategy Medline (January 1, 1946, to December 31, 2013), POPLINE (January 1, 1927, to December 31, 2013), and Cochrane Central Register of Controlled Trials (January 1, 1898 to December 31, 2013) were searched with the term “‘pregnancy, second trimester [MeSH Terms]’ AND ‘fetal
http://dx.doi.org/10.1016/j.ijgo.2014.11.011 0020-7292/© 2015 International Federation of Gynecology and Obstetrics. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
A.G. Mark et al. / International Journal of Gynecology and Obstetrics 129 (2015) 98–103
demise OR ruptured membranes OR incomplete abortion OR spontaneous abortion OR missed abortion OR miscarriage OR abortion’”. The reference lists of retrieved articles were searched for nonindexed citations. There were no language criteria. 2.2. Inclusion criteria In the anticipation that randomized controlled trials of the management of inevitable or incomplete abortion in the second trimester would be scarce, nonrandomized trials and comparative cohort studies were also included. The review included studies of women presenting in the second trimester (12–28 weeks of pregnancy). Because second-trimester PAC has not yet been clinically defined, it was defined for the purposes of the present review as care for women presenting after induced or spontaneous abortion in the second trimester who needed uterine evacuation. Thus, studies were included in which women presented with inevitable or incomplete abortion including fetal demise, ruptured membranes, retained placenta, or retained fetal tissue. If the study had a mixed population of both PAC and induced abortion without a PAC indication, it was included only if 50% or more of the population met the criteria for PAC. If less than 50% of the study population were PAC patients, the study was included if outcomes for PAC were analyzed separately. Termination of pregnancy for fetal anomalies was considered as induced abortion. If the study included both first- and second-trimester pregnancies but reported the second-trimester data separately, the secondtrimester data were extracted and included in the review. If the study contained both second- and third-trimester pregnancies, all data were included if more than 50% of the women had a second-trimester pregnancy. Studies were excluded if they did not specify the pregnancy duration/uterine size. In terms of interventions, the present review included studies that focused on WHO-recommended methods for second-trimester abortion or induction of labor appropriate for low-resource settings: misoprostol-based regimens and/or dilation and evacuation (D&E) or oxytocin [2,9]. If a WHO-recommended treatment was compared with outdated techniques (including instillation, ethacridine, or other prostaglandin preparations), the study was not included. 2.3. Outcomes The primary outcome was effectiveness, measured as time to successful pregnancy expulsion. For medical management, it was expected that effectiveness would be expressed as either time from initiation of medications to fetal expulsion or as expulsion within 24 and 48 hours. Because expulsion time is clinically relevant for women, providers, and health facilities, this outcome was used as the primary outcome if reported. Secondary outcomes were complications, adverse effects, rates of incomplete expulsion or retained placenta, and complications such as infection, hemorrhage, transfusion, or unplanned surgery. 2.4. Data collection and analysis The first and second reviewers (A.G.M. and L.B.) reviewed all articles and determined which studies met the inclusion criteria. In cases of disagreement, a third reviewer (A.E.) judged whether the inclusion criteria were met. If the information in the original study was insufficient, that study’s author was contacted for clarification. The first and second reviewers independently extracted data on design, participants, interventions, outcomes, and risk of bias from included studies. Any variance was resolved by mutual agreement. For randomized controlled trials, the risk of bias was evaluated using Cochrane methodology [10], which includes measures for the concealment of randomization and allocation, the blinding of participants and study personnel, the blinding of outcome assessment, and the completeness of the outcome data. Studies were judged to have a low, high, or unclear risk of bias. For cohort studies, the Newcastle–Ottawa
99
Scale [11] was used; this scale assigns points for the selection and comparability of the study groups and for the ascertainment of the outcomes, and is judged on a nine-point scale. When the interventions in two or more articles were comparable, RevMan version 5.2 (Cochrane Collaboration, Oxford, UK) was used to perform a meta-analysis. This was only possible for the route of administration of misoprostol (vaginal versus oral). Because most studies did not have comparable design features, interventions, or outcomes, they could not be combined for metaanalysis. If this was the case, individual study outcomes were reported. 3. Results 3.1. Search results The search returned 406 articles via the Cochrane Central Register of Controlled Trials, 3294 articles via MEDLINE, and 91 via POPLINE. Three articles were identified through hand searches of reference lists. In total, 3794 titles and 858 abstracts were reviewed (Fig. 1). The full text of 196 articles was read. In total, 3777 studies were excluded. 3.2. Included studies Seventeen studies—15 randomized trials and two cohort studies—from 12 countries met the inclusion criteria (Table 1). For four studies, only the subpopulations of women with a PAC indication were included. The studies came from countries where abortion care is legal and accessible, as well as from countries with legal restrictions or limited access to abortion care. One article from Iran [28] specifically excluded five women who had had a self-induced abortion, but all other articles did not distinguish between women who had experienced an unsafe, illegal, or self-induced abortion and those with a spontaneous abortion or fetal demise. 3.3. Study participants A total of 1419 women were included, of whom 1154 (81.3%) presented with intrauterine fetal demise, 113 (8.0%) presented with ruptured membranes, and 152 (10.7%) had had an induced abortion. All studies included women presenting with intrauterine fetal death, and five studies [18,19,22,23,28] included women with ruptured membranes. No studies included women presenting with other potential PAC diagnoses, such as retained placenta without fetal tissues, hemorrhage, foreign body, infection, or other complications after second-trimester abortion. In most studies, the mean pregnancy duration fell within the mid to late second trimester (Table 1). 3.4. Risk of bias The risk of bias was low in five randomized controlled trials, unclear in another five, and high in the remaining five (Table 1). The two cohort studies had a low risk of bias according to the Newcastle–Ottawa scale (Table 1). Many studies were brief reports and did not contain supporting information to adequately judge the risk of bias. 3.5. Interventions for uterine evacuation Every study that met the inclusion criteria had a misoprostol treatment group. Most studies [12–17,21,22,24,25] reported comparisons of misoprostol dose, route, and interval. One study [18] compared outpatient misoprostol administration with inpatient administration, and another [27] compared saline-moistened and dry vaginal misoprostol. Four studies compared misoprostol with misoprostol plus an adjunct treatment such as oxytocin [19,28], osmotic dilators [20], and mifepristone [26]. One study [23] compared misoprostol with oxytocin. No study compared medical methods with D&E.
100
A.G. Mark et al. / International Journal of Gynecology and Obstetrics 129 (2015) 98–103
Fig. 1. Identification and selection of studies for the systematic review of second-trimester postabortion care.
3.6. Route of misoprostol administration Four studies [16,17,22,25] evaluating vaginal versus oral administration at comparable doses and intervals were combined for metaanalysis (Fig. 2). The doses and frequencies of misoprostol were 100 μg every 4 hours [16,17], 200 μg every 6 hours [25], and 400 μg every 8 hours after an initial loading dose of misoprostol of 800 μg vaginally [22]. The overall effect of the route of administration was strongly in favor of vaginal misoprostol (P b 0.001) (Fig. 2). One study [17] showed a higher rate of retained placenta at 60 minutes in the vaginal group than in the oral group (27% vs 6%; P = 0.02). However, this rate did not differ in two studies [16,22], and one study [25] did not include information about retained placenta or other outcomes. One study [16] compared sublingual and oral misoprostol. Sublingual administration was associated with a significantly shorter mean induction-to-expulsion interval than was oral administration (10.5 ± 3.7 h vs 21.0 ± 10.5 h; P b 0.05). The rates of manual placental removal did not differ (46% vs 54%). The same study [16] compared vaginal and sublingual misoprostol. The mean induction-to-expulsion interval after sublingual administration was similar to that after vaginal administration (10.5 ± 3.7 h vs 13.9 ± 5.1 h; P N 0.05). There was no difference in the rate of manual placental removal (46% vs 54%).
