Postpartum hemorrhage is the fourth leading cause of maternal mortality in the United States, accounting for 13.4% of maternal deaths, and there is evidence that major postpartum hemorrhage is on the rise. Morbidity related to postpartum hemorrhage is substantial and includes anemia, a need for transfusion, disseminated intravascular coagulation, hysterectomy, and prolonged hospitalization.

The main cause of postpartum hemorrhage is uterine atony, risk factors for which include uterine overdistention, anesthesia with halogenated agents, rapid or prolonged labor, oxytocin stimulation, chorioamnionitis, and obesity. However, atony often occurs without predisposing risk factors. Active management of the third stage of labor (including prophylactic oxytocin given during or immediately after the third stage of labor) is now used in many deliveries in the developed world because it is effective in lowering the rate of postpartum hemorrhage.

But what happens when uterine atony and postpartum hemorrhage occur despite oxytocin prophylaxis? What medication should be administered next? In some units, misoprostol has become popular. This popularity almost certainly derives from clinical familiarity with misoprostol and its effectiveness for labor induction, medical management of miscarriage, and elective termination of pregnancy, its ease of use, and its low cost. Indeed, in our hospital (Women and Infants Hospital of Rhode Island), 1000 μg of rectal misoprostol is commonly given when postpartum hemorrhage is diagnosed, a practice endorsed by American College of Obstetricians and Gynecologists but only by reference to a non–peer-reviewed article that contains no supporting data.

To evaluate the use of misoprostol for postpartum hemorrhage, we searched PubMed from 1975 to September 2011 for English-language articles using the terms “postpartum hemorrhage,” “misoprostol,” and “cytotec.” We considered only randomized comparative studies.

On the basis of this review, we concluded that the use of misoprostol to control postpartum hemorrhage may be justified in developing countries without access to oxytocin. For example, in a field intervention trial in Tanzania, 1000 μg misoprostol given rectally was superior to no medical intervention in decreasing the need for blood-loss–initiated transfer to hospital facilities after a home birth. Three additional studies have evaluated misoprostol versus placebo in the absence of oxytocin. None showed definitively that misoprostol was statistically superior, but all found decreased rates of hemorrhage with misoprostol. However, these studies are not relevant to the developed world, in which the question is not “in the setting of postpartum hemorrhage, is misoprostol better than nothing?” (which it probably is) but rather “is misoprostol better than an alternative for the treatment of postpartum hemorrhage when prophylactic oxytocin has already been given (and is therefore available)?”

Recently 2 large, randomized studies were reported which answer this question. In the first, Blum et al studied 809 women who experienced postpartum hemorrhage as determined by clinical judgment or estimated blood loss of 700 mL despite prophylactic oxytocin (dosage not specified). Participants were randomized to either 800 μg of sublingual misoprostol or to 40 additional units of oxytocin given in 1 L of intravenous solution over 15 minutes. They reported 2 primary outcomes: cessation of active bleeding within 20 minutes of administration of medication and an additional blood loss of 300 mL or greater after the initiation of treatment. Women allocated to misoprostol were no more likely to cease bleeding within 20 minutes than those allocated to oxytocin, 89% versus 90%. Nor were they less likely to experience blood loss greater than 300 mL, 34% versus 31%. Women allocated to misoprostol were, however, significantly more likely to experience serious ongoing hemorrhage (blood loss greater than 1000 mL after randomization), 3% versus 1% (relative risk [RR], 3.62; 95% confidence interval [CI], 1.02–12.89). Side effects occurred significantly more frequently in the women allocated to misoprostol. Shivering occurred in 37% of women in the misoprostol group and 15% in the oxytocin group (RR, 2.54; 95% CI, 1.95–3.32). Fever to 38°C (100.4°F) occurred in 22% and 15% of the women, respectively (RR, 1.47; 95% CI, 1.09–1.99), and fever greater than 40°C (104°F) in 5% and 1%, respectively, of the women, although this last difference was not statistically significant.

The second trial demonstrates that the addition of misoprostol to oxytocin for management of postpartum hemorrhage is not more effective than simply giving oxytocin alone and that adding misoprostol causes more side effects. Widmer et al randomly allocated 1422 women who had received prophylactic oxytocin (10 U given intramuscularly or via slow intravenous infusion) and were subsequently diagnosed with postpartum hemorrhage to receive an additional 10 U of oxytocin (given intramuscularly or by slow intravenous injection) alone or in combination with 600 μg misoprostol given sublingually. The primary outcome was an estimated blood loss greater than 500 mL 1 hour after randomization. Although admission into the trial was based on subjective clinician assessment of hemorrhage, subsequent blood loss was estimated with the use of a calibrated bedpan. Women who received misoprostol in addition to oxytocin were no less likely to have blood loss greater than 500 mL, 14% versus 14% (RR, 1.02; 95% CI, 0.79–1.32). Shivering occurred in 73% of the women in the misoprostol group and 35% in the oxytocin group (RR, 2.08; 95% CI, 1.86–2.32). Fever to 38°C (100.4°F) occurred in 58% and 19%, respectively (RR, 3.00; 95% CI, 2.55–3.53), and fever greater than 40°C (104°F) in 7% and less than 1%, respectively (RR, 16.21; 95% CI, 5.07–51.78).

