Advertisement

A randomized controlled trial of methylergonovine prophylaxis after dilation and evacuation abortion

  • Jennifer L. Kerns
    Correspondence
    Corresponding author.
    Affiliations
    University of California, San Francisco, Department of Obstetrics, Gynecology and Reproductive Sciences; Zuckerberg San Francisco General, 1001 Potrero Avenue, Ward 6D, San Francisco, CA 94110, United States
    Search for articles by this author
  • Geffan Pearlson
    Affiliations
    University of California, San Francisco, Department of Obstetrics, Gynecology and Reproductive Sciences; Zuckerberg San Francisco General, 1001 Potrero Avenue, Ward 6D, San Francisco, CA 94110, United States
    Search for articles by this author
  • Biftu Mengesha
    Affiliations
    University of California, San Francisco, Department of Obstetrics, Gynecology and Reproductive Sciences; Zuckerberg San Francisco General, 1001 Potrero Avenue, Ward 6D, San Francisco, CA 94110, United States
    Search for articles by this author
  • Kristin Harter
    Affiliations
    University of California, San Francisco, School of Pharmacy; Zuckerberg San Francisco General, 1001 Potrero Avenue, Ward 6D, San Francisco, CA 94110, United States
    Search for articles by this author
  • Rebecca A. Jackson
    Affiliations
    University of California, San Francisco, Department of Obstetrics, Gynecology and Reproductive Sciences; Zuckerberg San Francisco General, 1001 Potrero Avenue, Ward 6D, San Francisco, CA 94110, United States
    Search for articles by this author
  • Eleanor A. Drey
    Affiliations
    University of California, San Francisco, Department of Obstetrics, Gynecology and Reproductive Sciences; Zuckerberg San Francisco General, 1001 Potrero Avenue, Ward 6D, San Francisco, CA 94110, United States
    Search for articles by this author

      Abstract

      Objective

      To evaluate the efficacy of intramuscular methylergonovine maleate as prophylaxis against excessive bleeding when given after dilation and evacuation (D&E) at 20–24 weeks.

      Study design

      We performed a randomized, double-blinded, placebo-controlled trial in patients without excessive bleeding requiring intervention after D&E completion. We administered study treatment within one minute of the end of the procedure. We primarily compared outcomes using a composite of indicators of excessive post-procedure blood loss (post-procedure measured blood loss exceeding 125 mL, uterine massage or compression for at least two minutes, administration of additional uterotonic medication, intrauterine balloon tamponade, uterine re-aspiration, blood transfusion, uterine artery embolization, hospital admission for bleeding, or major surgery). Secondary outcomes included individual indicator occurrences, satisfaction, and side effects.

      Results

      From March 3, 2015 to March 31, 2017, we randomized 284 participants (n = 140 methylergonovine, n = 144 placebo), five before we registered the trial with clinicaltrials.gov. Baseline characteristics were similar between groups. The composite outcome occurred in 78 (56%) methylergonovine and 75 (52%) placebo participants (p = 0.5). Methylergonovine recipients required more intrauterine balloon use (n = 20 [14%]) versus placebo (n = 10 [7%]), p = 0.04. We also observed a non-significant trend towards more uterotonic administration (n = 56 [40%] versus n = 43 [30%], p = 0.07) and hospital admissions for bleeding (n = 4 [3%] versus n = 0, p = 0.06) in the methylergonovine group compared to placebo.

      Conclusion

      We observed no improvement in the composite outcome for excessive bleeding with prophylactic post-procedure methylergonovine. In addition, individual excessive bleeding outcomes occurred more frequently in the methylergonovine group, potentially indicating harm with its prophylactic use after D&E.

      Implications

      When administered prophylactically immediately after dilation and evacuation abortion at 20–24 weeks, methylergonovine increases uterine bleeding. Given the lack of data for effectiveness as a prophylactic agent and our findings indicating harm, we do not recommend its use for post-operative prophylaxis.