In the study of buccal misoprostol [12], the higher dose had a reduced mean fetal and placental expulsion time (18.5 h vs 23.9 h; P = 0.02), and improved rates of expulsion within 24 hours of induction (64.9% vs 39.5%; P = 0.002) and within 48 hours of induction (77.9% vs 61.8%; P = 0.03). There was a significant increase in the frequency of diarrhea among women using 200 μg (32.5% vs 10.7%; P = 0.001), but other adverse effects (e.g. nausea, vomiting, chills, headache, and pain) were similar. Women in the 100-μg group were less satisfied with the procedure, duration of treatment, and length of hospitalization than were women given 200 μg. In the study of vaginal misoprostol [24], the mean induction-toexpulsion interval was shorter in the 600-μg group than in the 400-μg group (9.2 ± 6.7 h versus 12.2 ± 7.7 h; P = 0.003). Adverse effects did not differ significantly between the two groups.
3.8. Misoprostol dosing interval One study [21] compared 200 μg misoprostol given vaginally every 6 hours versus every 12 hours. The mean induction-to-expulsion intervals did not differ (13.8 h vs 14.0 h; P N 0.05). The 24-hour and 48-hour success rates, incomplete abortion rates, and adverse effect rates also did not differ.
3.9. Heterogeneous comparisons of misoprostol route, dose, and interval 3.7. Misoprostol dose Two studies [12,24] compared misoprostol doses using comparable intervals and routes of administration. One [12] compared 200 μg buccal misoprostol every 6 hours with 100 μg, whereas the other [24] compared 600 μg vaginally given every 12 hours with 400 μg. Both of these studies demonstrated faster expulsion times with higher doses of misoprostol.
Three studies of misoprostol [13–15] could not be used to form conclusions because the comparisons of dose, route, and interval were heterogeneous. It was unclear what role the dose, loading dose, route, or dosing interval played in the reported outcome. In all three studies, higher cumulative doses irrespective of route were superior to lower doses.
A.G. Mark et al. / International Journal of Gynecology and Obstetrics 129 (2015) 98–103
101
Table 1 Study details. First author, year [reference]
Study design
Location
Treatment groups
Number of participants
Total Bracken, 2014 [12] Chittacharoen, 2003 [13]
RCT RCT
Vietnam, USA Thailand
Dickinson, 2002 [14]c
RCT
Australia
Edward, 2005 [15]c
USA
Elhassan, 2008 [16]
Retrospective cohort RCT
Sudan
Fadalla, 2004 [17]
RCT
Sudan
Gonzalez, 2001 [18]
RCT
USA
Hidar, 2001 [19]
RCT
Tunisia
Jain, 1996 [20]
RCT
USA
Jain, 1999 [21] Kurshid, 2009 [22]
RCT RCT
USA India
Nakintu, 2001 [23]
RCT
Uganda
Niromanesh, 2005 [24] Nyende, 2004 [25]
RCT RCT
Iran South Africa
Stibbe, 2012 [26]c
Retrospective cohort RCT RCT
Netherlands
c
Yilmaz, 2007 [27] Zageneh, 2012 [28]
Turkey Iran
100 μg vs 200 μg buccal misoprostol q6h 400 μg oral misoprostol q4h vs 200 μg vaginal misoprostol q12h 200 μg vaginal misoprostol q6h vs 400 μg vaginal misoprostol q6h vs 600 μg loading dose vaginal misoprostol then 200 μg vaginal misoprostol q6h 200 μg vaginal misoprostol q12h vs 400 μg vaginal misoprostol q6h 100 μg vaginal misoprostol q4h vs 100 μg sublingual misoprostol q4h vs 100 μg oral misoprostol q4h 100 μg vaginal misoprostol q4h vs 100 μg oral misoprostol q4h 200 μg vaginal misoprostol q6h in hospital vs 200 μg vaginal misoprostol q6h at home after first dose 200 μg vaginal misoprostol q12h vs 200 μg vaginal misoprostol q12h with 15 mIU/min oxytocin 200 μg vaginal misoprostol with or without laminaria at start of induction 200 μg vaginal misoprostol q12h vs q6h 800 μg loading dose vaginal misoprostol then 400 μg vaginal misoprostol q8h vs 800 μg loading dose vaginal misoprostol then 400 μg oral misoprostol q8h 50 μg vaginal misoprostol doubled every 6 h vs intravenous oxytocin to maximum dose 40 mIU/min 400 μg vs 600 μg vaginal misoprostol q12h 200 μg vaginal misoprostol q6h vs 200 μg oral misoprostol q6h 200 μg oral misoprostol q3h vs 200 mg oral mifepristone then 200 μg oral misoprostol q3h 36 h later 800 μg dry vaginal misoprostol q6h × 3 dry vs moistened 600 μg loading dose vaginal misoprostol then 400 μg vaginal misoprostol q6h vs 600 μg loading dose vaginal misoprostol then intravenous oxytocin
Induced abortion
Fetal death
Pregnancy duration, wka
Risk of biasb
Ruptured membranes
153 80
0 0
153 80
0 0
18 (14–28) 23 (16–41)
Low Low
36
0
36
0
18 (14–30)
Low
29
0
29
0
20 (13–27)
Low (7)
150
0
150
0
18 (13–28)
High
70
0
70
0
23 (13–28)
Unclear
87
33
36
18
18 (13–28)
Unclear
83
15
49
19
20 (13–29)
Unclear
38
0
38
0
16 (12–22)
Unclear
84 100
32 25
52 51
0 24
17 (12–22) 22 (19–23)
Unclear High
120
0
120
0
25 (18–40)
Low
100 38
0 0
100 38
0 0
NS (14–25) 28 (NS)
High High
50
0
50
0
26 (17–40)
Low (7)
81 120
36 11
45 57
0 52
20 (14–24) 17 (14–25)
Low High
Abbreviations: RCT, randomized controlled trial; q6h, every 6 hours; q4h, every 4 hours; q12h, every 12 hours; q8h, every 8 hours; NS, not stated; q3h, every 3 hours. a Values are given as mean (range). b Cohort studies have the Newcastle–Ottawa score in parentheses. c Substudy.
3.10. Addition of oxytocin to misoprostol Two studies [19,28] compared oxytocin added to misoprostol versus misoprostol only. In the first study [28], all women had a loading dose of 600 μg vaginal misoprostol and were then randomized to receive 400 μg misoprostol every 6 hours or concentrated oxytocin. No differences in the mean induction-to-expulsion time were found with misoprostol plus oxytocin versus misoprostol only (12.3 ± 6.0 h vs 12.1 ± 6.0 h; P N 0.05). Adverse effects were also not different between the two groups. In the second study [19], among women receiving misoprostol 200 μg every 12 hours, the addition of oxytocin at a rate of 15 mIU/min starting with the first misoprostol dose decreased the mean expulsion time (22.1 ± 10.8 h v 27.1 ± 14.1 h; P b 0.05), but the rates of fetal
expulsion at 48 hours, complete expulsion (fetus and placenta), and adverse effects were similar between the two groups. These two studies were not combined for meta-analysis because of their heterogeneous designs. 3.11. Addition of mifepristone to misoprostol One cohort study [26] evaluated women treated with 200 μg misoprostol vaginally every 3 hours alone, and a subsequent cohort treated with mifepristone given 24–48 hours before misoprostol dosing. Women with intrauterine fetal demise who had mifepristone pretreatment had shorter induction-to-expulsion times than did women who received misoprostol alone (10.6 h vs 16.2 h; P = 0.04). Mifepristone
Fig. 2. Meta-analysis of difference in time from induction to abortion with vaginal versus oral misoprostol, in hours. Abbreviation: CI, confidence interval.
