Chapter 44. Management of Obstetrical Hemorrhage. Will Obs.

 Management of Obstetrical Hemorrhage

BS. Nguyễn Hồng Anh

Recognition o obstetrical hemorrhage severity is crucial to its management. However, visual estimates, especially when bloo losses are excessive, are notoriously inaccurate. In many cases, true volume losses are oten two to three times the clinical estimate. Moreover, in obstetrics, part an sometimes even all o the lost bloo may be conceale. Estimation is urther complicate in that peripartum hemorrhage also inclues the pregnancy-augmente bloo volume. Ater pregnancy hypervolemia is lost at elivery, bloo loss can be estimate by calculating 500 mL loss or each 3 volume percent rop in hematocrit. Its nair epens on the spee o resuscitation with intravenous crystallois an bloo proucts. With ongoing blood loss, the real-time hematocrit is at its maximum whenever measured in the delivery, operating, or recovery room. Pruently, i bloo loss is consiere more than average, the hematocrit is etermine, an plans are mae or close observation or potential physiological eterioration. Bloo loss etermination as recommene by the American College o Obstetricians an Gynecologists (2019b) is iscusse in etail in Chapter 42. Urine output measure hourly is one o the most important “vital signs.” Unless diuretic agents are given— and these are seldom indicated with active bleeding—accurately measured urine ow reects renal perusion. Tis in turn reects perusion o other vital organs. Te volume o urine output shoul be ≥30 mL/hr an preerably ≥50 mL/hr

Another important actor to consier with management o hemorrhage is whether there are aequate procoagulants to achieve clot ormation an stability. Many cases o severe obstetrical hemorrhage are urther complicate by isseminate intravascular coagulation (DIC), in which bloo has ysunctional coagulation (p. 775).

MANAGEMENT OF HEMORRHAGE

Hypovolemic Shock

Shock rom hemorrhage evolves through several stages (Cannon, 2018). Early, the mean arterial pressure, stroke volume, cariac output, central venous pressure, an pulmonary capillary wege pressure ecline. Greater ierences in arteriovenous oxygen content values reect enhance tissue oxygen extraction, although overall oxygen consumption alls. Bloo ow to capillary bes is controlle by arterioles, which are resistance vessels an partially controlle by the central nervous system (CNS). However, approximately 70 percent o total bloo volume is containe in venules, which are passive resistance vessels controlle by humoral actors. Tus, the catecholamine release uring hemorrhage prompts greater venular tone, an this provies an autotransusion rom this capacitance reservoir. Tis volume boost is accompanie by compensatory rises in heart rate, systemic an pulmonary vascular resistance, an myocarial contractility. At the same time, selective, CNS-meiate arteriolar constriction or relaxation, terme autoregulation, preerentially reistributes cariac output an bloo volume. Tus, more bloo ow is iverte to the heart, brain, an arenal glans, whereas perusion to the kineys, splanchnic bes, muscles, skin, an uterus is relatively iminishe.

When the bloo volume ecit excees approximately 25 percent, compensatory mechanisms usually are inaequate to maintain cariac output an bloo pressure. Importantly, aitional small losses o bloo will now cause rapi clinical eterioration. Following an initial augmente total oxygen extraction by maternal tissue, malistribution o bloo ow results in local tissue hypoxia an metabolic aciosis. Tis creates a vicious cycle o vasoconstriction, organ ischemia, an cellular eath.

Hemorrhage also activates lymphocytes an monocytes, which in turn prompts enothelial cell activation an platelet aggregation (Cannon, 2018). Tis enotheliopathy promotes release o vasoactive meiators that occlue small vessels an urther impair microcirculatory perusion. Comorbi preeclampsia or sepsis also leas to loss o capillary enothelial integrity, aitional loss o intravascular volume into the extracellular space, an platelet aggregation. Tese then can incite DIC.

Te pathophysiological events just escribe create important but oten overlooke extracellular ui an electrolyte shits involve in both the genesis an successul treatment o hypovolemic shock. Tese inclue changes in the cellular transport o various ions such as soium an water into skeletal muscle as well as potassium loss. Replacement o extracellular ui an intravascular volume are both necessary. Patient survival rates are enhance in acute hemorrhagic shock i bloo plus crystalloi solution are given compare with bloo trans- usions alone.

■ Fluid Resuscitation

Whenever excessive bloo loss is suspecte, steps are simultaneously taken to ientiy the bleeing source an to begin resuscitation. Reer to the algorithm or hemorrhage management (Fig. 42-3, p. 735). I the woman is unelivere, restoration o bloo volume benets both mother an etus. It also prepares or emergent elivery. I she is postpartum, immeiately ientiying uterine atony, retaine placental ragments, or genital tract lacerations is essential. One or two large-bore intravenous inusion systems are ieally establishe promptly, crystalloi solutions are rapily inuse, an bloo proucts are orere.

An operating room is reaie, an a surgical an anesthesia team are quickly assemble. Specic management o hemorrhage is urther epenent on its etiology. A neonatal resuscitation team is inclue i imminent elivery is planne.

Serious hemorrhage emans prompt an aequate relling o the intravascular compartment with crystalloi solutions. Tese rapily equilibrate into the extravascular space, an only 20 percent o crystalloi remains intravascular in critically ill patients ater 1 hour (Zuckerbraun, 2010). Because o this, initial uid is inused in a volume two to three times the estimated blood loss.