      Keywords

      1. Introduction

      Clinically significant hemorrhage after routine D&E is infrequent, with studies reporting an incidence of 0.9–10 per 1000 [
      • Kerns J.
      • Steinauer J.
      Management of postabortion hemorrhage: release date November 2012 SFP Guideline #20131.
      ,
      • Altman A.M.
      • Stubblefield P.G.
      • Schlam J.F.
      • Loberfeld R.
      • Osathanondh R.
      Midtrimester abortion with Laminaria and vacuum evacuation on a teaching service.
      ,
      • Peterson W.F.
      • Berry F.N.
      • Grace M.R.
      • Gulbranson C.L.
      Second-trimester abortion by dilatation and evacuation: an analysis of 11,747 cases.
      ]. Defintions of hemorrhage vary with some studies using blood loss and other studies using specific interventions like transfusion, admission or other procedures [
      • Kerns J.
      • Steinauer J.
      Management of postabortion hemorrhage: release date November 2012 SFP Guideline #20131.
      ]. No studies have described the incidence of excessive bleeding. Measures used by providers to decrease or control excessive bleeding after D&E include balloon tamponade, fundal massage, and of uterotonic administration [
      • White K.O.
      • Jones H.E.
      • Shorter J.
      • Norman W.V.
      • Guibert E.
      • Lichtenberg E.S.
      • et al.
      Second-trimester surgical abortion practices in the United States.
      ].
      Most D&Es are performed in an outpatient setting, about half of which have no immediate access to higher level care [
      • White K.O.
      • Jones H.E.
      • Shorter J.
      • Norman W.V.
      • Guibert E.
      • Lichtenberg E.S.
      • et al.
      Second-trimester surgical abortion practices in the United States.
      ]. In addition, because of the lack of providers in many counties, women travel great distances to access D&E and have limited ability to follow up with an experienced provider for complications [
      • Upadhyay U.D.
      • Johns N.E.
      • Barron R.
      • Cartwright A.F.
      • Tapé C.
      • Mierjeski A.
      • et al.
      Abortion-related emergency department visits in the United States: An analysis of a national emergency department sample.
      ]. Because of the landscape in which D&Es are provided, many providers are highly motivated to prevent any concerning bleeding before discharging patients home.
      About one-quarter of D&E providers report routinely using uterotonics before or after procedures over 20 weeks gestation. Methylergonovine is used most commonly, by almost two-thirds of D&E providers [
      • White K.O.
      • Jones H.E.
      • Shorter J.
      • Norman W.V.
      • Guibert E.
      • Lichtenberg E.S.
      • et al.
      Second-trimester surgical abortion practices in the United States.
      ]. Methylergonovine, a first-line agent in treating post-abortion hemorrhage [
      • Kerns J.
      • Steinauer J.
      Management of postabortion hemorrhage: release date November 2012 SFP Guideline #20131.
      ] is an ergot alkaloid with vasospastic and smooth muscle contractile effects [
      • de Groot A.N.
      • van Dongen P.W.
      • Vree T.B.
      • Hekster Y.A.
      • van Roosmalen J.
      Ergot alkaloids. Current status and review of clinical pharmacology and therapeutic use compared with other oxytocics in obstetrics and gynaecology.
      ]. One low-quality study provides evidence that methylergonovine may be associated with fewer first-trimester abortion complications [
      • Sands R.X.
      • Burnhill M.S.
      • Hakim-Elahi E.
      Postabortal uterine atony.
      ]. While there is evidence to guide abortion providers in identifying women at risk of hemorrhage, there is little evidence to guide practice with prevention of bleeding, specifically regarding use of prophylactic uterotonic medications. Given the widespread use of prophylactic methylergonovine without evidence for benefit in preventing excessive bleeding from D&E, we sought to evaluate its effectiveness as a prophylactic agent.

