Figure 1.
Figure 2.
Figure 3.
Cont OB/GYN 48(3):34-53, 2003. © 2003 Medical Economics Company, Inc.
Although you can't always anticipate a potentially catastrophic obstetric hemorrhage, rapid diagnosis and intervention can make all the difference in the world. The authors review interventions -- new and old -- and tell clinicians what to have on hand to implement them.
Massive uncontrolled hemorrhage after childbirth is the leading cause of pregnancy-related death in the United States and one of the most common causes of maternal death.[1] Even though infant mortality has steadily declined since 1982, thanks to modern advances in neonatal intensive care unit technology, maternal mortality in the US has not improved for two decades -- holding steady at 7.7 deaths per 100,000 live births between 1982 and 1996.[2] Of 1,459 reported pregnancy-related deaths between 1987 and 1990, hemorrhage accounted for about three out of every 10 (29%).[3] Moreover, obstetric hemorrhage can cause shock, renal failure, and Sheehan's syndrome (postpartum pituitary necrosis); if the bleeding cannot be stopped in time, hysterectomy is often necessary.
Postpartum hemorrhage (PPH) is traditionally defined as a blood loss of more than 500 mL after vaginal delivery and more than 1,000 mL after cesarean delivery, but intraoperative estimations of blood loss are notoriously inaccurate. Thus, the American College of Obstetricians and Gynecologists defines it as a decrease in hematocrit of more than 10% from before to after delivery.[4]
Although PPH cannot always be anticipated, the cornerstones of effective treatment remain rapid diagnosis and intervention. Our goal here is to review new interventions -- and revisit some old ones -- that can assist in controlling acute hemorrhage.
A prostaglandin E1 analogue, misoprostol is best known in the obstetric community for its use in labor induction and medical abortion. When administered rectally or orally in doses of 400 to 1,000 µg, however, it is rapidly absorbed and can effectively prevent and treat uterine atony.[5-12] The drug's few side effects are limited to shivering and fever in a small percentage of patients. Unlike other prostaglandin agents, misoprostol is not associated with MI and bronchospasm and is therefore safe for asthmatic patients. Similarly, it's also safe for hypertensive and preeclamptic patients, having no effect on blood pressure. By way of contrast, ergometrine is contraindicated in these patient groups. The drug's other significant advantages are its heat-stable preparation (it requires no refrigeration) and its low cost (a little more than $1.00 for one 200-µg tablet).
The pharmacokinetics of vaginally and orally administered misoprostol is well established. The drug is well absorbed by both routes with peak levels seen within 60 minutes.[5] Vaginal administration in patients with PPH has not been evaluated in clinical trials, but it is unlikely that this route would be effective because excessive bleeding probably inhibits the drug's absorption. While there are no studies detailing the pharmacokinetics of rectal absorption, clinical observation has shown that the drug increases muscle tone in the uterus in as little as 3 minutes.[6]
For preventing PPH, misoprostol has been evaluated in five randomized controlled trials and has been shown to be as effective as oxytocin when given rectally in doses of 400 µg or orally at 600 µg.[7-11] Although no RCTs have evaluated misoprostol for acute PPH due to uterine atony, observational studies suggest that it holds promise, particularly when other agents are contraindicated or unavailable. An observational study of 14 patients with PPH who did not respond to oxytocin and ergometrine found that 1,000 µg of misoprostol given rectally had uterotonic effects within 3 minutes of administration and no patient required additional intervention after receiving it.[12]
With these studies in mind, we recommend the following:
In recent years, interest has surged in the surgical compression suture for treating PPH brought on by uterine atony. Of the several different techniques using this approach, the B-Lynch suture initially described by Christopher B-Lynch in 1997 has gained the most popularity (Figures 1, 2, and 3).[13,14]
Figure 1.
Figure 2.
Figure 3.
The theory behind each technique is the same: the mechanical compression of uterine vascular sinuses prevents further engorgement with blood and continued hemorrhage. When used to treat atony and hemorrhage that does not respond to pharmacologic intervention, the B-Lynch appears to be very effective.
To date, this technique has been used in a series of 11 patients, in each case successfully averting hysterectomy.[15-17] Subsequently, two of these women had uncomplicated pregnancies and were delivered at term. One of these patients had an elective C/S, and no adhesions or uterine defects were identified at the time of surgery. A third patient had an MRI and hysterosalpingogram following the B-Lynch suture and was noted to have normal uterine anatomy.
In our combined experience, we have used this technique eight times after traditional interventions failed (methylgonovine maleate, carboprost tromethamine, oxytocin, surgical vessel ligation). One patient ultimately required a hysterectomy for intractable hemorrhage and hemodynamic instability. The other seven responded immediately, and no further interventions were required.