Crystalloid Versus Colloid Solutions

Resuscitation o hypovolemic shock with colloi versus crystalloi solutions has been ebate. In a Cochrane review o resuscitation o nonpregnant critically ill patients, Lewis an coworkers (2018) oun selection o collois versus crystallois probably makes little or no ierence to mortality rates. Similar results were oun in the Saline versus Albumin Flui Evaluation (SAFE) ranomize trial o almost 7000 nonpregnant patients (Finer, 2004). At Parklan Hospital, acute volume resuscitation is preerably one with crystalloi solutions an bloo.

Either a saline-base or a balance crystalloi solution may be inuse. Te latter inclues Ringer lactate an PlasmaLyte solutions, which have electrolyte compositions similar to plasma. In some stuies, but not in all, balance crystalloi solutions were oun to be superior to saline-base ones (Semler, 2018; Yule, 2020). Excessive saline-base crystalloi solution can theoretically lea to hyperchloremic aciosis. In comparisons, the ifculty lies in separating the eects o the unerlying pathophysiology riving the aciosis.

■ Blood Replacement

Type and Screen Versus Crossmatch

A bloo type an antiboy screen shoul be perorme or any woman at signicant risk or hemorrhage. Because preiction o women at risk or postpartum hemorrhage is poor, we per- orm type an screen in all women upon amission to labor an elivery. Screening involves mixing maternal serum with stanar reagent re cells that carry antigens to which most o the common clinically signicant antiboies react. Instea, crossmatching involves the use o actual onor erythrocytes rather than the stanarize re cells. Tis process is efcient, an only 0.03 to 0.07 percent o patients ientie as having no antiboies are subsequently oun to them (Boral, 1979). Importantly, administration o screened blood rarely results in adverse clinical sequelae.

Transfusion Thresholds

Te hematocrit level or hemoglobin concentration threshol that manates bloo transusion is controversial (Bienstock, 2021). Cariac output oes not substantively rop until the hemoglobin concentration alls to approximately 7 g/L or when the hematocrit approximates 20 volume percent. Military combat trauma units have use a target hematocrit o 21 volume percent, an at this level, most experts recommen consieration o re cell transusion (Carson, 2017; Kogutt, 2019). In general, with ongoing obstetrical hemorrhage, we recommen rapi bloo inusion when the hematocrit is <25 volume percent. Other ecisionmoiying actors are whether the etus is unelivere; surgery is imminent or ongoing operative bloo loss is expecte; or maternal hypoxia, vascular collapse, or other actors are present. Limite ata aress these issues. In a stuy rom the Cana- ian Critical Care rials Group, nonpregnant patients were ranomly assigne to restrictive re cell transusions to maintain hemoglobin concentration >7 g/L or to liberal transusions to maintain the hemoglobin level at 10 to 12 g/L. Te 30-ay mortality rate was similar—19 versus 23 percent—in the restrictive versus liberal groups, respectively (Hébert, 1999). ransusion therapy in nonpregnant patients with septic shock ha similar mortality rates when 7 g/L was compare with 9 g/L as targets or transusions (Holst, 2014). Te number o units transused in a given woman to reach a target hematocrit depends on her body mass and on the expectations o additional blood loss.

Blood Component Products

Te content an eects o transusion o various bloo components are shown in Table 44-1. Because whole bloo is rarely available, most women with obstetrical hemorrhage an ongoing massive bloo loss are given packe re cells, crystalloi solutions, an bloo components. As subsequently iscusse, many institutions use massive transusion protocols (MPs) esigne to anticipate all acets o massive obstetrical hemorrhage. Tese protocols commonly contain plasma, cryoprecipitate, an platelets in various ratios (Cunningham, 2015; Shiels, 2011).

Several stuies have assesse plasma:re cell ratios with MPs use in civilian trauma units an military combat hospitals (Har- in, 2014; Rahouma, 2018). Patients receiving a massive transusion—ene as 10 or more units o bloo—ha much higher survival rates as the ratio o plasma to re cell units neare 1:1.4, that is, one unit o plasma given or each 1.4 units o packe re cells. By way o contrast, the highest mortality group ha a ratio o 1:8.

From the oregoing, when re cell replacement excees ve units or so, evaluation o platelet count, clotting stuies, an plasma brinogen concentration is reasonable (Bienstock, 2021; Pacheco, 2019). In the woman with obstetrical hemorrhage, the platelet count shoul be maintaine >50,000/µL by the inusion o platelet concentrates. A brinogen level <150 mg/L or a sufciently prolonge P or P in a woman with surgical bleeing is an inication or replacement. Fresh-rozen plasma is aministere in oses o 10 to 15 mL/kg, or alternatively, cryoprecipitate is inuse (see able 44-1).

At Parklan Hospital, we use a stanarize response to active hemorrhage. I whole bloo is not available, the ollowing escalation o bloo proucts is applie: 2 units packe re bloo cells, an i continue bleeing, 2 units packe re bloo cells an 2 units resh-rozen plasma are automatically orere. Beyon this, MP is initiate.

Whole Blood and Packed Red Blood Cells

Compatible whole bloo is ieal or management o severe obstetrical hemorrhage. It has a shel lie o 40 ays, an 70 percent o the transuse re cells unction or at least 24 hours ollowing transusion. One unit raises the hematocrit by 3 to 4 volume percent. Important or obstetrical hemorrhage, whole bloo replaces many coagulation actors neee in obstetrics— especially brinogen—an its plasma treats hypovolemia. A collateral erivative is that women with severe hemorrhage are resuscitate with ewer bloo onor exposures than with packe re cells an components.

