52: Obstetric emergencies

Obstetric emergencies

Karin Fox1,2, Alexandria J. Hill3,4,5, and Stephanie R. Martin6

1Maternal‐Fetal Medicine, Baylor College of Medicine, Houston, TX, USA

2Texas Children’s Hospital, Pavilion for Women, Houston, TX, USA

3High Risk Pregnancy Center, Las Vegas, NV, USA

4Texas A&M College of Medicine, College Station, TX, USA

5University of Arizona, Phoenix, AZ, USA

6Clinical Concepts in Obstetrics, Scottsdale, AZ, USA


This chapter focuses on the role of high‐performing team‐based care and the use of standardized protocols, particularly in the setting of obstetrical emergencies. There are two clinical scenarios used to illustrate the utility of multidisciplinary team care, one involving the evaluation and management of pulmonary embolism, and the second involving cardiac arrest during pregnancy.


In the last 25 years, there has been a dramatic paradigm shift in medical care. Previously, the physician’s role was that as a staunchly independent, all‐knowing informational source and leader, with availability at all hours day or night. The pace at which medical knowledge and complexity has advanced makes relying upon one’s own resources not only impractical, but also potentially unsafe. According to a discussion paper published by the Institute of Medicine in 2012, entitled, “Core Principles and Values of Effective Team‐Based Health Care,” the US National Clearinghouse lists over 2700 clinical practice guidelines and more than 25 000 new clinical trials are published annually [1]. No one individual could possibly effectively read, process and apply such a vast amount of information, let alone care for patients with ever more complex chronic and acute disease.

Much like in other industries, such as the commercial aviation or nuclear power industries, in which groups of people must work together in hazardous conditions while maintaining a high level of safety, there has been a push toward development of highly efficient, highly reliable, team‐based systems within healthcare. The impetus for this change is in part due to increasing concomitant public demands to improve patient safety and control rising healthcare costs. In 2000, the Institute of Medicine published the report “To Err Is Human,” [2] which highlighted the devastating effects of medical error both on patient mortality and national costs, with an estimated 44 000–98 000 deaths per year due to medical errors and estimated adverse health care costs between $17 and $29 billion annually. Since then, national efforts have been led by various medical leadership, education, and credentialing organizations including the Joint Commission [3], the American College of Obstetrics and Gynecology (ACOG) [4], and the American Council for Graduate Medical Education (ACGME) [5], to develop organizational systems that promote a culture in which physicians function as “leaders and participants in team‐oriented care.” [5]. Standardization in practice may also be leveraged to improve communication and outcomes [4]. The goal of this chapter is to explore the role of high‐reliability, team‐oriented care specifically in the setting of obstetrical emergencies.


Venous thromboembolism (VTE) remains one of the leading causes of death in industrialized countries [68]. Monitoring data from the United Kingdom (UK) indicates that a significant decrease in death due to pulmonary embolism has contributed to the slight decrease in overall maternal deaths between 2006 and 2008. Pulmonary embolism dropped from its former spot as the leading cause of death in the UK for the first time since 1985 [8]. This follows the Royal College’s emphasis on prompt recognition and treatment of acute VTEs and recommendations for thromboprophylaxis, initially published in 2001 and updated in 2015 [9, 10]. Similar efforts have been made in the United States, including recommendations from the American College of Obstetricians and Gynecologists [11], the American College of Chest Physicians [12], and the American Thoracic Society and American Society of Radiologists [13, 14].

The aims of this section are to review available evidence regarding diagnosis of pulmonary embolism in pregnancy, to discuss recommendations from the aforementioned medical societies and expert consensus where data is sparse, and to elucidate strategies that may be used to aid healthcare teams in early identification of emergencies such as a VTE. Treatment of venous thrombus embolism is covered in detail in Chapter 34.

Clinical questions

  1. Do the normal physiologic and hematologic changes (tests) in the pregnant patient (population) alter the evaluation in working up a patient for pulmonary embolism (outcome)?
  2. Are the Wells’ criteria of assessment of pretest probability of pulmonary embolism (assessment), useful in pregnancy (population)? Is there another formal assessment that may more accurately determine pretest probability in pregnancy (comparison/outcome)?
  3. In pregnant patients admitted to the hospital (population), are the number of deaths from venous thrombotic events reduced (outcome) when using a Modified Early Obstetrical Warning System (MEOWS) (comparison)?