      2. Material and methods

      We conducted a randomized, double-blinded, placebo-controlled trial of methylergonovine versus placebo after D&E at 20–24 weeks. We recruited patients from March 3, 2015 to March 31, 2017 from a hospital-based, outpatient abortion clinic, five of whom we randomized before registration with clinicaltrials.gov (NCT02408965) on March 11, 2015. In preparation for the study, we measured post-D&E blood loss among 25 consecutive patients within this gestational range to define expected post-procedural blood loss. Blood loss ranged from 25 to 170 mL with 95% of patients less than 125 mL; therefore, we used 125 mL as an indicator of excessive bleeding for our study.
      Eligible patients were women ages 18 and older who were between 20 weeks 0 days and 24 weeks 0 days dated by ultrasonography, and who spoke English or Spanish. Exclusion criteria were hypertension on the preoperative or operative day (defined as systolic blood pressure greater than 140 mmHg or diastolic blood pressure greater than 90 mmHg), D&E procedures with more than one day of cervical preparation with dilators, use of protease inhibitors, known coagulopathy, or known or suspected morbidly adherent placenta.
      A research assistant not involved in the study generated a random sequence of numbers from a computer-generated randomization scheme in a 1:1 fashion. We use block randomization in blocks of 10 stratified by two gestational groups, 20 weeks 0 days to 21 weeks 6 days and 22 weeks 0 days to 24 weeks 0 days. The list was electronically delivered to the inpatient pharmacy where pharmacists prepared and dispensed blinded study drugs to research staff. Pharmacists prepared sequentially numbered syringes of methylergonovine maleate 0.2 mg or saline placebo as 1 mL in a 3 mL syringe, and labeled as study drug. All clinic staff, research staff, D&E providers, and study participants remained blinded to the study drug assignment.
      We consented participants after ultrasound examination, counseling, and consent for the abortion, and before osmotic dilator placement at the pre-operative visit (day 1). On the procedure day (day 2), experienced physicians, including attendings and second-year family planning fellows, did all D&E procedures. The standard D&E procedure included a paracervical block of chloroprocaine 1% 20 mL with vasopressin 4 units, mechanical cervical dilation if necessary, amniotic fluid evacuation with suction, forceps evacuation of fetal parts, then placenta, sharp curettage, and final suction. Most patients over 20 weeks in our hospital-based clinic receive anesthesiologist-administered deep sedation with Propofol, with the remainder receiving nurse-administered moderate sedation with fentanyl and midazolam.
      At the end of the D&E, defined as the end of final suction, we assessed participants’ bleeding and blood pressure. We enrolled (provided study treatment) to participants not requiring treatment for active bleeding and without elevated blood pressure. Within one minute of the end of the procedure, a nurse administered the next sequential study drug intramuscularly in the participant’s right thigh. Treating physicians were allowed to administer additional uterotonics after study drug administration at their discretion. Participants recovered for approximately one hour after the D&E, consistent with clinic standard of care. We measured post-procedure blood loss in the recovery room by weighing pads and linens in addition to a semi-quantitative assessment of blood loss in the toilet.
      We collected baseline demographics, gestational duration, obstetric and medical histories, cervical preparation method (number of osmotic dilators and adjunctive medications used), baseline blood pressure, anesthesia type, need for mechanical cervical dilation, length of D&E procedure, difficulty of procedure (reported by the D&E provider as a 4-point Likert scale), and procedural measured blood loss [
      • Serapio E.T.
      • Pearlson G.A.
      • Drey E.A.
      • Kerns J.L.
      Estimated versus measured blood loss during dilation and evacuation: an observational study.
      ].
      Our primary outcome was a composite of post-procedure excessive bleeding indicators. This composite outcome included post-procedure blood loss greater than 125 mL, uterine massage or compression for at least two minutes, administration of additional uterotonic medication, intrauterine balloon tamponade, uterine re-aspiration, blood transfusion, uterine artery embolization, hospital admission for bleeding, or major surgery. Any one indicator met criteria for excessive bleeding. Secondary outcomes included each individual outcome in the composite, total number of interventions for excessive bleeding per participant, provider and nurse satisfaction with post-abortion bleeding, and participant side effects after study drug injection.
      To estimate sample size, we first reviewed our own database of cases at 20–24 weeks from September–December 2014; among 162 cases, we found that 35% met criteria for the composite outcome. We assumed a 15% reduction of excessive bleeding in the methylergonovine group (20%) versus placebo (35%), consistent with another study evaluating interventions for bleeding at the time of D&E that used a 15% reduction as the basis for sample size [
      • Micks E.
      • Edelman A.
      • Botha R.
      • Bednarek P.
      • Nichols M.
      • Jensen J.
      The effect of sevoflurane on interventions for blood loss during dilation and evacuation procedures at 18–24 weeks of gestation: a randomized controlled trial.
      ]. We estimated 284 patients would be needed using a two-sided alpha of 0.05 and 80% power.
      A data safety monitoring board reviewed interim results after we enrolled one-third of participants. Their criteria for stopping included any serious bleeding outcomes that were statistically significantly different between groups. We used an intent-to-treat analysis and conducted the statistical analysis in blinded fashion. We evaluated demographic characteristics and secondary outcomes using chi-square and Fisher’s exact tests as appropriate for categorical variables, as well as t-tests and Kruskall–Wallis tests for continuous variables, depending on normality of distribution. We calculated relative risks and 95% confidence intervals for the primary composite outcome to compare methylergonovine and placebo groups both in the overall study population and gestational subgroups. We did sensitivity analyses excluding the outcome of uterine massage and compression as criteria for the composite outcome to evaluate any effect of the study drug on more invasive interventions for excessive bleeding. We considered p-values less than 0.05 to be statistically significant. We performed all statistical analyses using Stata version 14.1 (College Station, TX USA). Subejcts were compensated for participation. The study was approved by the University of California, San Francisco (UCSF) Committee on Human Research.
      Clinical Trial Registration: ClinicalTrials.gov, registration number NCT02408965.