Experience at Parklan Hospital supports the preerable use o whole bloo or massive hemorrhage (Alexaner, 2009; Hernanez, 2012). O more than 66,000 eliveries, women with obstetrical hemorrhage treate with whole bloo ha signicantly lower inciences o severe maternal morbiity that inclue renal ailure, acute respiratory istress synrome, pulmonary eema, hypobrinogenemia, intensive care unit amissions, or maternal eath compare with those given packe re cells an component therapy. One unit o packe erythrocytes is erive rom one unit o whole bloo to have a hematocrit o 55 to 80 volume percent. One unit will increase the hematocrit by 3 to 4 volume percent epening on the size o the woman.

Dilutional Coagulopathy

A major rawback o massive hemorrhage treatment with crystalloi solutions an packe re bloo cells is epletion o platelets an clotting actors. Discusse later, this can lea to a ilutional coagulopathy that is clinically inistinguishable rom DIC (Cunningham, 2015; Hossain, 2013). Trombocytopenia is the most requent coagulation eect oun with bloo loss an multiple transusions. In aition, packe re cells have only very small amounts o soluble clotting actors, an store whole bloo is ecient in platelets an in actors V, VIII, an XI. Tus, hypofbrinogenemia an prolongation o the prothrombin (P) an partial thromboplastin times (P) are other sequelae. Because many causes o obstetrical hemorrhage also cause DIC, the istinction between ilutional an consumptive coagulopathy can be conusing. Fortunately, treatment or both is similar.

Platelets

Platelet transusions are consiere with ongoing obstetrical hemorrhage i the platelet count alls below 50,000/µL (Kenny, 2015). In the nonsurgical patient, bleeing is rarely encountere i the platelet count is 10,000/µL or higher (Murphy, 2010). Te preerable source o platelets is one “bag” obtaine by single-onor apheresis. Tis contains the equivalent o one unit each rom six iniviual onors. Depening on maternal size, each single-onor-apheresis, six-unit bag raises the platelet count by approximately 20,000/µL. I these bags are not available, then iniviual-onor platelet units are use, an six to eight such units are generally transuse.

Importantly, the onor plasma in platelet units must be compatible with recipient erythrocytes. Further, because some re bloo cells are invariably transuse along with the platelets, only units rom D-negative onors shoul be given to D-negative recipients.

Fresh-frozen Plasma

Tis component is prepare by separating plasma rom whole bloo an then reezing it. Approximately 30 minutes are require or rozen plasma to thaw. It is a source o all stable an labile clotting actors, incluing brinogen. Tus, it is oten use or treatment o women with consumptive or ilutional coagulopathy. Plasma is not a suitable volume expander in the absence o specifc clotting actor defciencies. It is consiere in a bleeing woman with a brinogen level <150 mg/L or with an abnormal P or P.

An alternative to rozen plasma is liquid plasma (LQP). Tis never-rozen plasma is store at 1 to 60 C or up to 40 ays, an its use compares avorably with resh-rozen plasma (Backholer, 2017). Liqui plasma is not universally available in many centers.

Cryoprecipitate and Fibrinogen Concentrate

Each unit o cryoprecipitate is prepare rom one unit o resh-rozen plasma. Each 10- to 15-mL unit contains at least 200 mg o brinogen along with actor VIII:C, actor VIII:von Willebran actor, actor XIII, an bronectin (American Association o Bloo Banks, 2017). It is usually given as a “pool” or “bag” using an aliquot o brinogen concentrate taken rom 8 to 120 onors. Cryoprecipitate is an ieal source o brinogen when levels are angerously low an surgical incisions show oozing. Because o transit time, the return o cryoprecipitate to transusion services is oten not easible an is consiere wastage i not eploye. Nearly 17 percent o orere cryoprecipitate has been reporte to be waste in labor an elivery units, an this is the most commonly waste prouct (Yee, 2019).

Another alternative is virus-inactivate brinogen concentrate (Ng, 2020). Tese poole brinogen proucts are markete as RiaSAP an Fibryna. Each gram o concentrate raises the plasma brinogen level approximately 40 mg/L. Tese concentrates are use to treat congenital hypobrinogenemia synromes.

Recombinant Activated Factor VII

Tis synthetic vitamin K–epenent protein is available as NovoSeven. Recombinant activate actor VII (rFVIIa) bins to expose tissue actor at the injury site to generate thrombin that activates platelets an the coagulation cascae. Since its introuction, rFVIIa has been use to help control hemorrhage rom surgery, trauma, an obstetrical causes (Goonough, 2016; Murakami, 2015). Many Level I trauma centers inclue it in MPs. Importantly, rFVIIa will not be eective i the plasma brinogen level is <50 mg/L or the platelet count is <30,000/µL.

One concern with rFVIIa is associate arterial—an to a lesser egree—venous thrombosis. In a review o 35 ranomize trials, arterial thromboembolism evelope in 55 percent (Levi, 2010). A secon concern is that it has been oun to have only marginal efcacy (Pacheco, 2019).

Tranexamic Acid

Normally, plasminogen bins with tissue plasminogen activator (tPA) to orm plasmin. Tis bining egraes brin into brinogen–brin egraation proucts an leas to clot lysis. Instea, tranexamic aci (XA) reversibly bins to plasminogen, an thereby blocks plasmin bining to brin. Fibrin strans are not broken, an a clot persists to slow own bleeing.