These 2 well-done randomized trials with a total enrollment of more than 2200 women diagnosed with postpartum hemorrhage despite oxytocin prophylaxis demonstrate that treating such women with additional oxytocin alone is as effective as or in fact better than misoprostol in stopping their bleeding while sparing them the hyperthermic side effects of misoprostol. The effectiveness of oxytocin to treat postpartum hemorrhage seems to be independent of the dose of prophylactic oxytocin. A blinded randomized trial showed that additional oxytocin to treat postpartum hemorrhage was as effective after high-dose oxytocin prophylaxis (80 U in 500 mL of lactated Ringer’s over 30 minutes) as it was after low-dose prophylaxis (10 U over 30 minutes).

The hyperthermic effects of misoprostol are not only unpleasant, but they may also be life threatening: in one reported case, a woman who received 800 μg of oral misoprostol for postpartum hemorrhage prophylaxis subsequently developed chills, rigors, hyperpyrexia to 41.9°C (107.4°F), and tachycardia to 180 beats per minute. She required ice saline lavage for cooling. Even lesser degrees of misoprostol-associated hyperthermia can be of consequence, in that they can lead to erroneous infection diagnoses and subsequent treatment with antibiotics with the potential sequelae and consequences thereof (eg, allergic reactions, higher costs, and longer hospitalizations).

Unlike misoprostol, at very high doses (up to 100 IU over 1 hour), oxytocin does not raise maternal temperature. Potential side effects of oxytocin include water intoxication and hemodynamic instability, most notably hypotension and tachycardia. However, both of these complications are obviated by infusing oxytocin in isotonic solutions, such as lactated Ringer’s or normal saline as is now common practice. The side effect profiles of misoprostol and oxytocin aside, pharmacokinetics favor oxytocin. Oxytocin has a faster onset of action: 3-5 minutes when given intravenously. In contrast, oral misoprostol has an onset of action of 8 minutes and a duration of action of 2 hours. Sublingual, vaginal, and rectal misoprostol have increasingly slower onsets of action (11, 20, and 100 minutes, respectively) but longer durations of action (3, 4, and 4 hours, respectively). The slow onset of action of rectal misoprostol in particular argues against its use for acute postpartum hemorrhage.

The low cost of misoprostol is often cited as a reason for its use, particularly in areas with limited access to health care. However, at our institution, the hospital cost for a premixed bag of oxytocin 30 U in lactated Ringer’s 500 mL is $6.75, whereas the cost for misoprostol 800 μg is $8.04. Additionally, laboring women at most US hospitals receive intravenous fluids as part of their intrapartum care, thus nullifying the additional cost of intravenous access required for oxytocin infusion.

Even in the absence of prophylactic oxytocin, treatment with oxytocin is superior to misoprostol in arresting postpartum hemorrhage, a fact borne out in a recent trial in which 978 patients were randomly allocated to 800 μg sublingual misoprostol or 40 IU oxytocin in 1 L of normal saline administered intravenously over 15 minutes for the treatment of postpartum hemorrhage. None of the women in this trial received routine prophylactic oxytocin. Fewer women in the misoprostol group stopped bleeding within 20 minutes, 90% versus 96% (RR, 0.94; 95% CI, 0.91–0.98), and more women had an additional blood loss of greater than 300 mL, 30% versus 17% (RR, 1.78; 95% CI, 1.40–2.26). Women in the misoprostol group more often required additional uterotonics, 13% versus 6% (RR, 1.98; CI, 1.31–2.99). Notably, fever greater than 40°C (104°F) occurred in 14% treated with misoprostol and no women who received oxytocin ( P < .0001).

We conclude that oxytocin is more effective than misoprostol in treating postpartum hemorrhage and has a better side effect profile, regardless of whether oxytocin has been given as a prophylactic agent. Unfortunately, there have been no randomized studies of other uterotonic agents for the treatment of postpartum hemorrhage. Case series support the safety and effectiveness of carboprost tromethamine (Hemabate; Pfizer, New York, NY) in appropriately selected patients, and long clinical experience suggests that methylergonovine maleate (Methergine; Novartis Pharmaceuticals Corporation, East Hanover, NJ) is often effective. Given the apparent rise in the incidence of postpartum hemorrhage, randomized trials of these agents are warranted, as are further trials of tranexamic acid, an antifibrinolytic agent approved by the Food and Drug Administration for the management of menorrhagia, which has shown promise in the treatment of postpartum hemorrhage. Despite misoprostol’s ready availability, ease of use, and utility for other pregnancy indications, in the developed world, oxytocin should remain the mainstay of both prophylaxis and first-line medical treatment of postpartum hemorrhage cause by uterine atony, and misoprostol should be used infrequently for this indication.