      3. Results

      During the study period, we assessed 1384 subjects for eligibility. We consented 320 subjects and enrolled 284 (Fig. 1). Only 14 (4.4%) consented subjects had significant bleeding at the end of the procedure which excluded enrollment. Subject characteristics are presented in Table 1. While mean intraoperative blood loss was similar between groups, the methylergonovine group had a higher proportion of cases with intraoperative blood loss greater than 90% of the mean (n = 28 [20%] versus n = 13 [9%], p = 0.03).
      Figure thumbnail gr1
      Fig. 1Study flow of subjects randomized to post-operative prophylactic intramuscular methylergonovine or placebo for dilation and evacuation abortion at 20–24 weeks.
      Table 1Demographic and procedural characteristics of women randomized to post-operative prophylactic intramuscular methylergonovine or placebo for dilation and evacuation abortion at 20–24 weeks.
      Methylergonovine

      n = 140
      Placebo

      n = 144
      p-value
      Age24.5 (21–29)25 (21–31)0.6
      Race/Ethnicity0.7
       Non-Hispanic White39 (28)48 (33)
       Non-Hispanic Black26 (19)33 (23)
       Hispanic37 (27)33 (23)
       Asian7 (5)7 (5)
       Multiracial19 (13)15 (10)
       Other11 (8)8 (6)
      Body mass index26.5 (23–31)26 (23–29)0.3
      Nulliparous35 (25)37 (26)0.9
      Prior cesarean section24 (17)27 (19)0.7
      Current or recent cocaine or methamphetamine use11 (8)19 (13)0.1
      History of bleeding disorder
      Missing data, n = 9.
      2 (1)3 (2)0.7
      Gestation, weeks21.6 (21–23)21.5 (20–23)0.4
      Cervical preparation
      Missing data, n = 1.
      0.6
       Dilators only106 (76)108 (75)
       Dilators + mifepristone33 (24)35 (24)
       Dilators + misoprostol0 (0)1 (1)
       Needed mechanical dilation7 (5)5 (3)0.5
      Procedural anesthesia0.8
       Deep sedation125 (89)125 (87)
       Moderate sedation13 (10)17 (12)
       General anesthesia2 (1)2 (1)
      Length of procedure, minutes12 (8–16)12 (9–15)0.4
      Procedural measured blood loss300 (150–400)250 (150–425)0.8
      All data presented as n (%) or median (interquartile range).
      a Missing data, n = 9.
      b Missing data, n = 1.
      Table 2 demonstrates bleeding outcomes with study treatment. The composite outcome of excessive bleeding occurred in 78 (56%) of the methylergonovine group and 75 (52%) of the placebo group. Our sensitivity analysis did not show any difference when we excluded massage or compression from the composite outcome. Because intraoperative blood loss greater than 90th percentile of the mean was different between groups, we did an additional analysis controlling for intraoperative blood loss; the association between the intervention and the composite outcome was unchanged. The methylergonovine group had significantly more use of intrauterine balloon tamponade compared to placebo (n = 20 [14%] versus n = 10 [7%], p = 0.04) and a trend towards more frequent administration of at least one uterotonic medication (n = 56 [40%] versus 43 [30%], p = 0.07). In addition, the methylergonovine group experienced more admissions for bleeding (n = 4 [3%] versus n = 0, p = 0.05). Two subjects underwent uterine artery embolization, both in the methylergonovine group. The mean number of excessive bleeding outcomes was higher in the methylergonovine group versus placebo (1.3 versus 0.9, p = 0.008). While the proportions of subjects who received additional uterotonics was not significantly different between groups, the mean number of uterotonics used was higher in the methylergonovine group versus placebo (0.9 versus 0.6, p = 0.01). We observed no differences between groups in time to first or second uterotonic administration.
      Table 2Effect of post-operative prophylactic intramuscular methylergonovine versus placebo on excessive bleeding after dilation and evacuation abortion.
      Methylergonovine