102
A.G. Mark et al. / International Journal of Gynecology and Obstetrics 129 (2015) 98–103
pretreatment reduced the risk of infection (P = 0.049) and lowered the need for pain relief (P = 0.022). The risks of hemorrhage, retained placenta, and surgery were not different between the two groups. 3.12. Addition of laminaria to misoprostol One study [20] randomized women to receive laminaria or no intervention before undergoing misoprostol induction for fetal demise with 200 μg vaginal misoprostol every 12 hours. The induction-toexpulsion time was similar in the group receiving misoprostol only and the group receiving misoprostol plus laminaria (15.7 h vs 17.4 h; P = 0.54). Adverse effects (including vomiting, diarrhea, fever, and moderate pain) were similar in the two groups, but the rate of severe pain was higher among women treated with laminaria (15.2% vs 34.3%; P = 0.09). 3.13. Comparison of dry and moistened misoprostol One randomized controlled trial [27] compared dry misoprostol (800 μg vaginally) every 6 hours with moistened misoprostol (800 μg vaginally) every 6 hours. There was no difference in the induction-toexpulsion time or adverse effects between the two groups. 3.14. Comparison of home use to in-hospital misoprostol One randomized trial [18] evaluated beginning misoprostol at home versus the hospital. One group of women was hospitalized and treated with 200 μg of misoprostol vaginally every 6 hours until expulsion, whereas the second group was instructed on self-insertion at home, with instructions to return to the hospital when bleeding or contractions occurred, or when 24 hours had elapsed. Expulsion times in the home and hospital groups did not differ (14.0 h vs 12.0 h; P = 0.28), but the home administration group spent less time in the hospital (11 h vs 24 h; P b 0.05). There were two extramural deliveries without complications in the home group. All women favored the privacy and intimacy of home use. 3.15. Comparison of misoprostol with oxytocin alone One randomized trial [23] compared oxytocin alone with misoprostol alone for women presenting with fetal death at more than 18 weeks of pregnancy. The oxytocin dose was increased every 30 minutes until contractions started or until a maximum dose of 40 mIU/min was reached. Vaginal misoprostol was started at 50 μg, and was doubled for each subsequent dose; doses were given every 6 hours. The maximum misoprostol dose was 750 μg. The induction-to-abortion time was significantly shorter among women who received misoprostol than among those receiving oxytocin (12.4 h vs 23.3 hours; P = 0.004). 3.16. Other outcomes of interest Only one study [14] stated that women with a history of cesarean delivery were not excluded. Of the remainder, seven studies [12,16,20–22,25,27] excluded women with a previous cesarean delivery, one study [13] excluded women with classic cesarean delivery only, three studies [17,19,28] excluded women with more than one cesarean delivery, and five studies [15,18,23,24,26] did not specify whether women with a previous cesarean delivery were excluded. There was one report [14] of hysterectomy for hemorrhage and placenta accreta in a woman with four previous cesarean deliveries after using vaginal misoprostol. There were no reported uterine ruptures.
4. Discussion Overall, the present review has shown that the published literature lacks comparable, high-quality data on the management of PAC in the second trimester. There were limited data available for meta-analysis. However, several patterns emerge from the present review to guide clinical practice. First, misoprostol given either vaginally or sublingually is more effective than oral misoprostol. Second, misoprostol doses at or above 200 μg decrease the induction-to-expulsion time, but the maximum dose based on the risk of adverse effects or adverse outcomes is not known. Third, pretreatment with mifepristone may decrease the overall induction-to-expulsion time. Finally, misoprostol is associated with a shorter time to expulsion than oxytocin alone. These findings are analogous to those observed for managing induced abortion in the second trimester [28]. Non-oral (vaginal, sublingual, and possibly buccal) routes of misoprostol administration are more effective than is the oral route at the same doses. The present review confirmed that vaginal and sublingual misoprostol are more effective than is oral misoprostol. One study [12] used buccal dosing in both groups without comparison with other routes. Given the efficacy of buccal misoprostol in induced abortion in the first [29] or second [30] trimester and the similar uterotonic profile to vaginal dosing [31], buccal dosing seems appropriate. In both studies that compared misoprostol dosing [12,24], higher doses of misoprostol had shorter induction-to-abortion intervals than did lower doses. Given the superiority of the 200-μg dose over the 100-μg dose in one high-quality trial [12], misoprostol doses of 200 μg or more are likely to have shorter induction-to-abortion intervals than lower doses. The optimal dose, with which the risk of adverse effects is outweighed by the benefit of the shortened induction interval, is not yet established. In addition, as pregnancy advances into the late second trimester, lower doses of misoprostol could be used to achieve the same effect. In the included studies, the risk of adverse effects was not systematically evaluated and none of the studies was adequately powered to detect differences in serious adverse events. The recommended misoprostol dose for induced abortion in the second trimester is 400 μg [2]. This dose could be appropriate for PAC as well and should be studied further. The only article [21] designed to compare the dosing interval showed no difference in the induction-to-delivery time between 6-hourly and 12-hourly doses. In other studies [16,17,26], dosing intervals of less than 6 hours were used but not compared. The recommended dosing interval for induced abortion is 3 hours, which has improved outcomes when compared with 6-hourly dosing [32]. The addition of pretreatment with mifepristone would be expected to improve outcomes, on the basis of evidence from studies of induced abortion [32]. In the present review, this notion was supported by one small study in women with fetal demise [26]. More research is needed on this promising intervention. In women who do not require immediate intervention, mifepristone could be beneficial. Other interventions—e.g. moistening the misoprostol tablets, adding oxytocin to misoprostol, giving misoprostol at home, or adding laminaria—did not improve outcomes but could impact women’s experience and satisfaction with care. No studies compared D&E with medical methods of uterine evacuation or D&E variations (e.g. different surgical techniques or cervical preparation agents) in the PAC population. Dilation and evacuation is a WHO-recommended method for uterine evacuation in the second trimester [2]. Because D&E compares favorably with medical induction for second-trimester induced abortion, women could be offered D&E for PAC when skilled providers and supportive facilities are in place [33]. The strength of the present review is the inclusion of studies from multiple countries. Because studies were included from countries with restricted abortion access, the review is likely to include women presenting with intrauterine demise or rupture of membranes after unsafe or self-induced abortion.
A.G. Mark et al. / International Journal of Gynecology and Obstetrics 129 (2015) 98–103
Box 1 Recommendations for second-trimester postabortion care. The following recommendations are made on good and consistent scientific evidence (Grade A): • Misoprostol should be used in preference to oxytocin alone. • Vaginal or sublingual administration of misoprostol is superior to oral administration. The following recommendations are made on limited or inconsistent scientific evidence (Grade B): • Misoprostol should be given at a dose of at least 200 μg. • The frequency of misoprostol dosing should be at least every 6 hours. • When available and time permitting, mifepristone should be used before misoprostol. • Osmotic dilators should not be used. The following recommendations are based primarily on expert opinion (Grade C): • Buccal administration of misoprostol can be used. • Where skilled providers and supportive facilities exist, dilation and evacuation could be offered.
The limitations of the present review are the lack of systematic investigation, the heterogeneous nature of the interventions, the quality of the included studies, and the small number of women. The low quality of the studies meant adverse effects were not consistently measured or reported. Because of the low numbers, rare serious adverse events could not be assessed. Moreover, D&E could not be evaluated. In conclusion, misoprostol with or without mifepristone is an effective medical treatment for women presenting for secondtrimester PAC. Women with indications including fetal demise or ruptured membranes should be treated with misoprostol vaginally, sublingually, or buccally according to the evidence generated by the review (Box 1). More research is needed to see if recommended regimens for second-trimester induced abortion can be used for women who need PAC [2]; however, the evidence indicates that these regimens would be acceptable. Conflict of interest The authors have no conflicts of interest. References [1] Postabortion Care Consortium Community Task Force. Essential elements of postabortion care: an expanded and updated model. http://www.pac-consortium.org/ attachments/article/16/Essential%20Elements%20of%20Postabortion%20Care.pdf. Published 2002. Accessed May 13, 2014. [2] World Health Organization. Safe abortion: technical and policy guidance for health systems. Second edition. http://apps.who.int/iris/bitstream/10665/70914/1/9789241548434_ eng.pdf?ua=1. Published 2012. Accessed October 1, 2014. [3] Neilson JP, Gyte GM, Hickey M, Vazquez JC, Dou L. Medical treatments for incomplete miscarriage. Cochrane Database Syst Rev 2013;3:CD007223. [4] Fetters T, Vonthanak S, Picardo C, Rathavy T. Abortion-related complications in Cambodia. BJOG 2008;115(8):957–68. [5] Kalilani-Phiri L, Gebreselassie H, Levandowski BA, Kuchingale E, Kachale F, Kangaude G. The severity of abortion complications in Malawi. Int J Gynecol Obstet 2015; 128(2):160–4.