A woman meets the criteria for the B-Lynch compression suture if bimanual compression decreases the amount of uterine bleeding by abdominal and perineal inspection. Although originally described using No. 2 chromic catgut, variations using 0 Vicryl (Ethicon, Somerville, N.J.) suture have been equally successful.[17] In our experience, we have also used No. 1 chromic catgut and 0 loop PDS (polydioxanone surgical) suture with no complications noted. There are theoretical concerns about bowel complications with PDS suture because the suture material may not completely degrade for up to 6 months.
When using the B-Lynch suture technique, which is illustrated in Figures 1, 2, and 3, we recommend using absorbable, as opposed to delayed absorbable, suture material.
For more on this technique, visit www.contemporaryobgyn.net for a Clinical Dialogue (August 1998) in which Ronald A. Chez, MD, interviews Mr. Christopher B-Lynch, who developed the B-Lynch suture for control of massive postpartum hemorrhage.
The rate of blood transfusion during C/S ranges from 1% to 14%.[18-26] Intraoperative blood salvage (IBS) was initially described in 1874 as an alternative to homologous blood transfusion and its attendant risks of infection and transfusion reaction.[27] But the use of cell saver technology during C/S delivery -- intraoperative erythrocyte salvage and autotransfusion -- has been limited by theoretical concerns about amniotic fluid embolism (AFE) and infection.
Although qualitative studies examining the ability of various IBS devices to remove fetal products from salvaged blood have been performed, they are limited by one key fact: To date, the exact cause of AFE is unknown. Nonetheless, various studies have shown that cell saver devices are capable of filtering tissue factor, lamellar bodies, fetal squamous cells, and alpha-fetoprotein.[28-30] But it's also been shown that fetal erythrocytes, which are more difficult to filter, remain in filtered salvaged blood in concentrations higher than in maternal serum.
There's but a single case report of AFE following intraoperative erythrocyte salvage, which involved a patient who was delivered preterm by C/S due to HELLP (Hemolysis, Elevated Liver Enzymes, Low Platelets) syndrome.[31] Cell salvage was performed because the woman was a Jehovah's Witness. Although the estimated blood loss was 600 mL, the decision was made to transfuse the salvaged 200 mL of blood because of a suspected coagulopathy resulting from HELLP syndrome. Approximately 10 minutes after starting the transfusion, the patient developed cardiopulmonary collapse and died. The cause of death was presumed to be AFE, although autopsy results were not specific. While a leukocyte depletion filter was not used in this instance, adding such a filter to the standard IBS device can further reduce the amount of fetal cellular debris in salvaged blood.[29]
In contrast to this single case report, two clinical trials using IBS technology with leukocyte depletion filters found no increased complication rate in patients receiving salvaged blood. Our large, multicenter historical cohort study using Haemonetics Cell Saver found no differences in the rate of adult respiratory distress syndrome (ARDS), AFE, disseminated intravascular coagulation (DIC), need for ventilatory support, infection, or length of hospital stay.[32] In fact, there were no reports of ARDS or AFE in either group. The volumes of autotransfused blood ranged from 125 mL to 11,250 mL in 139 patients.
A subsequent RCT using the Dideco machine -- which involved 34 patients who underwent blood salvage and 34 controls -- produced similar results.[33] The women who had undergone blood salvage required significantly less homologous blood transfusion (1/34 patients in the study group versus 8/34 in the control group) and had significantly shorter postoperative hospital stays. There were no cases of ARDS or AFE in either group.
Women at risk for intraoperative hemorrhage (previa, known accreta, preoperative anemia) who object to homologous blood transfusion may benefit from IBS technology. More importantly, IBS may be lifesaving in remote regions with limited blood banking services. When at all possible, delivery should be planned with the cell saver set up prior to the procedure.
Our recommendations are as follows:
Selective arterial embolization (SAE) has been used in a variety of clinical settings to control hemorrhage. In obstetrics and gynecology, it has been shown effective in the management of ectopic pregnancy, postabortal hemorrhage, malignancy, cervical hemorrhage post-conization, and PPH. Some authors argue that this technique isn't being used enough to combat PPH. With more than 150 cases in the literature, SAE has reported success rates of up to 97%.[34] Table 1 lists the published case series that contain more than five patients and their success rates.[34-41] Selective arterial embolization was considered successful in these cases if bleeding was controlled and hysterectomy was avoided.