      n = 140
      Placebo

      n = 144
      RR (95%CI)p-value
      Composite outcome
      Composite outcome includes post-procedure measured blood loss greater than 125 mL, uterine massage or compression for at least two minutes, administration of additional uterotonic medication, intrauterine balloon tamponade, uterine re-aspiration, blood transfusion, uterine artery embolization, hospital admission for bleeding, or major surgery.
      78 (56)75 (52)1.1 (0.9–1.3)0.5
      Balloon tamponade20 (14)10 (7)2.1 (1.0–4.2)0.04
      Post-procedure blood loss > 125 cc32 (23)23 (16)1.4 (0.9–2.3)0.1
      Post-procedure blood loss126 ± 24176 ± 690.02
      Massage/compression51 (36)46 (32)1.1 (0.8–1.6)0.4
      Uterotonic given56 (40)43 (30)1.3 (0.97–1.8)0.07
      Reaspiration4 (2.9)00.06
      Admission for bleeding4 (3)00.06
      Transfusion3 (2)00.1
      Uterine artery embolization2 (1)00.2
      Major surgery00
      All data presented as n (%) or mean ± standard deviation.
      a Composite outcome includes post-procedure measured blood loss greater than 125 mL, uterine massage or compression for at least two minutes, administration of additional uterotonic medication, intrauterine balloon tamponade, uterine re-aspiration, blood transfusion, uterine artery embolization, hospital admission for bleeding, or major surgery.
      Table 3 delineates bleeding outcome by gestational age. In the 22–24 week gestational duration group, methylergonovine was associated with more excessive bleeding outcomes, including increased use of intrauterine balloon tamponade (n = 13 [19%] versus n = 5 [7%], p = 0.04), and more frequent administration of additional uterotonic medication (n = 33 [47%] versus n = 22 [31%], p = 0.04).
      Table 3Effect of post-operative prophylactic methylergonovine versus placebo on excessive bleeding after dilation and evacuation abortions, by gestational group.
      20–22 week gestational group22–24 week gestational group
      Methylergonovine