103
[6] Republic of Kenya, Ministry of Health. Incidence and Complications of Unsafe Abortion in Kenya: Key Findings of a National Study. http://www.guttmacher.org/ pubs/abortion-in-Kenya.pdf. Published August 2013. Accessed October 1, 2014. [7] Dragoman M, Sheldon WR, Qureshi Z, Blum J, Winikoff B, Ganatra B, et al. Overview of abortion cases with severe maternal outcomes in the WHO Multicountry Survey on Maternal and Newborn Health: a descriptive analysis. BJOG 2014;121(Suppl. 1): 25–31. [8] Moher D, Liberati A, Tetzlaff J, Altman DG, PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Ann Intern Med 2009;151(4):264–9. [9] World Health Organization. WHO recommendations for induction of labour. http:// whqlibdoc.who.int/publications/2011/9789241501156_eng.pdf. Published 2011. Accessed October 1, 2014. [10] Higgins JPT, Green S. Cochrane Handbook for Systematic Reviews of Interventions, vol. 4. New York, NY: John Wiley & Sons; 2011. [11] Wells GA, Shea B, O'Connell D, Peterson J, Welch V, Losos M, et al. The NewcastleOttawa Scale (NOS) for assessing the quality of nonrandomised studies in metaanalyses. http://www.ohri.ca/programs/clinical_epidemiology/oxford.htm. Accessed May 13, 2014. [12] Bracken H, Ngoc NT, Banks E, Blumenthal PD, Derman RJ, Patel A, et al. Buccal misoprostol for treatment of fetal death at 14-28 weeks of pregnancy: a double-blind randomized controlled trial. Contraception 2014;89(3):187–92. [13] Chittacharoen A, Herabutya Y, Punyavachira P. A randomized trial of oral and vaginal misoprostol to manage delivery in cases of fetal death. Obstet Gynecol 2003;101(1): 70–3. [14] Dickinson JE, Evans SF. The optimization of intravaginal misoprostol dosing schedules in second-trimester pregnancy termination. Am J Obstet Gynecol 2002;186(3):470–4. [15] Edwards RK, Sims SM. Outcomes of second-trimester pregnancy terminations with misoprostol: comparing 2 regimens. Am J Obstet Gynecol 2005;193(2):544–50. [16] Elhassan EM, Abubaker MS, Adam I. Sublingual compared with oral and vaginal misoprostol for termination of pregnancy with second-trimester fetal demise. Int J Gynecol Obstet 2008;100(1):82–3. [17] Fadalla FA, Mirghani OA, Adam I. Oral misoprostol vs. vaginal misoprostol for termination of pregnancy with intrauterine fetal demise in the second-trimester. Int J Gynecol Obstet 2004;86(1):52–3. [18] Gonzalez JA, Carlan SJ, Alverson MW. Outpatient second trimester pregnancy termination. Contraception 2001;63(2):89–93. [19] Hidar S, Fekih M, Chaïeb A, Bibi M, Mellouli R, Khaïri H. Oxytocin and misoprostol administered intravaginally for termination of pregnancy at 13 to 29 weeks of amenorrhea. A prospective randomized trial. J Gynecol Obstet Biol Reprod (Paris) 2001;30(5):439–43. [20] Jain JK, Mishell Jr DR. A comparison of misoprostol with and without laminaria tents for induction of second-trimester abortion. Am J Obstet Gynecol 1996;175(1):173–7. [21] Jain JK, Kuo J, Mishell Jr DR. A comparison of two dosing regimens of intravaginal misoprostol for second-trimester pregnancy termination. Obstet Gynecol 1999; 93(4):571–5. [22] Kurshid R, Ahmed A, Mir S, Shamas I. To assess the efficacy of two regimens of misoprostol for second trimester pregnancy termination-a randomized comparison. Internet J Gynecol Obstet 2009;14(1). [23] Nakintu N. A comparative study of vaginal misoprostol and intravenous oxytocin for induction of labour in women with intra uterine fetal death in Mulago Hospital, Uganda. Afr Health Sci 2001;1(2):55–9. [24] Niromanesh S, Hashemi-Fesharaki M, Mosavi-Jarrahi A. Second trimester abortion using intravaginal misoprostol. Int J Gynecol Obstet 2005;89(3):276–7. [25] Nyende L, Towobola OA, Mabina MH. Comparison of vaginal and oral misoprostol, for the induction of labour in women with intra-uterine foetal death. East Afr Med J 2004;81(4):179–82. [26] Stibbe KJ, de Weerd S. Induction of delivery by mifepristone and misoprostol in termination of pregnancy and intrauterine fetal death: 2nd and 3rd trimester induction of labour. Arch Gynecol Obstet 2012;286(3):795–6. [27] Yilmaz B, Kelekci S, Ertas IE, Ozel M, Sut N, Mollamahmutoglu L, et al. Randomized comparison of second trimester pregnancy termination utilizing saline moistened or dry misoprostol. Arch Gynecol Obstet 2007;276(5):511–6. [28] Zangeneh M, Malek-Khosravi S, Veisi F, Rezavand N, Rezaee M, Rajatee M. Multipledose vaginal misoprostol and single-dose misoprostol plus oxytocin for termination of second-trimester pregnancy. Int J Gynecol Obstet 2012;117(1):78–80. [29] Winikoff B, Dzuba IG, Creinin MD, Crowden WA, Goldberg AB, Gonzales J, et al. Two distinct oral routes of misoprostol in mifepristone medical abortion: a randomized controlled trial. Obstet Gynecol 2008;112(6):1303–10. [30] Ellis SC, Kapp N, Vragpvoc O, Borgata L. Randomized trial of buccal versus vaginal misoprostol for induction of second trimester abortion. Contraception 2010;81(5): 441–5. [31] Meckstroth KR, Whitaker AK, Bertisch S, Goldberg AB, Darney PD. Misoprostol administered by epithelial routes: Drug absorption and uterine response. Obstet Gynecol 2006;108(3 Pt 1):582–90. [32] Wildschut H, Both MI, Medema S, Thomee E, Wildhagen MF, Kapp N. Medical methods for mid-trimester termination of pregnancy. Cochrane Database Syst Rev 2011;1:CD005216. [33] Lohr PA, Hayes JL, Gemzell-Danielsson K. Surgical versus medical methods for second trimester induced abortion. Cochrane Database Syst Rev 2008;1:CD006714.
Contents lists available at ScienceDirect
International Journal of Gynecology and Obstetrics journal homepage: www.elsevier.com/locate/ijgo
REVIEW ARTICLE
Second-trimester postabortion care for ruptured membranes, fetal demise, and incomplete abortion Alice G. Mark a,⁎, Alison Edelman a,b, Lynn Borgatta c a b c
Ipas, Chapel Hill, NC, USA Department of Obstetrics and Gynecology, Oregon Health and Science University, Portland, OR, USA Department of Obstetrics and Gynecology, Boston University, Boston, MA, USA
a r t i c l e
i n f o
Article history: Received 29 May 2014 Received in revised form 5 November 2014 Accepted 9 January 2015 Keywords: Abortion Intrauterine fetal death Mifepristone Misoprostol Postabortion care Premature rupture of membranes Second trimester Unsafe abortion
a b s t r a c t Background: Guidance for postabortion care (PAC) is established for the first trimester but limited in the second trimester. Objectives: To establish evidence-based recommendations for PAC in the second trimester. Search strategy: Medline, POPLINE, and the Cochrane Central Register of Controlled Trials were searched with terms related to second-trimester PAC, including fetal demise, ruptured membranes, and incomplete abortion. The reference lists of retrieved articles were also searched. Selection criteria: Clinical trials and comparative studies of women presenting in the second trimester (12–28 weeks) were included if more than 50% of participants met PAC criteria or if outcomes for PAC were analyzed separately. Data collection and analysis: Data were extracted from included studies. When interventions in at least two articles were comparable, a meta-analysis was performed. Main results: Overall, 17 studies of 1419 women met inclusion criteria. Misoprostol given vaginally, sublingually, or buccally was associated with shorter expulsion times than was oral misoprostol. Additionally, 200 μg of misoprostol was more effective than lower doses. Pretreatment with mifepristone decreased expulsion time. Misoprostol was more effective than oxytocin. Conclusion: Misoprostol with or without mifepristone is an effective treatment for second-trimester PAC. The minimum misoprostol dose is 200 μg vaginally, sublingually, or buccally every 6–12 hours. © 2015 International Federation of Gynecology and Obstetrics. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
1. Introduction Postabortion care (PAC) is the comprehensive treatment of women presenting after spontaneous or induced, safe or unsafe abortion. PAC includes several key components: community and service provider partnerships, counseling, treatment of incomplete abortion with uterine evacuation, management of complications, contraception, and reproductive health services [1]. Following unsafe abortion, PAC services reduce maternal morbidity and mortality. Evidence-based guidelines [2,3] exist to support medical management and vacuum aspiration for incomplete abortion in women with a uterine size less than 13 weeks of pregnancy. However, many women in countries with restricted abortion access present later in pregnancy with fetal demise, ruptured membranes, retained placenta, hemorrhage, a foreign body, or infectious complications resulting from unsafe or self-induced abortion. These women could receive inconsistent care from different providers in obstetrics and gynecology wards or emergency departments. Magnitude studies from Sub-Saharan ⁎ Corresponding author at: PO Box 9990, Chapel Hill, NC, 27515, USA. Tel.: +1 919 960 5581; fax: +1 919 929 0258. E-mail address: [email protected] (A.G. Mark).