Ahmazia an coworkers (2021) recently reporte the pharmacoynamics o XA at cesarean elivery using rotational thromboelastometry. XA inhibite tPA-inuce clot lysis in a ose-epenent manner. In the ranomize WOMAN trial o gravias with hemorrhage ollowing vaginal birth or uring cesarean elivery, mortality rates rom obstetrical hemorrhage were 1.2 percent in those given a 1-g intravenous XA ose plus traitional care or bleeing. Tis rate was statistically lower than the 1.7-percent eath rate in women given traitional care alone (WOMAN rial Collaborators, 2017). Te trial was carrie out in eveloping nations an because o the results, the rug is recommene by the Worl Health Organization (2017). In a large ranomize trial o tranexamic aci given or prophylaxis, the rug i not ecrease the rate o postpartum hemorrhage ollowing vaginal elivery (Sentilhes, 2018). In the ranexamic Aci or Preventing Postpartum Hemorrhage Following a Cesarean Delivery—RAAP2 trial—women unergoing cesarean elivery who were given tranexamic aci prophylaxis ha a lower incience o re-cell transusions (Sentilhes, 2021).

Gayet-Ageron an coworkers (2018) conucte a metaanalysis with more than 1000 patients with postpartum hemorrhage treate with XA. Tey also analyze ata rom two ranomize trials that inclue more than 40,000 women treate with XA. Tey conclue that treatment must be given soon ater bleeing onset to reuce maternal mortality rates. Shakur an colleagues (2018) perorme a Cochrane atabase review an also conclue that early XA aministration ecreases maternal eaths.

Most o these stuies were conucte in impoverishe countries, an the American College o Obstetricians an Gynecologists (2019a) oes not recommen XA or either prophylaxis or rst-line treatment o obstetrical hemorrhage. I use, aministration within 3 hours o elivery is recommene.

Massive Transfusion Protocols

As note, these protocols are initiate once our to ve units o packe re cells have been given within 2 hours or so. Tese protocols spee bloo prouct elivery to provie early resuscitation an help avoi ilutional coagulopathy. Once activate, packe re cells, plasma, platelets, an brinogen are given set ratios (Table 44-2). Some protocols inclue rFVIIa, XA, or prothrombin complex concentrates (Jackson, 2018).

Te ata supporting the superiority o MPs compare with traitional component replacement to improve survival rates in obstetrical stuies are limite. Most reports escribe nonpregnant trauma victims, but some observational stuies aress obstetrical hemorrhage (Green, 2016; Pacheco, 2016).

Viscoelastic Assays

Tromboelastography (EG) an rotational thromboelastometry (ROEM) are point-o-care tests that assess coagulation in whole bloo uring massive transusions. Tese tests work by analyzing both clot ormation an breakown in a whole bloo sample rom a given patient. esting prouces a prole o coagulation ynamics, an isplaye values inicate the spee an quality o clot ormation (Fig. 44-1). Tese assays provie inormation regaring time to clot ormation, clot strength, an brinolysis (Arnols, 2020). Currently, they guie bloo prouct replacement in trauma, liver transplant, an cariac surgery patients. Stuies o EG an ROEM techniques in pregnant women have conrme the hypercoagulable state o pregnancy an provie reerence ranges or use in this population (Lee, 2021; McNamara, 2019; Pacheco, 2019).

Although promising, they also have several limitations. For example, they are less inormative in a woman with torrential hemorrhage in whom clotting unction woul nee to be measure minute-by-minute. Tey cannot be use to etect isorers o primary hemostasis (Solomon, 2012). A major rawback is the risk o misinterpretation when tests are use by inaequately traine personnel. Although use is gaining avor, we an others recommen urther stuy beore these tests are wiely applie or obstetrical hemorrhage treatment (Amgalan, 2020). We have oun this testing moality to be more applicable ater initial management o hemorrhage an with intensive care recovery than uring acute hemorrhage.

Cell Salvage and Autologous Transfusion

Preoperative patient phlebotomy an autologous blood storage or transusion in obstetrics has been isappointing. Exceptions are women with a rare bloo type or with unusual antiboies (Pacheco, 2013; Sullivan, 2019). Intraoperative blood salvage with reinusion is consiere a sae intervention in obstetrical patients. Tis practice may ai women eclining transusion (Chap. 30, p. 549). Prior concern centere on amnionic ui contamination an embolism (Dhariwal, 2014; Goucher, 2015). o evaluate benets, one trial ranomly assigne 3028 women at risk or hemorrhage an unergoing cesarean elivery either to routine care or to cell salvage. Te rate o nonautologous onor bloo transusion was not signicantly reuce in the cell salvage group—2.5 versus 3.5 percent (Khan, 2017). In another stuy, Sullivan an colleagues (2019) routinely set up most cesarean eliveries or autologous transusions an reporte a lower rate o nonautologous onor transusion. Similar to prior reports, no cases o amnionic ui embolism were reporte in these two stuies.

■ Transfusion Complications

O serious known risks, transusion o an incompatible blood component may result in acute hemolysis. I severe, this can cause DIC, acute kiney injury, an eath. Preventable errors responsible or most o these reactions requently inclue mislabeling o a specimen or incorrectly transusing a patient not slate or those proucts.

Te rate o such errors in the Unite States is estimate to be 1 case in 14,000 units, but these events are likely unerreporte (Lerner, 2010). A transusion reaction is characterize by ever, hypotension, tachycaria, yspnea, chest or back pain, ushing, severe anxiety, an hemoglobinuria. Immeiate supportive measures inclue stopping the transusion, treating hypotension an hyperkalemia, provoking iuresis, an alkalinizing the urine. ransusion-related acute lung injury (RALI) an transusion-associated circulatory overload (ACO) are the most common causes o transusion-relate mortality (Semple, 2019).