      n = 70
      Placebo

      n = 72
      RR (95% CI)Methylergonovine

      n = 70
      Placebo

      n = 72
      RR (95% CI)
      Composite outcome
      Composite outcome includes post-procedure measured blood loss greater than 125 mL, uterine massage or compression for at least two minutes, administration of additional uterotonic medication, intrauterine balloon tamponade, uterine re-aspiration, blood transfusion, uterine artery embolization, hospital admission for bleeding, or major surgery.
      34 (48)37 (51)0.9 (0.7–1.3)45 (64)40 (55)1.2 (0.8–1.5)
      Balloon tamponade7 (10)5 (7)1.4 (0.5–4.3)13 (19)5 (7)2.7 (1.0–7.1)
      Post-procedure blood loss >125 cc14 (20)12 (17)1.2 (0.6–2.4)19 (27)12 (17)1.6 (0.9–3.1)
      Massage or compression18 (26)23 (32)0.8 (0.5–1.3)33 (47)23 (32)1.4 (1.0–2.2)
      Uterotonic given23 (33)21 (29)1.1 (0.7–1.8)33 (47)22 (31)1.5 (1.0–2.4)
      Reaspiration1 (1)1 (1)3 (4)0
      Admission for bleeding2 (3)02 (3)0
      Transfusion1 (1)02 (3)0
      Uterine artery embolization1 (1)01 (1)0
      All data are presented as n (%).
      a Composite outcome includes post-procedure measured blood loss greater than 125 mL, uterine massage or compression for at least two minutes, administration of additional uterotonic medication, intrauterine balloon tamponade, uterine re-aspiration, blood transfusion, uterine artery embolization, hospital admission for bleeding, or major surgery.
      Compared to placebo, participants randomized to methylergonovine reported more nausea (n = 42 [30%] versus n = 29 [12%], p < 0.001). We observed a trend towards increased vomiting (n = 18 [13%] versus n = 9 [6%], p = 0.06 and cramping pain (n = 99 [71%] versus n = 86 [58%], p = 0.07) in the methylergonovine group; however, most participants reported they were not significantly bothered by these side effects. Overall, subjects reported headache (n = 37 [13%]), shortness of breath (n = 17 [6%]), vision changes (n = 10 [3.5%]), and palpitations (n = 3 [1%]), with no differences between groups.
      Satisfaction outcomes are presented in online Appendix 1. Physicians were less likely to report being satisfied with post-procedure bleeding in the methylergonovine group, both overall (p = 0.03) and in the 20–22 week gestational duration group (p = 0.01). Nurses’ and subjects’ satisfaction with bleeding were similar between groups, as was surgeon-rated procedure difficulty. Physicians correctly guessed group assignment in 8% of cases, nurses in 30% of cases, and subjects in 8% of cases.