Africa and Southeast Asia show high rates of hospital presentation for PAC in the second trimester: 17% of women presenting for PAC in Cambodia are in the second trimester [4], whereas this rate is 35% in Malawi [5] and 41% in Kenya [6]. As the pregnancy duration increases, so does the risk of complications [7]. Creating evidence-based guidance for PAC in the second trimester could reduce harm for women who need PAC at a later stage in pregnancy. The aim of the present review was to establish recommendations for women needing second-trimester PAC. 2. Materials and methods The review was prepared in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines [8] and included the following review protocol. 2.1. Search strategy Medline (January 1, 1946, to December 31, 2013), POPLINE (January 1, 1927, to December 31, 2013), and Cochrane Central Register of Controlled Trials (January 1, 1898 to December 31, 2013) were searched with the term “‘pregnancy, second trimester [MeSH Terms]’ AND ‘fetal
http://dx.doi.org/10.1016/j.ijgo.2014.11.011 0020-7292/© 2015 International Federation of Gynecology and Obstetrics. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
A.G. Mark et al. / International Journal of Gynecology and Obstetrics 129 (2015) 98–103
demise OR ruptured membranes OR incomplete abortion OR spontaneous abortion OR missed abortion OR miscarriage OR abortion’”. The reference lists of retrieved articles were searched for nonindexed citations. There were no language criteria. 2.2. Inclusion criteria In the anticipation that randomized controlled trials of the management of inevitable or incomplete abortion in the second trimester would be scarce, nonrandomized trials and comparative cohort studies were also included. The review included studies of women presenting in the second trimester (12–28 weeks of pregnancy). Because second-trimester PAC has not yet been clinically defined, it was defined for the purposes of the present review as care for women presenting after induced or spontaneous abortion in the second trimester who needed uterine evacuation. Thus, studies were included in which women presented with inevitable or incomplete abortion including fetal demise, ruptured membranes, retained placenta, or retained fetal tissue. If the study had a mixed population of both PAC and induced abortion without a PAC indication, it was included only if 50% or more of the population met the criteria for PAC. If less than 50% of the study population were PAC patients, the study was included if outcomes for PAC were analyzed separately. Termination of pregnancy for fetal anomalies was considered as induced abortion. If the study included both first- and second-trimester pregnancies but reported the second-trimester data separately, the secondtrimester data were extracted and included in the review. If the study contained both second- and third-trimester pregnancies, all data were included if more than 50% of the women had a second-trimester pregnancy. Studies were excluded if they did not specify the pregnancy duration/uterine size. In terms of interventions, the present review included studies that focused on WHO-recommended methods for second-trimester abortion or induction of labor appropriate for low-resource settings: misoprostol-based regimens and/or dilation and evacuation (D&E) or oxytocin [2,9]. If a WHO-recommended treatment was compared with outdated techniques (including instillation, ethacridine, or other prostaglandin preparations), the study was not included. 2.3. Outcomes The primary outcome was effectiveness, measured as time to successful pregnancy expulsion. For medical management, it was expected that effectiveness would be expressed as either time from initiation of medications to fetal expulsion or as expulsion within 24 and 48 hours. Because expulsion time is clinically relevant for women, providers, and health facilities, this outcome was used as the primary outcome if reported. Secondary outcomes were complications, adverse effects, rates of incomplete expulsion or retained placenta, and complications such as infection, hemorrhage, transfusion, or unplanned surgery. 2.4. Data collection and analysis The first and second reviewers (A.G.M. and L.B.) reviewed all articles and determined which studies met the inclusion criteria. In cases of disagreement, a third reviewer (A.E.) judged whether the inclusion criteria were met. If the information in the original study was insufficient, that study’s author was contacted for clarification. The first and second reviewers independently extracted data on design, participants, interventions, outcomes, and risk of bias from included studies. Any variance was resolved by mutual agreement. For randomized controlled trials, the risk of bias was evaluated using Cochrane methodology [10], which includes measures for the concealment of randomization and allocation, the blinding of participants and study personnel, the blinding of outcome assessment, and the completeness of the outcome data. Studies were judged to have a low, high, or unclear risk of bias. For cohort studies, the Newcastle–Ottawa
99
Scale [11] was used; this scale assigns points for the selection and comparability of the study groups and for the ascertainment of the outcomes, and is judged on a nine-point scale. When the interventions in two or more articles were comparable, RevMan version 5.2 (Cochrane Collaboration, Oxford, UK) was used to perform a meta-analysis. This was only possible for the route of administration of misoprostol (vaginal versus oral). Because most studies did not have comparable design features, interventions, or outcomes, they could not be combined for metaanalysis. If this was the case, individual study outcomes were reported. 3. Results 3.1. Search results The search returned 406 articles via the Cochrane Central Register of Controlled Trials, 3294 articles via MEDLINE, and 91 via POPLINE. Three articles were identified through hand searches of reference lists. In total, 3794 titles and 858 abstracts were reviewed (Fig. 1). The full text of 196 articles was read. In total, 3777 studies were excluded. 3.2. Included studies Seventeen studies—15 randomized trials and two cohort studies—from 12 countries met the inclusion criteria (Table 1). For four studies, only the subpopulations of women with a PAC indication were included. The studies came from countries where abortion care is legal and accessible, as well as from countries with legal restrictions or limited access to abortion care. One article from Iran [28] specifically excluded five women who had had a self-induced abortion, but all other articles did not distinguish between women who had experienced an unsafe, illegal, or self-induced abortion and those with a spontaneous abortion or fetal demise. 3.3. Study participants A total of 1419 women were included, of whom 1154 (81.3%) presented with intrauterine fetal demise, 113 (8.0%) presented with ruptured membranes, and 152 (10.7%) had had an induced abortion. All studies included women presenting with intrauterine fetal death, and five studies [18,19,22,23,28] included women with ruptured membranes. No studies included women presenting with other potential PAC diagnoses, such as retained placenta without fetal tissues, hemorrhage, foreign body, infection, or other complications after second-trimester abortion. In most studies, the mean pregnancy duration fell within the mid to late second trimester (Table 1). 3.4. Risk of bias The risk of bias was low in five randomized controlled trials, unclear in another five, and high in the remaining five (Table 1). The two cohort studies had a low risk of bias according to the Newcastle–Ottawa scale (Table 1). Many studies were brief reports and did not contain supporting information to adequately judge the risk of bias. 3.5. Interventions for uterine evacuation Every study that met the inclusion criteria had a misoprostol treatment group. Most studies [12–17,21,22,24,25] reported comparisons of misoprostol dose, route, and interval. One study [18] compared outpatient misoprostol administration with inpatient administration, and another [27] compared saline-moistened and dry vaginal misoprostol. Four studies compared misoprostol with misoprostol plus an adjunct treatment such as oxytocin [19,28], osmotic dilators [20], and mifepristone [26]. One study [23] compared misoprostol with oxytocin. No study compared medical methods with D&E.