Aecte women characteristically have severe yspnea, hypoxia, an pulmonary eema that evelop within 6 hours o transusion (Peters, 2015). O the two, ACO is more common, an its incience nears 1 percent. RALI is estimate to complicate at least 1 in 12,000 transusions (Carson, 2017). Although their pathogenesis is incompletely unerstoo, injury to the pulmonary capillaries may arise rom human leukocyte antigen (HLA) antiboies an human neutrophil antiboies (HNA) in onor plasma (McCullough, 2016). Management is supportive an may inclue mechanical ventilation (Chap. 50, p. 885)

Bacterial inection rom transusion o a contaminate bloo component is unusual because organism growth is iscourage by rerigeration. Te most oten implicate contaminants o re cells inclue Yersinia, Pseudomonas, Serratia, Acinetobacter, an Escherichia species. Te more important risk is rom bacterial contamination o platelets, which are store at room temperature. Current estimates are that 1 in 1000 to 2000 platelet units are contaminate. Death rom transusion-relate sepsis is 1 case in 17,000 transuse single-onor platelets an 1 case in 61,000 transuse apheresis-onor packs (Lerner, 2010). Viral inection risks rom transusion have been curtaile. Te estimate risk o human immunoeciency virus (HIV) or o hepatitis C virus inection in screene bloo is 1 case in 1 to 2 million transuse units (Carson, 2017). Te estimate hepatitis B transmission rate is <1 case in 100,000 transuse units (Jackson, 2003). Because o the high prevalence o the virus, cytomegalovirus-inecte leukocytes are oten transuse.

Risks or transmitting West Nile virus, parvovirus B19, human -lymphotropic virus type 1, an toxoplasmosis are slight (American Association o Bloo Banks, 2013; ForoutanRa, 2016). Although rare, Zika virus has emerge as another relevant transusion-transmitte inection (Motta, 2016). Collection o all whole bloo components now inclues testing or Zika virus (Centers or Disease Control an Prevention, 2018). At this time, transmission o SARS-CoV-2 virus rom a COVID-positive onor is hypothetical (Leblanc, 2020).

OBSTETRICAL COAGULOPATHIES

Te terms consumptive coagulopathy, defbrination syndrome, an disseminated intravascular coagulation (DIC) are oten use interchangeably, but istinctions are important. Actual consumption o procoagulants within the intravascular tree results in a consumptive coagulopathy. In contrast, massive loss o procoagulants rom hemorrhage results in a ilutional coagulopathy. Semantics asie, consumptive coagulopathy culminates in a systemic intravascular activation that completely isrupts natural hemostasis. As a result, an ineective balance o natural anticoagulant mechanisms leas to wiesprea brin eposition that can cause multiorgan ailure (Levi, 2016).

■ Pregnancyinduced Coagulation Changes

During normal pregnancy, the balance between coagulation an brinolysis changes to create a procoagulant state. Changes that promote coagulation inclue appreciable elevation in the plasma concentrations o actors I (brinogen), VII, VII, IX, an X. A partial list o these normal values is oun in the Appenix (p. 1228). Concurrently, levels o plasminogen, which lyses brin, rise consierably. However, plasminogen activator inhibitor 1 an 2 (PAI-1 an PAI-2) levels also increase. Tus, plasmin activity usually eclines until ater elivery (Hale, 2012; Hui, 2012). As shown in Figure 4-7, (p. 62), mean platelet count rops by 10 percent uring pregnancy, an platelet activation is enhance (Kenny, 2015; Reese, 2018).

Te net results o these changes inclue greater levels o brinopeptie A, β-thromboglobulin, platelet actor 4, an brinogen–brin egraation proucts, which inclues d-imers. In aition to lower concentrations o the anticoagulant protein S, pregnancy-relate hypercoagulability, an ecrease brinolysis, there is augmente—yet compensate—intravascular coagulation that may unction to maintain the uteroplacental interace.

■ Disseminated Intravascular Coagulation

Because o the many enitions an variable severity, the incience o consumptive coagulopathy in gravias varies an ranges rom 0.03 to 0.35 percent (Erez, 2014; Rattray, 2012). For example, some egree o signicant coagulopathy is oun in virtually all cases o placental abruption an amnionic ui embolism. Other instances in which requently occurring but less recognize egrees o coagulation activation can be oun inclue sepsis; thrombotic microangiopathies; acute kiney injury; acute atty liver; severe preeclampsia; an hemolysis, elevate liver enzyme levels, low platelet count (HELLP) syn- rome (Cunningham, 2015).

When consumptive coagulopathy is severe, the likelihoo o maternal an perinatal morbiity an mortality rises. In one stuy o 49 cases, anteceent causes inclue those liste above, an 59 percent receive bloo transusions, 18 percent unerwent hysterectomy, 6 percent require renal ialysis, an three mothers ie (Rattray, 2012). In this series, the perinatal mortality rate was 30 percent. From 2010 to 2011, DIC was the secon most common severe maternal morbiity inicator (Creanga, 2014). DIC was associate with nearly a ourth o maternal eaths uring this stuy perio. Despite these statistics, consumptive coagulopathy as the sole cause o maternal eath is uncommon an accounts or only 0.2 percent o pregnancy-relate eaths in the Unite States (Creanga, 2015).

■ Activation of Normal Coagulation

Instea o the “waterall” sequential activation o the clotting cascae, a current theory proposes that tissue actor—an integral membrane glycoprotein—serves as the principal initiator o coagulation (Levi, 2016). Coagulation then moves orwar but incorporates a eeback loop. o begin, tissue actor orms complexes with actor VII/VIIa to activate actors IX an X. issue actor is oun in highly vascularize organs such as the brain, lungs, an placenta; in amnionic ui; an in certain other cells (Kuczyski, 2002; Østeru, 2006; Uszyski, 2001). issue actor–actor VIIa complexes ultimately generate activate actor X (Xa) to initiate clotting. Subsequently, the previously labele “intrinsic” pathway amplies this process. Specically, the initial thrombin prouce irectly activates actor XI by proviing a eeback amplication loop (Fig. 44-2). Te en result o this amplie coagulation process is brin ormation. Tis is then counterbalance by the brinolytic system, in which plasminogen is activate. Even this process is tie initially to tissue actor. Te nal prouct is brin egraation proucts, sometimes calle brin split proucts, which inclue d-imers.