The procedure is usually performed by the interventional radiologist under fluoroscopic guidance in the angiography suite. Using regional anesthesia or conscious sedation, the physician introduces a catheter via the femoral artery and directs it into the target vessel -- usually either the hypogastric, uterine, or ovarian artery. With gelfoam pledgets, coils, or a balloon catheter, the targeted artery is occluded.[42] In many cases, both sides of the pelvis can be accessed through a single puncture. Patients usually respond immediately. Unlike other interventions, SAE can be highly effective when coagulopathy is present. Once the patient is in the angiography suite, embolization can usually be completed within 30 minutes. Although long-term follow-up is unavailable for most of the reported cases, menses typically returns within 3 months, and subsequent normal pregnancies have resulted. In our experience, one patient became pregnant following embolization and had no obstetric complications during that pregnancy.
When women are at increased risk for PPH (suspected accreta, previa), catheters can be placed prophylactically, prior to a planned C/S delivery in anticipation of need. One study found that prophylactically placed catheters reduced the total blood loss and incidence of coagulopathy, compared with catheterization performed in an emergent setting.[41]
The procedure's potential complications include fever, buttock ischemia, hematoma, vascular perforation, and infection. One woman developed uterine necrosis and sepsis 53 days following SAE for PPH and required a hysterectomy for definitive treatment.[43] During her initial hemorrhage, the bleeding failed to respond to bilateral uterine artery embolization, so surgical ligation of the uterine and utero-ovarian arteries was performed and that stopped the bleeding. The patient then presented 53 days later with high fever and severe pelvic pain; hysterectomy was performed, and pathology was consistent with uterine necrosis. In another instance, SAE was performed to treat hemorrhage following a cervical cone biopsy.[44] Inadvertently, the external iliac artery was embolized instead of the hypogastric, resulting in subsequent surgery to amputate her leg. This outcome underscores the importance of having an experienced interventional radiologist in attendance.
While the availability of SAE varies from institution to institution, if it is available at your institution, here are some tips to keep in mind:
Often, the consequences of PPH are more severe because of baseline maternal anemia. Erythropoietin (EPO) is a hormone secreted by the kidney that regulates the differentiation and maturation of erythrocytes. In the normal pregnancy, EPO levels rise to two to four times that seen in nonpregnant, nonanemic controls and peak in the third trimester. In the nonpregnant patient, recombinant human erythropoietin (rHuEPO) is used in a variety of settings, including anemia due to chronic renal failure, cancer chemotherapy, and hereditary disorders.
In pregnancy, rHuEPO has been shown to be a safe adjunct to iron therapy in treating severe anemia. With its large molecular weight, rHuEPO does not cross the placenta. Furthermore, it appears to be safe for breastfeeding mothers. Erythropoietic activity begins 36 hours after a single dose, peaks at 84 hours, and fades after 6 days. Side effects are limited to mild flu-like symptoms that resolve with repeat doses. In order for the drug to be effective, the patient must have adequate iron stores and supplementation.
Use of rHuEPO during pregnancy has been reported in women with severe anemia due to renal failure, thalassemia, leukemia, and iron-deficiency anemia. RHuEPO is useful in: (1) the antepartum management of the severely anemic patient at risk for hemorrhage; (2) the antepartum management of the nonanemic woman at risk for hemorrhage to allow for autologous blood transfusion; and, (3) the postpartum augmentation of anemia due to hemorrhage.
In one series of 26 pregnant patients with severe iron deficiency anemia (Hgb <8.5 mg/dL) that had not improved after 8 weeks of iron supplementation alone, 73% achieved normal hemoglobin levels within 2 weeks of therapy (Epogen 150 IU/kg subcutaneously three times weekly plus daily parenteral iron).[45] All patients in this series had low serum iron and ferritin levels before rHuEPO. Each of these women experienced a dramatic increase in reticulocyte count within 7 days of initiating therapy. No maternal or fetal adverse effects were linked to rHuEPO use, although three patients had an allergic reaction that was attributed to the IV iron. It remains unclear, however, why 25% of patients failed to respond with an increase in hemoglobin levels despite IV iron supplementation and an increase in their reticulocyte counts.
We recommend the following:
Although uterine packing was advocated for treating PPH in the past, it fell out of use largely due to concerns of concealed hemorrhage and uterine overdistension. In recent years, however, several modifications of this procedure have allayed these concerns. Balloon tamponade using either a Foley catheter or a Sengstaken-Blakemore tube has been shown to effectively control postpartum bleeding -- and may be useful in several settings: uterine atony, retained placental tissue, and placenta accreta.[46-49]
Both the Foley catheter and the Sengstaken-Blakemore tube have open tips, which permit continuous drainage from the uterus. Furthermore, if the concern for concealed hemorrhage still exists, ultrasound can more effectively detect a developing hematoma when the contrast is a fluid-filled balloon as opposed to blood-saturated gauze. Thus, this technique has the advantage of being not only therapeutic but also diagnostic -- when used in combination with U/S -- in differentiating the various etiologies described above. Additionally, if intrauterine blood loss exceeds 5 cm/sec, the actual site of arterial bleeding can be pinpointed sonographically using power angiography mode against the contrast of the fluid-filled balloon.