100
A.G. Mark et al. / International Journal of Gynecology and Obstetrics 129 (2015) 98–103
Fig. 1. Identification and selection of studies for the systematic review of second-trimester postabortion care.
3.6. Route of misoprostol administration Four studies [16,17,22,25] evaluating vaginal versus oral administration at comparable doses and intervals were combined for metaanalysis (Fig. 2). The doses and frequencies of misoprostol were 100 μg every 4 hours [16,17], 200 μg every 6 hours [25], and 400 μg every 8 hours after an initial loading dose of misoprostol of 800 μg vaginally [22]. The overall effect of the route of administration was strongly in favor of vaginal misoprostol (P b 0.001) (Fig. 2). One study [17] showed a higher rate of retained placenta at 60 minutes in the vaginal group than in the oral group (27% vs 6%; P = 0.02). However, this rate did not differ in two studies [16,22], and one study [25] did not include information about retained placenta or other outcomes. One study [16] compared sublingual and oral misoprostol. Sublingual administration was associated with a significantly shorter mean induction-to-expulsion interval than was oral administration (10.5 ± 3.7 h vs 21.0 ± 10.5 h; P b 0.05). The rates of manual placental removal did not differ (46% vs 54%). The same study [16] compared vaginal and sublingual misoprostol. The mean induction-to-expulsion interval after sublingual administration was similar to that after vaginal administration (10.5 ± 3.7 h vs 13.9 ± 5.1 h; P N 0.05). There was no difference in the rate of manual placental removal (46% vs 54%).
In the study of buccal misoprostol [12], the higher dose had a reduced mean fetal and placental expulsion time (18.5 h vs 23.9 h; P = 0.02), and improved rates of expulsion within 24 hours of induction (64.9% vs 39.5%; P = 0.002) and within 48 hours of induction (77.9% vs 61.8%; P = 0.03). There was a significant increase in the frequency of diarrhea among women using 200 μg (32.5% vs 10.7%; P = 0.001), but other adverse effects (e.g. nausea, vomiting, chills, headache, and pain) were similar. Women in the 100-μg group were less satisfied with the procedure, duration of treatment, and length of hospitalization than were women given 200 μg. In the study of vaginal misoprostol [24], the mean induction-toexpulsion interval was shorter in the 600-μg group than in the 400-μg group (9.2 ± 6.7 h versus 12.2 ± 7.7 h; P = 0.003). Adverse effects did not differ significantly between the two groups.
3.8. Misoprostol dosing interval One study [21] compared 200 μg misoprostol given vaginally every 6 hours versus every 12 hours. The mean induction-to-expulsion intervals did not differ (13.8 h vs 14.0 h; P N 0.05). The 24-hour and 48-hour success rates, incomplete abortion rates, and adverse effect rates also did not differ.
3.9. Heterogeneous comparisons of misoprostol route, dose, and interval 3.7. Misoprostol dose Two studies [12,24] compared misoprostol doses using comparable intervals and routes of administration. One [12] compared 200 μg buccal misoprostol every 6 hours with 100 μg, whereas the other [24] compared 600 μg vaginally given every 12 hours with 400 μg. Both of these studies demonstrated faster expulsion times with higher doses of misoprostol.
Three studies of misoprostol [13–15] could not be used to form conclusions because the comparisons of dose, route, and interval were heterogeneous. It was unclear what role the dose, loading dose, route, or dosing interval played in the reported outcome. In all three studies, higher cumulative doses irrespective of route were superior to lower doses.
A.G. Mark et al. / International Journal of Gynecology and Obstetrics 129 (2015) 98–103
101
Table 1 Study details. First author, year [reference]
Study design
Location
Treatment groups
Number of participants
Total Bracken, 2014 [12] Chittacharoen, 2003 [13]
RCT RCT
Vietnam, USA Thailand
Dickinson, 2002 [14]c
RCT
Australia
Edward, 2005 [15]c
USA
Elhassan, 2008 [16]
Retrospective cohort RCT
Sudan
Fadalla, 2004 [17]
RCT
Sudan
Gonzalez, 2001 [18]
RCT
USA
Hidar, 2001 [19]
RCT
Tunisia
Jain, 1996 [20]
RCT
USA
Jain, 1999 [21] Kurshid, 2009 [22]
RCT RCT
USA India
Nakintu, 2001 [23]
RCT
Uganda
Niromanesh, 2005 [24] Nyende, 2004 [25]
RCT RCT
Iran South Africa
Stibbe, 2012 [26]c
Retrospective cohort RCT RCT
Netherlands
c
Yilmaz, 2007 [27] Zageneh, 2012 [28]
Turkey Iran
100 μg vs 200 μg buccal misoprostol q6h 400 μg oral misoprostol q4h vs 200 μg vaginal misoprostol q12h 200 μg vaginal misoprostol q6h vs 400 μg vaginal misoprostol q6h vs 600 μg loading dose vaginal misoprostol then 200 μg vaginal misoprostol q6h 200 μg vaginal misoprostol q12h vs 400 μg vaginal misoprostol q6h 100 μg vaginal misoprostol q4h vs 100 μg sublingual misoprostol q4h vs 100 μg oral misoprostol q4h 100 μg vaginal misoprostol q4h vs 100 μg oral misoprostol q4h 200 μg vaginal misoprostol q6h in hospital vs 200 μg vaginal misoprostol q6h at home after first dose 200 μg vaginal misoprostol q12h vs 200 μg vaginal misoprostol q12h with 15 mIU/min oxytocin 200 μg vaginal misoprostol with or without laminaria at start of induction 200 μg vaginal misoprostol q12h vs q6h 800 μg loading dose vaginal misoprostol then 400 μg vaginal misoprostol q8h vs 800 μg loading dose vaginal misoprostol then 400 μg oral misoprostol q8h 50 μg vaginal misoprostol doubled every 6 h vs intravenous oxytocin to maximum dose 40 mIU/min 400 μg vs 600 μg vaginal misoprostol q12h 200 μg vaginal misoprostol q6h vs 200 μg oral misoprostol q6h 200 μg oral misoprostol q3h vs 200 mg oral mifepristone then 200 μg oral misoprostol q3h 36 h later 800 μg dry vaginal misoprostol q6h × 3 dry vs moistened 600 μg loading dose vaginal misoprostol then 400 μg vaginal misoprostol q6h vs 600 μg loading dose vaginal misoprostol then intravenous oxytocin
Induced abortion
Fetal death
Pregnancy duration, wka
Risk of biasb
Ruptured membranes
153 80
0 0
153 80
0 0
18 (14–28) 23 (16–41)
Low Low
36
0
36
0
18 (14–30)
Low
29
0
29
0
20 (13–27)
Low (7)
150
0
150
0
18 (13–28)
High
70
0
70
0
23 (13–28)
Unclear
87
33
36
18
18 (13–28)
Unclear
83
15
49
19
20 (13–29)
Unclear
38
0
38
0
16 (12–22)
Unclear
84 100
32 25
52 51
0 24
17 (12–22) 22 (19–23)
Unclear High
120
0
120
0
25 (18–40)
Low
100 38
0 0
100 38
0 0
NS (14–25) 28 (NS)
High High
50
0
50
0
26 (17–40)
Low (7)
81 120
36 11
45 57
0 52
20 (14–24) 17 (14–25)
Low High
Abbreviations: RCT, randomized controlled trial; q6h, every 6 hours; q4h, every 4 hours; q12h, every 12 hours; q8h, every 8 hours; NS, not stated; q3h, every 3 hours. a Values are given as mean (range). b Cohort studies have the Newcastle–Ottawa score in parentheses. c Substudy.