Activation of Pathological Coagulation

With DIC, pathological entities prompt tissue actor release rom subenothelial tissue an stimulate monocytes, which in turn provoke cytokine release rom the enothelium. With generalize enothelial activation, iuse activation o coagulation ollows. Tis pathological cycle o coagulation an brinolysis becomes clinically important when coagulation actors an platelets are sufciently eplete to create consumptive coagulopathy.

Purpura ulminans is a severe—oten lethal—orm o consumptive coagulopathy. It is cause by microthrombi in small bloo vessels leaing to skin necrosis an sometimes vasculitis. Debriement o large areas o skin over the extremities an buttocks requently requires treatment in a burn unit. Rarely, our-extremity amputation is require (Bhatti, 2019). Purpura ulminans usually complicates sepsis in women with heterozygous protein C eciencies an low protein C serum levels (Bhatti, 2019; Levi, 2016). Importantly, homozygous protein C or S eciency results in neonatal purpura ulminans, which is atal (Chap. 55, p. 976).

■ Inciting Conditions

Several obstetrical synromes can trigger consumptive coagulopathy. O these, placental abruption is the most common cause o severe consumptive coagulopathy in obstetrics an is iscusse more ully in Chapter 43 (p. 749). With preeclampsia, eclampsia, an HELLP syndrome, enothelial activation is a hallmark. Perhaps 10 percent o HELLP cases have some isseminate coagulation (Haram, 2017). In general, the clinical severity o preeclampsia is irectly correlate with thrombocytopenia an brin egraation proucts (Geik 2017; Kenny, 2015). Tat sai, intravascular coagulation is selom severe enough to be clinically worrisome (Pritchar, 1976). Acute atty liver o pregnancy also causes intravascular coagulation in aition to ecrease procoagulant synthesis (Chap. 58, p. 1034).

Amnionic uid embolism is usually cause by intravenous embolization o meconium-laen amnionic ui. It results in rapi cariorespiratory collapse an prooun consumptive coagulopathy. Te Society or Maternal-Fetal Meicine (2021) has evelope a checklist or the initial management o amnionic ui embolism. It is iscusse in Chapter 42 (p. 745). Sepsis stemming rom various inections can be accompanie by eno-or exotoxin release. Although a eature o sepsis syn- rome inclues activation o coagulation, selom oes sepsis alone cause massive procoagulant consumption. Escherichia coli bacteremia is requently seen with antepartum pyelonephritis an puerperal inections, however, accompanying consumptive coagulopathy is usually not severe. Some notable exceptions are septicemia cause by exotoxins release rom group A Streptococcus pyogenes, Staphylococcus aureus, or Clostridium perringens, C sordellii, or C novyi (Herrera, 2016; Pritchar, 1971). reatment o sepsis an septic shock is iscusse in Chapter 50 (p. 889). Septic abortion—especially associate with these organisms—can incite coagulation an worsen hemorrhage.

Second-trimester induced abortions can stimulate intravascular coagulation even in the absence o sepsis. Ben-Ami an associates (2012) escribe a 1.6-percent incience in 1249 late-secon- trimester pregnancies terminate by ilation an evacuation. wo thirs were one or etal emise, which may have contribute to coagulopathy (Kerns, 2019). Another source o intense coagulation is instillation o hypertonic solutions to eect mitrimester abortions. Tese are not commonly use currently or pregnancy terminations. Te mechanism is thought to initiate coagulation by thromboplastin release into maternal circulation rom the placenta, etus, an eciua by the necrobiotic eect o hypertonic solutions.

Prolonged retention o a dead etus is an unusual cause o consumptive coagulopathy toay. Namely, etal eath can be easily conrme, an highly eective methos or labor inuction are available. Moreover, i the ea etus is unelivere, most women enter spontaneous labor within 2 weeks an gross isruption o maternal coagulation rarely evelops beore 4 weeks (Pritchar, 1959, 1973). Ater 1 month, however, almost a ourth will evelop consumptive coagulopathy.

■ Diagnosis

Bioassay is an excellent metho to etect or suspect clinically signicant coagulopathy. Excessive bleeing at sites o mo- est trauma characterizes eective hemostasis. Examples inclue persistent bleeing rom venipuncture sites, nicks rom preoperative shaving, trauma rom blaer catheterization, an spontaneous bleeing rom the gums, nose, or gastrointestinal tract. Purpura or petechiae at pressure sites such as sphygmomanometer cus or tourniquets suggest signicant thrombocytopenia. Any surgical proceure provies the ultimate bioassay an elicits generalize oozing rom abominal wall layers, the retroperitoneal space, the episiotomy, or incisions an issections or cesarean elivery or hysterectomy.

O laboratory tests, brinogen an d-imer levels can be inormative. In late pregnancy, plasma brinogen levels typically have risen to 300 to 600 mg/L. Tus, even with severe consumptive coagulopathy, levels may sometimes be sufciently high to protect against clinically harmul hypo- brinogenemia. For example, ebrination cause by a placental abruption might lower an initial brinogen level o 600 mg/L to 250 mg/L. Although this woul inicate massive brinogen consumption, levels are still aequate to promote clinical coagulation. Te clinically impactul threshol usually approximates 150 mg/L. Te eects o serious hypo- brinogenemia—less than 50 mg/L—can be illustrate with clinical thrombin clot test. With low brinogen levels, the clot orme rom whole bloo in a glass phlebotomy tube will initially be sot an cannot retract in volume. Instea, uring the next hal hour, platelet-induced clot retraction ensues, an this causes a tighter clot to orm.