      4. Discussion

      We found that methylergonovine, when given as a prophylactic medication immediately after D&E at 20–24 weeks, is associated with increased rather than decreased bleeding. Providers commonly use methylergonovine after D&E as a prophylactic medication in an attempt to prevent excessive bleeding; however, we demonstrated increased admissions for bleeding, increased use of balloon tamponade, and increased use of additional uterotonic medications to control bleeding. Physicians were less satisfied with post-D&E bleeding in patients who received prophylactic methylergonovine after D&E, and patients experienced more side effects.
      Much of our understanding of the actions of methylergonovine comes from obstetric literature. Given our divergent findings, we question whether methylergonovine may have different actions on the uterus at 20–24 weeks than it does at term. Methylergonovine has vasoconstrictive, vasospastic, and uterine-specific smooth muscle contractile properties [
      • Vallera C.
      • Choi L.O.
      • Cha C.M.
      • Hong R.W.
      Uterotonic medications: oxytocin, methylergonovine, carboprost, misoprostol.
      ]. The uterine contractile effect decreases uteroplacental blood flow and closes open vasculature in the vacant postpartum placental bed, which results in a reduction in bleeding [
      • Cunningham F.G.
      • Leveno K.J.
      • Bloom S.L.
      • Dashe J.S.
      • Hoffman B.L.
      • Casey B.M.
      • et al.
      Implantation and placental development [Internet].
      ]. We hypothesize that the effects of methylergonovine may be modulated by the fewer number of available serotonin receptors on the uterus that the medication targets, specifically of the 5HT2 subtype [
      • Karlsson C.
      • Bodelsson G.
      • Bodelsson M.
      • Stjernquist M.
      5-Hydroxytryptamine contracts human uterine artery smooth muscle predominantly via 5-HT2 receptors.
      ,
      • Minosyan T.Y.
      • Lu R.
      • Eghbali M.
      • Toro L.
      • Stefani E.
      Increased 5-HT contractile response in late pregnant rat myometrium is associated with a higher density of 5-HT2A receptors.
      ]. With fewer uterine receptors, methylergonovine’s vasoconstrictive and vasospatic properties and subsequent increase in central venous blood pressure may overpower the reduction in uterine contractile activity. This unchecked vasoconstrictive and vasospastic activity on the open vasculature of a post-D&E placental bed along with an increase in central venous pressure may increase blood flow out of the open vasculature, thereby causing greater blood loss. Supporting this hypothesis, the trauma literature describes reduced blood loss by employing hypotensive resuscitation during hemorrhagic shock, whereby a patient’s blood pressure is maintained at a level low enough to reduce bleeding, but high enough for perfusion [
      • Morrison C.A.
      • Carrick M.M.
      • Norman M.A.
      • Scott B.G.
      • Welsh F.J.
      • Tsai P.
      • et al.
      Hypotensive resuscitation strategy reduces transfusion requirements and severe postoperative coagulopathy in trauma patients with hemorrhagic shock: preliminary results of a randomized controlled trial.
      ]. Lastly, the vasoactive effects of methylergonovine may have an additive bleeding effect with the already dilated low-resistance uterine spiral arteries that have little smooth muscle [
      • Cunningham F.G.
      • Leveno K.J.
      • Bloom S.L.
      • Dashe J.S.
      • Hoffman B.L.
      • Casey B.M.
      • et al.
      Obstetrical Hemorrhage [Internet].
      ].
      The lack of benefit from methylergonovine after D&E in our trial may be related to the timing of administration – specifically immediately after placental removal which is the end of the D&E and not before placental removal. A systematic review of methylergonovine as a prophylactic medication for postpartum hemorrhage (PPH) found no effect on preventing PPH when administered after placental removal. Furthermore, a different systematic review found that methylergonovine was effective in decreasing PPH when it was given in the third stage of labor, before the placenta was removed [
      • Liabsuetrakul T.
      • Choobun T.
      • Peeyananjarassri K.
      • Islam Q.M.
      Prophylactic use of ergot alkaloids in the third stage of labour.
      ]. We cannot determine from our study whether the effect of methylergonovine on post-D&E bleeding would be different had the drug been administered before placenta was removed.
      Finally, we must consider that methylergonovine may act differently when administered prophylactically versus as a treatment intervention. Robust evidence demonstrates that methylergonovine is effective in obstetrics as both a prophylactic and treatment medication. By contrast, our study demonstrates that methylergonovine is harmful when used prophylactically after D&E. While we currently have no reason to believe that methylergonovine is harmful or ineffective as a treatment medication for post-D&E bleeding, the rationale for its use is based on obstetric literature and not from any abortion-specific studies.
      The strengths of our study include a robust randomization and blinding scheme whereby clinicians, researchers and subjects were blinded to assignment. We had no loss to follow up. To our knowledge, this is the first study to examine the prophylactic effect of a commonly used medication after D&E, and our findings may change routine practice. Limitations include the timing of prophylactic methylergonovine at the time of D&E. Specifically, our study only evaluated its use after D&E completion. Our findings may not be applicable to cases with administation before D&E, or before the placenta is removed. Furthermore, our findings may not be applicable to the use of prophylactic methylergonovine much longer after D&E completion, such as in the recovery room or at home.
      D&E providers commonly use prophylactic uterotonic medications after D&E to avoid excessive bleeding. When administered prophylactically as an intramuscular injection immediately after D&E at 20–24 weeks, methylergonovine actually causes harm in addition to side effects and should not be used. Studies evaluating the effect of other uterotonic agents are warranted to better guide D&E practices.

      Appendix A. Supplementary data

      The following are the Supplementary data to this article:

      References

        • Kerns J.
        • Steinauer J.
        Management of postabortion hemorrhage: release date November 2012 SFP Guideline #20131.
        Contraception. 2013; 87: 331-342
        • Altman A.M.
        • Stubblefield P.G.
        • Schlam J.F.
        • Loberfeld R.
        • Osathanondh R.
        Midtrimester abortion with Laminaria and vacuum evacuation on a teaching service.
        J Reprod Med. 1985; 30: 601-606
        • Peterson W.F.
        • Berry F.N.
        • Grace M.R.
        • Gulbranson C.L.
        Second-trimester abortion by dilatation and evacuation: an analysis of 11,747 cases.
        Obstet Gynecol. 1983; 62: 185-190
        • White K.O.
        • Jones H.E.
        • Shorter J.
        • Norman W.V.
        • Guibert E.
        • Lichtenberg E.S.
        • et al.
        Second-trimester surgical abortion practices in the United States.
        Contraception. 2018; 98: 95-99
        • Upadhyay U.D.
        • Johns N.E.
        • Barron R.
        • Cartwright A.F.
        • Tapé C.
        • Mierjeski A.
        • et al.
        Abortion-related emergency department visits in the United States: An analysis of a national emergency department sample.
        BMC Med. 2018; 16: 88
        • de Groot A.N.
        • van Dongen P.W.
        • Vree T.B.
        • Hekster Y.A.
        • van Roosmalen J.
        Ergot alkaloids. Current status and review of clinical pharmacology and therapeutic use compared with other oxytocics in obstetrics and gynaecology.
        Drugs. 1998; 56: 523-535
        • Sands R.X.
        • Burnhill M.S.
        • Hakim-Elahi E.
        Postabortal uterine atony.
        Obstet Gynecol. 1974; 43: 595-598
        • Serapio E.T.
        • Pearlson G.A.
        • Drey E.A.
        • Kerns J.L.
        Estimated versus measured blood loss during dilation and evacuation: an observational study.
        Contraception. 2018; 97: 451-455
        • Micks E.
        • Edelman A.
        • Botha R.
        • Bednarek P.
        • Nichols M.
        • Jensen J.
        The effect of sevoflurane on interventions for blood loss during dilation and evacuation procedures at 18–24 weeks of gestation: a randomized controlled trial.
        Contraception. 2015; 91: 488-494
        • Vallera C.
        • Choi L.O.
        • Cha C.M.
        • Hong R.W.
        Uterotonic medications: oxytocin, methylergonovine, carboprost, misoprostol.
        Anesthesiol Clin. 2017; 35: 207-219
        • Cunningham F.G.
        • Leveno K.J.
        • Bloom S.L.
        • Dashe J.S.
        • Hoffman B.L.
        • Casey B.M.
        • et al.
        Implantation and placental development [Internet].
        Williams Obstetrics, 25e. McGraw-Hill Education, New York, NY2018 ([cited 2019 Sep 30]. Available from: accessmedicine.mhmedical.com/content.aspx?aid=1160772378)
        • Karlsson C.
        • Bodelsson G.
        • Bodelsson M.
        • Stjernquist M.
        5-Hydroxytryptamine contracts human uterine artery smooth muscle predominantly via 5-HT2 receptors.
        Hum Reprod. 1997; 12: 361-367https://doi.org/10.1093/humrep/12.2.361
        • Minosyan T.Y.
        • Lu R.
        • Eghbali M.
        • Toro L.
        • Stefani E.
        Increased 5-HT contractile response in late pregnant rat myometrium is associated with a higher density of 5-HT2A receptors.
        J Physiol (Lond). 2007; 581: 91-97
        • Morrison C.A.
        • Carrick M.M.
        • Norman M.A.
        • Scott B.G.
        • Welsh F.J.
        • Tsai P.
        • et al.
        Hypotensive resuscitation strategy reduces transfusion requirements and severe postoperative coagulopathy in trauma patients with hemorrhagic shock: preliminary results of a randomized controlled trial.
        J Trauma. 2011; 70: 652-663https://doi.org/10.1097/TA.0b013e31820e77ea
        • Cunningham F.G.
        • Leveno K.J.
        • Bloom S.L.
        • Dashe J.S.
        • Hoffman B.L.
        • Casey B.M.
        • et al.
        Obstetrical Hemorrhage [Internet].
        Williams Obstetrics. 25e. McGraw-Hill Education, New York, NY2018 ([cited 2019 Sep 28]. Available from: accessmedicine.mhmedical.com/content.aspx?aid=1160784056)
        • Liabsuetrakul T.
        • Choobun T.
        • Peeyananjarassri K.
        • Islam Q.M.
        Prophylactic use of ergot alkaloids in the third stage of labour.
        Cochrane Database Syst Rev. 2007; : 1-85https://doi.org/10.1002/14651858.CD005456.pub2