3.10. Addition of oxytocin to misoprostol Two studies [19,28] compared oxytocin added to misoprostol versus misoprostol only. In the first study [28], all women had a loading dose of 600 μg vaginal misoprostol and were then randomized to receive 400 μg misoprostol every 6 hours or concentrated oxytocin. No differences in the mean induction-to-expulsion time were found with misoprostol plus oxytocin versus misoprostol only (12.3 ± 6.0 h vs 12.1 ± 6.0 h; P N 0.05). Adverse effects were also not different between the two groups. In the second study [19], among women receiving misoprostol 200 μg every 12 hours, the addition of oxytocin at a rate of 15 mIU/min starting with the first misoprostol dose decreased the mean expulsion time (22.1 ± 10.8 h v 27.1 ± 14.1 h; P b 0.05), but the rates of fetal
expulsion at 48 hours, complete expulsion (fetus and placenta), and adverse effects were similar between the two groups. These two studies were not combined for meta-analysis because of their heterogeneous designs. 3.11. Addition of mifepristone to misoprostol One cohort study [26] evaluated women treated with 200 μg misoprostol vaginally every 3 hours alone, and a subsequent cohort treated with mifepristone given 24–48 hours before misoprostol dosing. Women with intrauterine fetal demise who had mifepristone pretreatment had shorter induction-to-expulsion times than did women who received misoprostol alone (10.6 h vs 16.2 h; P = 0.04). Mifepristone
Fig. 2. Meta-analysis of difference in time from induction to abortion with vaginal versus oral misoprostol, in hours. Abbreviation: CI, confidence interval.
102
A.G. Mark et al. / International Journal of Gynecology and Obstetrics 129 (2015) 98–103
pretreatment reduced the risk of infection (P = 0.049) and lowered the need for pain relief (P = 0.022). The risks of hemorrhage, retained placenta, and surgery were not different between the two groups. 3.12. Addition of laminaria to misoprostol One study [20] randomized women to receive laminaria or no intervention before undergoing misoprostol induction for fetal demise with 200 μg vaginal misoprostol every 12 hours. The induction-toexpulsion time was similar in the group receiving misoprostol only and the group receiving misoprostol plus laminaria (15.7 h vs 17.4 h; P = 0.54). Adverse effects (including vomiting, diarrhea, fever, and moderate pain) were similar in the two groups, but the rate of severe pain was higher among women treated with laminaria (15.2% vs 34.3%; P = 0.09). 3.13. Comparison of dry and moistened misoprostol One randomized controlled trial [27] compared dry misoprostol (800 μg vaginally) every 6 hours with moistened misoprostol (800 μg vaginally) every 6 hours. There was no difference in the induction-toexpulsion time or adverse effects between the two groups. 3.14. Comparison of home use to in-hospital misoprostol One randomized trial [18] evaluated beginning misoprostol at home versus the hospital. One group of women was hospitalized and treated with 200 μg of misoprostol vaginally every 6 hours until expulsion, whereas the second group was instructed on self-insertion at home, with instructions to return to the hospital when bleeding or contractions occurred, or when 24 hours had elapsed. Expulsion times in the home and hospital groups did not differ (14.0 h vs 12.0 h; P = 0.28), but the home administration group spent less time in the hospital (11 h vs 24 h; P b 0.05). There were two extramural deliveries without complications in the home group. All women favored the privacy and intimacy of home use. 3.15. Comparison of misoprostol with oxytocin alone One randomized trial [23] compared oxytocin alone with misoprostol alone for women presenting with fetal death at more than 18 weeks of pregnancy. The oxytocin dose was increased every 30 minutes until contractions started or until a maximum dose of 40 mIU/min was reached. Vaginal misoprostol was started at 50 μg, and was doubled for each subsequent dose; doses were given every 6 hours. The maximum misoprostol dose was 750 μg. The induction-to-abortion time was significantly shorter among women who received misoprostol than among those receiving oxytocin (12.4 h vs 23.3 hours; P = 0.004). 3.16. Other outcomes of interest Only one study [14] stated that women with a history of cesarean delivery were not excluded. Of the remainder, seven studies [12,16,20–22,25,27] excluded women with a previous cesarean delivery, one study [13] excluded women with classic cesarean delivery only, three studies [17,19,28] excluded women with more than one cesarean delivery, and five studies [15,18,23,24,26] did not specify whether women with a previous cesarean delivery were excluded. There was one report [14] of hysterectomy for hemorrhage and placenta accreta in a woman with four previous cesarean deliveries after using vaginal misoprostol. There were no reported uterine ruptures.
4. Discussion Overall, the present review has shown that the published literature lacks comparable, high-quality data on the management of PAC in the second trimester. There were limited data available for meta-analysis. However, several patterns emerge from the present review to guide clinical practice. First, misoprostol given either vaginally or sublingually is more effective than oral misoprostol. Second, misoprostol doses at or above 200 μg decrease the induction-to-expulsion time, but the maximum dose based on the risk of adverse effects or adverse outcomes is not known. Third, pretreatment with mifepristone may decrease the overall induction-to-expulsion time. Finally, misoprostol is associated with a shorter time to expulsion than oxytocin alone. These findings are analogous to those observed for managing induced abortion in the second trimester [28]. Non-oral (vaginal, sublingual, and possibly buccal) routes of misoprostol administration are more effective than is the oral route at the same doses. The present review confirmed that vaginal and sublingual misoprostol are more effective than is oral misoprostol. One study [12] used buccal dosing in both groups without comparison with other routes. Given the efficacy of buccal misoprostol in induced abortion in the first [29] or second [30] trimester and the similar uterotonic profile to vaginal dosing [31], buccal dosing seems appropriate. In both studies that compared misoprostol dosing [12,24], higher doses of misoprostol had shorter induction-to-abortion intervals than did lower doses. Given the superiority of the 200-μg dose over the 100-μg dose in one high-quality trial [12], misoprostol doses of 200 μg or more are likely to have shorter induction-to-abortion intervals than lower doses. The optimal dose, with which the risk of adverse effects is outweighed by the benefit of the shortened induction interval, is not yet established. In addition, as pregnancy advances into the late second trimester, lower doses of misoprostol could be used to achieve the same effect. In the included studies, the risk of adverse effects was not systematically evaluated and none of the studies was adequately powered to detect differences in serious adverse events. The recommended misoprostol dose for induced abortion in the second trimester is 400 μg [2]. This dose could be appropriate for PAC as well and should be studied further. The only article [21] designed to compare the dosing interval showed no difference in the induction-to-delivery time between 6-hourly and 12-hourly doses. In other studies [16,17,26], dosing intervals of less than 6 hours were used but not compared. The recommended dosing interval for induced abortion is 3 hours, which has improved outcomes when compared with 6-hourly dosing [32]. The addition of pretreatment with mifepristone would be expected to improve outcomes, on the basis of evidence from studies of induced abortion [32]. In the present review, this notion was supported by one small study in women with fetal demise [26]. More research is needed on this promising intervention. In women who do not require immediate intervention, mifepristone could be beneficial. Other interventions—e.g. moistening the misoprostol tablets, adding oxytocin to misoprostol, giving misoprostol at home, or adding laminaria—did not improve outcomes but could impact women’s experience and satisfaction with care. No studies compared D&E with medical methods of uterine evacuation or D&E variations (e.g. different surgical techniques or cervical preparation agents) in the PAC population. Dilation and evacuation is a WHO-recommended method for uterine evacuation in the second trimester [2]. Because D&E compares favorably with medical induction for second-trimester induced abortion, women could be offered D&E for PAC when skilled providers and supportive facilities are in place [33]. The strength of the present review is the inclusion of studies from multiple countries. Because studies were included from countries with restricted abortion access, the review is likely to include women presenting with intrauterine demise or rupture of membranes after unsafe or self-induced abortion.
A.G. Mark et al. / International Journal of Gynecology and Obstetrics 129 (2015) 98–103
Box 1 Recommendations for second-trimester postabortion care. The following recommendations are made on good and consistent scientific evidence (Grade A): • Misoprostol should be used in preference to oxytocin alone. • Vaginal or sublingual administration of misoprostol is superior to oral administration. The following recommendations are made on limited or inconsistent scientific evidence (Grade B): • Misoprostol should be given at a dose of at least 200 μg. • The frequency of misoprostol dosing should be at least every 6 hours. • When available and time permitting, mifepristone should be used before misoprostol. • Osmotic dilators should not be used. The following recommendations are based primarily on expert opinion (Grade C): • Buccal administration of misoprostol can be used. • Where skilled providers and supportive facilities exist, dilation and evacuation could be offered.