As epicte in Figure 44-2, brinolysis cleaves brin into various brin egraation proucts, an one o the smaller ones is the d-imer. Several sensitive assays contain monoclonal antiboies specic or d-imers (Johnson, 2019). Assay values are almost always abnormally high with clinically signicant consumptive coagulopathy. At least in obstetrical isorers, quantication oes not correlate with outcomes. Examples o the magnitue o brin split prouct elevations in various obstetrical coagulopathies are shown in Figure 44-3.

Trombocytopenia is likely i petechiae are abunant or i clotte bloo ails to retract within an hour or so. Conrmation is provie by a low platelet count. I severe preeclampsia syn- rome is comorbi, qualitative platelet ysunction may coexist (Chap. 40, p. 696).

Prothrombin time (P) an partial thromboplastin time (P) are stanar coagulation tests. Prolongation may stem rom very low brinogen concentrations, rom appreciably reuce levels o the procoagulants neee to generate thrombin, or rom large amounts o circulating brinogen–brin egraation proucts.

Tromboelastometry an thromboelastography are point-o- care tests use as ajuncts to conventional laboratory stu- ies (McNamara, 2019; Pacheco, 2019). Teir current role may serve to guie bloo prouct replacements as iscusse earlier (p. 774). Using many o these tests, several organizations have attempte to establish a more uniorm enition o DIC. One is the International Society on Trombosis an

Haemostasis (ISH) scoring system. Te score is use only ater a conition known to cause intravascular coagulation is ientie an is calculate using a combination o laboratory tests. Composite ISH-DIC scores <5 suggest nonovert DIC, whereas scores ≥5 are compatible with overt DIC.

A scoring system or nonovert DIC has been evelope by Alhousseini an coworkers (2020). Tese scoring systems have not been applie wiely in obstetrics (Hizkiyahu, 2019; Jonar, 2016; Nelson, 2014).

■ General Management

o halt ongoing ebrination, ientiying an removing the inciting source is a priority. With incisions or extensive lacerations accompanie by severe hemorrhage, rapi replacement o procoagulants is usually inicate. Vigorous restoration an maintenance o the circulation to treat hypovolemia cannot be overemphasize. Aequate perusion restores hepatic an enothelial synthesis o procoagulants an permits prompt removal o activate coagulation actors, brin, an brin egraation proucts by the reticuloenothelial system.

Asie rom these unamental steps, ew other agents have prove sounly eective. Although its use is seemingly counterintuitive, unractionate heparin has now been abanone. Other agents not recommene or rst-line use inclue tranexamic aci or epsilon-aminocaproic aci, both anti- brinolytic agents (American College o Obstetricians an Gynecologists, 2019a; Pacheco, 2017). Tis is because the brinolytic system is necessary or issolution o wiesprea brin thromboses cause by generalize intravascular coagulation (Levi, 2016). Discusse earlier (p. 773), recombinant actor VIIa (rFVIIa) can be use to help control severe obstetrical hemorrhage rom other causes. However, current clinical evience is insufcient to make rm recommenations on its aministration or obstetrical coagulopathies.

SURGICAL MANAGEMENT OF HEMORRHAGE

Several invasive proceures can help to arrest severe postpartum hemorrhage. A report rom the Agency or Healthcare Research an Quality conclue that most stuies aressing these methos are o poor quality (Likis, 2015). In one stuy o 6660 women with postpartum hemorrhage, 4 percent unerwent an invasive proceure, an 1 percent ha a hysterectomy (Kayem, 2016). Te ailure rate o conservative surgical an embolization proceures was 15 percent. O specic methos, use o uterine balloon tamponae or treatment o atony has ouble in the past ecae (Merriam, 2020; Suarez, 2020). It is escribe in Chapter 42 (p. 737). For low-pressure bleeing, topical hemostatic agents can be use to control persistent surgical oozing. Tese were reviewe by Stachowicz an colleagues (2020). Other than or cesarean hysterectomy, these are selom use in obstetrical hemorrhage.

■ Uterine Compression Sutures

Tis surgical technique uses a no. 2 chromic suture to compress the anterior an posterior uterine walls together (B-Lynch, 1997). Because they give the appearance o suspeners, they are also calle braces (Fig. 44-4). Several moications o the B-Lynch technique have been escribe (Gilmanyar, 2019; Matsubara, 2013; Nelson, 2007).

B-Lynch sutures are mainly use to treat atony. Tese may help iminish DIC-associate bleeing, while coagulation is correcte by transusion. Success rates vary by inication. For example, B-Lynch (2005) cite 948 cases with only 7 ailures. Conversely, in other series, overall ailure rates were 20 to 25 percent (Kaya, 2016; Kayem, 2011). Some unique complications can rarely ollow compression sutures (Matsubara, 2013). Most involve variations o uterine ischemic necrosis with peritonitis (Gottlieb, 2008; Joshi, 2004; reloar, 2006). In most cases, subsequent pregnancies are uneventul i compression sutures are use (An, 2013). A ew women with B-Lynch sutures, however, evelope uterine wall eects or uterine cavity synechiae (Akoury, 2008; Alouini, 2011; Ibrahim, 2013).