The technique is simple. A Foley catheter with a 30-mL balloon capacity is easy to acquire and may routinely be stocked on labor and delivery suites. Using a No. 24F Foley catheter, the tip is guided into the uterine cavity and inflated with 60 to 80 mL of saline. Additional Foley catheters can be inserted if necessary. If bleeding stops, the patient can be observed with the catheters in place and then removed after 12 to 24 hours.
Originally developed for the tamponade of bleeding esophageal varices, the Sengstaken-Blakemore tube has the advantage over the Foley catheter due to the larger capacity of its balloon tip. Like the Foley catheter, the Sengstaken-Blakemore tube has an open tip that permits continuous drainage. But, unlike the Foley catheter, this device may be more difficult to obtain in an emergency setting because it is not routinely stored on labor and delivery suites. Like the Foley catheter, the Sengstaken-Blakemore tube should be guided through the cervix into the uterus and the balloon can then be inflated to achieve the desired tamponade and can be removed in 12 to 24 hours.
We recommend that Sengstaken-Blakemore tubes or 30-mL balloon Foley catheters be made readily available on the labor floor.
Like IBS, this is another technique that can be used when excessive blood loss is anticipated to avoid the need for homologous blood transfusion. The goal of acute normovolemic hemodilution is to reduce the total red cell mass lost intraoperatively by reducing baseline hematocrit.
To date, two case series have been reported that focused on this technique in 40 patients.[50,51] The larger series describes 38 patients in whom the technique was used.[51] All patients included in this series were thought to be at risk of hemorrhage during C/S delivery and had the following diagnoses: placenta previa, large fibroids, placenta accreta, and abnormal placental attachment along a uterine septum. Patients were taken to the operating room approximately 1.5 hours prior to scheduled C/S delivery where phlebotomy of 750 to 1,000 mL of blood was performed through a 16-gauge IV. At the same time, the women underwent infusion of an equal volume of 10% pentastarch in the opposite arm. The collected blood was then transfused after the case was completed or during the case in the event of excessive bleeding.
All patients had a starting hematocrit of at least 29% with a mean hematocrit of 32% and tolerated hemodilution well. None experienced any nausea, vomiting, dizziness, or lightheadedness. Every woman underwent continuous fetal monitoring during phlebotomy, and no fetal heart rate abnormalities were noted. In terms of fetal outcomes, all infants had APGAR scores greater than or equal to eight at 5 minutes and normal venous cord gases (arterial cord gases were not reported). Only one woman -- a patient who had a posterior placenta previa -- required homologous blood transfusion of 2 units. Her estimated blood loss was 3L. Interestingly, 63% of the 38 patients predonated autologous blood (in addition to the blood withdrawn just prior to delivery), and more than half of this group received this blood perioperatively.
With these results in mind, we recommend establishing protocols for normovolemic hemodilution for patients who are at risk for excessive obstetric hemorrhage at the time of C/S.
We've described a variety of novel techniques, in addition to our traditional arsenal of uterotonic medications and surgical vascular ligation procedures, that can be highly successful in managing PPH. Depending on the availability of resources (anesthesia, interventional radiology, cost), most of these approaches can be incorporated into routine practice with little hardship in the majority of modern obstetrical suites. Introducing these techniques can help to lower maternal death rates, minimize hysterectomies and blood transfusions, and improve patient satisfaction.
Series Indication for SAE Materials and methods Success rate Descargues et al.[35] Postpartum hemorrhage Case series 5/7 Borgatta et al.[36] Postabortal hemorrhage Case series 10/11 Deux et al.[37] Postpartum hemorrhage Case series 24/25 Pelage et al.[38] Delayed postpartum hemorrhage unresponsive to drugs or curettage Case series 14/14 Hansch et al.[39] Postpartum hemorrhage Case series 4/5 Pelage et al.[40] Primary postpartum hemorrhage unresponsive to drugs and vaginal packing Case series 24/25 Vendantham et al.[34] Postpartum hemorrhage Literature review 65/67 Alvarez et al.[41] Obstetric hemorrhage: five postpartum hemorrhages two abdominal pregnancies Case series 7/7 Total 153/161 (95%)
SAE -- selective arterial embolization