The limitations of the present review are the lack of systematic investigation, the heterogeneous nature of the interventions, the quality of the included studies, and the small number of women. The low quality of the studies meant adverse effects were not consistently measured or reported. Because of the low numbers, rare serious adverse events could not be assessed. Moreover, D&E could not be evaluated. In conclusion, misoprostol with or without mifepristone is an effective medical treatment for women presenting for secondtrimester PAC. Women with indications including fetal demise or ruptured membranes should be treated with misoprostol vaginally, sublingually, or buccally according to the evidence generated by the review (Box 1). More research is needed to see if recommended regimens for second-trimester induced abortion can be used for women who need PAC [2]; however, the evidence indicates that these regimens would be acceptable. Conflict of interest The authors have no conflicts of interest. References [1] Postabortion Care Consortium Community Task Force. Essential elements of postabortion care: an expanded and updated model. http://www.pac-consortium.org/ attachments/article/16/Essential%20Elements%20of%20Postabortion%20Care.pdf. Published 2002. Accessed May 13, 2014. [2] World Health Organization. Safe abortion: technical and policy guidance for health systems. Second edition. http://apps.who.int/iris/bitstream/10665/70914/1/9789241548434_ eng.pdf?ua=1. Published 2012. Accessed October 1, 2014. [3] Neilson JP, Gyte GM, Hickey M, Vazquez JC, Dou L. Medical treatments for incomplete miscarriage. Cochrane Database Syst Rev 2013;3:CD007223. [4] Fetters T, Vonthanak S, Picardo C, Rathavy T. Abortion-related complications in Cambodia. BJOG 2008;115(8):957–68. [5] Kalilani-Phiri L, Gebreselassie H, Levandowski BA, Kuchingale E, Kachale F, Kangaude G. The severity of abortion complications in Malawi. Int J Gynecol Obstet 2015; 128(2):160–4.
103
[6] Republic of Kenya, Ministry of Health. Incidence and Complications of Unsafe Abortion in Kenya: Key Findings of a National Study. http://www.guttmacher.org/ pubs/abortion-in-Kenya.pdf. Published August 2013. Accessed October 1, 2014. [7] Dragoman M, Sheldon WR, Qureshi Z, Blum J, Winikoff B, Ganatra B, et al. Overview of abortion cases with severe maternal outcomes in the WHO Multicountry Survey on Maternal and Newborn Health: a descriptive analysis. BJOG 2014;121(Suppl. 1): 25–31. [8] Moher D, Liberati A, Tetzlaff J, Altman DG, PRISMA Group. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Ann Intern Med 2009;151(4):264–9. [9] World Health Organization. WHO recommendations for induction of labour. http:// whqlibdoc.who.int/publications/2011/9789241501156_eng.pdf. Published 2011. Accessed October 1, 2014. [10] Higgins JPT, Green S. Cochrane Handbook for Systematic Reviews of Interventions, vol. 4. New York, NY: John Wiley & Sons; 2011. [11] Wells GA, Shea B, O'Connell D, Peterson J, Welch V, Losos M, et al. The NewcastleOttawa Scale (NOS) for assessing the quality of nonrandomised studies in metaanalyses. http://www.ohri.ca/programs/clinical_epidemiology/oxford.htm. Accessed May 13, 2014. [12] Bracken H, Ngoc NT, Banks E, Blumenthal PD, Derman RJ, Patel A, et al. Buccal misoprostol for treatment of fetal death at 14-28 weeks of pregnancy: a double-blind randomized controlled trial. Contraception 2014;89(3):187–92. [13] Chittacharoen A, Herabutya Y, Punyavachira P. A randomized trial of oral and vaginal misoprostol to manage delivery in cases of fetal death. Obstet Gynecol 2003;101(1): 70–3. [14] Dickinson JE, Evans SF. The optimization of intravaginal misoprostol dosing schedules in second-trimester pregnancy termination. Am J Obstet Gynecol 2002;186(3):470–4. [15] Edwards RK, Sims SM. Outcomes of second-trimester pregnancy terminations with misoprostol: comparing 2 regimens. Am J Obstet Gynecol 2005;193(2):544–50. [16] Elhassan EM, Abubaker MS, Adam I. Sublingual compared with oral and vaginal misoprostol for termination of pregnancy with second-trimester fetal demise. Int J Gynecol Obstet 2008;100(1):82–3. [17] Fadalla FA, Mirghani OA, Adam I. Oral misoprostol vs. vaginal misoprostol for termination of pregnancy with intrauterine fetal demise in the second-trimester. Int J Gynecol Obstet 2004;86(1):52–3. [18] Gonzalez JA, Carlan SJ, Alverson MW. Outpatient second trimester pregnancy termination. Contraception 2001;63(2):89–93. [19] Hidar S, Fekih M, Chaïeb A, Bibi M, Mellouli R, Khaïri H. Oxytocin and misoprostol administered intravaginally for termination of pregnancy at 13 to 29 weeks of amenorrhea. A prospective randomized trial. J Gynecol Obstet Biol Reprod (Paris) 2001;30(5):439–43. [20] Jain JK, Mishell Jr DR. A comparison of misoprostol with and without laminaria tents for induction of second-trimester abortion. Am J Obstet Gynecol 1996;175(1):173–7. [21] Jain JK, Kuo J, Mishell Jr DR. A comparison of two dosing regimens of intravaginal misoprostol for second-trimester pregnancy termination. Obstet Gynecol 1999; 93(4):571–5. [22] Kurshid R, Ahmed A, Mir S, Shamas I. To assess the efficacy of two regimens of misoprostol for second trimester pregnancy termination-a randomized comparison. Internet J Gynecol Obstet 2009;14(1). [23] Nakintu N. A comparative study of vaginal misoprostol and intravenous oxytocin for induction of labour in women with intra uterine fetal death in Mulago Hospital, Uganda. Afr Health Sci 2001;1(2):55–9. [24] Niromanesh S, Hashemi-Fesharaki M, Mosavi-Jarrahi A. Second trimester abortion using intravaginal misoprostol. Int J Gynecol Obstet 2005;89(3):276–7. [25] Nyende L, Towobola OA, Mabina MH. Comparison of vaginal and oral misoprostol, for the induction of labour in women with intra-uterine foetal death. East Afr Med J 2004;81(4):179–82. [26] Stibbe KJ, de Weerd S. Induction of delivery by mifepristone and misoprostol in termination of pregnancy and intrauterine fetal death: 2nd and 3rd trimester induction of labour. Arch Gynecol Obstet 2012;286(3):795–6. [27] Yilmaz B, Kelekci S, Ertas IE, Ozel M, Sut N, Mollamahmutoglu L, et al. Randomized comparison of second trimester pregnancy termination utilizing saline moistened or dry misoprostol. Arch Gynecol Obstet 2007;276(5):511–6. [28] Zangeneh M, Malek-Khosravi S, Veisi F, Rezavand N, Rezaee M, Rajatee M. Multipledose vaginal misoprostol and single-dose misoprostol plus oxytocin for termination of second-trimester pregnancy. Int J Gynecol Obstet 2012;117(1):78–80. [29] Winikoff B, Dzuba IG, Creinin MD, Crowden WA, Goldberg AB, Gonzales J, et al. Two distinct oral routes of misoprostol in mifepristone medical abortion: a randomized controlled trial. Obstet Gynecol 2008;112(6):1303–10. [30] Ellis SC, Kapp N, Vragpvoc O, Borgata L. Randomized trial of buccal versus vaginal misoprostol for induction of second trimester abortion. Contraception 2010;81(5): 441–5. [31] Meckstroth KR, Whitaker AK, Bertisch S, Goldberg AB, Darney PD. Misoprostol administered by epithelial routes: Drug absorption and uterine response. Obstet Gynecol 2006;108(3 Pt 1):582–90. [32] Wildschut H, Both MI, Medema S, Thomee E, Wildhagen MF, Kapp N. Medical methods for mid-trimester termination of pregnancy. Cochrane Database Syst Rev 2011;1:CD005216. [33] Lohr PA, Hayes JL, Gemzell-Danielsson K. Surgical versus medical methods for second trimester induced abortion. Cochrane Database Syst Rev 2008;1:CD006714.