■ Artery Ligation

Ligation o arterial bloo ow to the uterus can halt or iminish obstetrical hemorrhage. Common inications inclue uterine hysterotomy incision laceration, uterine atony, an abnormal placentation. Uterine artery ligation may be unilateral or bilateral an is use mainly or lacerations at the lateral part o a hysterotomy incision (Fig. 44-5). In our experiences, this proceure is less helpul or other hemorrhage etiologies. In rare cases with torrential hemorrhage rom placenta accreta synrome, a tourniquet wrappe aroun the lower uterine segment may slow bleeing sufciently while resuscitation ensues an issection or hysterectomy is perorme.

Internal iliac artery ligation may also be unilateral or bilateral. For years, ligation has been use to reuce pelvic hemorrhage. However, the proceure may be technically ifcult an is successul in only hal o cases (American College o Obstetricians an Gynecologists, 2019a; Bienstock, 2021). Wei an colleagues (2019) reporte it to be successul or improve hemostasis with placenta previa. It is not particularly helpul or abating hemorrhage with postpartum atony (Clark, 1985).

For internal iliac ligation, aequate exposure is obtaine by opening the peritoneum over the common iliac artery an issecting own to the biurcation o the external an internal iliac arteries (Fig. 44-6). Branches istal to the external iliac arteries are palpate to veriy pulsations at or below the inguinal area. Ligation o the internal iliac artery at a point 5 cm istal to the common iliac biurcation will usually avoi the internal iliac artery's posterior ivision branches (Bleich, 2007). Te areolar sheath o the artery is incise longituinally, an a right-angle clamp is careully passe just beneath the artery rom lateral to meial. Care must be taken not to perorate contiguous large veins, especially the internal iliac vein. Suture—usually nonabsorbable—is passe uner the artery with a clamp, an the vessel is then securely ligate.

Te most important mechanism o action with internal iliac artery ligation is an 85- percent reuction in pulse pressure in those arteries istal to the ligation (Burchell, 1968). Tis converts an arterial pressure system into one with pressures approaching those in the venous circulation. Tis creates vessels more amenable to hemostasis via pressure an clot ormation. Even bilateral internal iliac artery ligation oes not appear to interere with subsequent repro- uction. Nizar an colleagues (2003) reporte subsequent pregnancy in 17 such women. From a total o 21 pregnancies, 13 were normal, 3 ene with miscarriage, 3 were terminate, an 2 were ectopic.

■ Angiographic Embolization

Tis moality is now use or many causes o intractable hemorrhage when surgical access is ifcult. Its use has increase remarkably in the past ecae (Merriam, 2020). In more than 500 pregnancy-relate cases, embolization was eective in 90 percent (Grönvall, 2014; Lee, 2012; Poujae, 2011; Zhang, 2015). Ater his review, Rouse (2013) conclue that embolization can be use to arrest reractory postpartum hemorrhage. Other reports have been less enthusiastic. Fertility is not impaire, an many subsequent successul pregnancies have been reporte (Fiori, 2009; Kolomeyevskaya, 2009). An important caveat or these procedures is that women with hemodynamic instability related to active bleeding should not be removed rom the operating room.

Complications are relatively uncommon but can be severe. Case reports etail instances o iatrogenic iliac artery rupture, uterine ischemic necrosis, an uterine inection (Grönvall, 2014; Katakam, 2009; Nakash, 2012). A rare case o massive buttock necrosis an paraplegia ollowe bilateral internal iliac artery embolization (Al-Tunyun, 2012). A myometrial eect in subsequent pregnancy also has been reporte (Choo, 2019).

In a ew instances, massive bloo loss an ifcult surgical issection is anticipate. Te use o balloon-tippe catheters preoperatively inserte into the iliac or uterine arteries is escribe in management o placenta accreta spectrum (Chap. 43, p. 759).

■ Aortic Compression

Prophylactic use o an aortic balloon occlusion evice has been reporte to reuce bloo loss an the nee or hysterectomy (Chen, 2019). Also known as resuscitative endovascular balloon occlusion o the aorta (REBOA), such occlusion lowers the perusion pressure istally, increases cariac aterloa, an reistributes bloo volume to the heart an brain (Cannon, 2018). Prolonge occlusion can result in lower limb ischemia requiring amputation. Whittington an colleagues (2020) escribe its successul prophylactic use in 11 women with placenta accreta spectrum isorers. Instea, manual compression o the aorta above the sacral promontory can be use in cases in which pelvic hemorrhage is torrential. I the abomen is close, pressure place above the umbilicus reuces bloo pressure istally (Barbieri, 2018).

■ Pelvic Packing

I hysterectomy ails to curtail hemorrhage, then pelvic packing with gauze an termination o the operation may be consiere (Pacheo, 2018; ouhami, 2018). Rolls o gauze are packe to provie constant local pressure. In some cases, this may serve as a temporizing step prior to interventional embolization. In other cases, packing alone may be let or 48 to 72 hours. I the patient is stable an bleeing appears to have stoppe, packing is remove. In one review o 104 cases, the success rate was 79 percent. Packing aile in 24 patients, an 13 o these women ie, yieling an overall mortality rate o 13 percent (ouhami, 2018).

Te umbrella or parachute pack uses a similar concept (Logothetopulos, 1926). Although selom use toay, it can be liesaving i all other measures have aile, especially in low-resource areas (Dily, 2006; ouhami, 2018). Te pack is constructe o a stury sterile plastic bag that is lle with gauze rolls that are unwoun an knotte together. Sufcient rolls are use to provie aequate volume in the bag to ll the pelvis. Te pack is introuce transabominally with the stalk exiting the vagina. Mil traction is applie by tying the stalk to a 1-liter ui bag, which is hung over the oot o the be. Te umbrella pack is remove vaginally ater 24 hours.

Nhận xét