General search strategy

COCHRANE: pulmonary embolism AND pregnancy (yielded 1 Cochrane review, 1 other review, 9 clinical trials, 2 economic evaluations) and PUBMED: pulmonary embolism and pregnancy.

Critical appraisal of the literature

  1. Do the normal physiologic and hematologic changes (tests) in the pregnant patient (population) alter the evaluation in working up a patient for pulmonary embolism (outcome)?

Search Strategy

  • PUBMED: “pregnancy” AND “pulmonary embolism” AND “test” AND (case–control OR cohort OR meta‐analysis).

Many of the symptoms of pulmonary embolism such as chest pain and shortness of breath are non‐specific, and may lead one to a wide differential diagnosis (Table 52.1). The considerable prevalence of these symptoms during normal pregnancy can either mimic or mask an embolic event (Table 52.2).

Table 52.1 Brief list of differential diagnoses based on symptoms of pulmonary embolism

Symptom/signs of VTE Differential diagnosis
Shortness of breath Normal pregnancy changes
Tachypnea Acute asthma exacerbation
Decreased oxygenation Pulmonary edema (+/− pre‐eclampsia)
Cough +/− hemoptysis Pneumonia
Congestive heart failure/myocarditis
Mildly elevated temperature/fever Systemic infection/sepsis
Tachycardia Cardiac tachyarrhythmia
Thyrotoxicosis/thyroid storm
Drug toxicity
Cocaine use
Chest pain (acute) Myocardial infarction
Aortic or coronary artery dissection
Costochondritis/musculoskeletal pain
Calf pain/edema Normal pregnancy changes
Muscle spasm
Pre‐eclampsia (although no longer part of diagnostic criteria)
Congestive heart failure/myocarditis
Deep venous thrombosis

Table 52.2 Comparison of symptom overlap between physiologic changes in pregnancy and VTE

Physiologic change of normal pregnancy [15] Symptom of DVT/PE
Shortness of breath (15–75% of patients in 1st and 3rd trimester respectively) Shortness of breath
Minute ventilation increases 15% Tachypnea
Physiologic respiratory alkalosis (pH 7.44) Respiratory alkalosis or acidosis
Functional residual capacity decreases ∼20% Sudden oxygen desideration
Heart rate increases 15–20% (normal range in pregnancy 61–81 bpm) Tachycardia (mild to >150 bpm)
Lower extremity edema (often bilateral) Edema of affected limb
Elevation of D‐Dimer of 32–39% [16] Elevated D‐Dimer

Stasis within pelvic vessels increases as the uterus enlarges [11]. The surge in estrogen levels during normal pregnancy increase the levels of prothrombotic coagulation factors to fourfold in normal parturients [17], and the risk of VTE is increased in women with high‐risk gene carrier status, such as Factor V Leiden or antithrombin‐III deficiency, previous VTE, or a family history of VTE [12]. Other complications of pregnancy such as infection, premature rupture of membranes or preeclampsia, which can lead to hospitalization and bed rest, may both compound these underlying risks or may lead to signs and symptoms that overlap those associated with VTE such as tachycardia, fever, shortness of breath, and decreased oxygenation [15]. Up to 30% of women have no signs of deep venous thrombosis (DVT) prior development of pulmonary embolus [18], therefore a very high index of suspicion is essential to prompt diagnosis and treatment in a pregnant patient. Diagnostic workup to exclude pulmonary embolism is warranted in any pregnant patient with shortness of breath, and immediate initiation of therapeutic anticoagulation with unfractionated or low‐molecular weight heparin (LMWH) is recommended if the pretest probability for pulmonary embolism (PE) is high, until testing can be completed [912, 18].

Proposed diagnostic algorithms by expert panels all involve a step‐wise assessment involving clinical screening and assessment of risk for VTE using a combination of symptoms and bedside testing such as chest x‐ray, electrocardiogram (EKG), pulse oximetry, arterial blood gas evaluation, and initiation of anti‐coagulation therapy if clinical suspicion is high. Diagnostic testing follows, however the order varies by specific recommendations and may vary by availability of local resources. The optimal method of imaging during pregnancy remains somewhat contested, especially considering the need to minimize risks of radiation exposure and invasive or repetitive procedures while avoiding a missed diagnosis. These tests include duplex compression sonography of the lower extremities to evaluate for the presence DVT, ventilation‐perfusion (V/Q) scanning, computed tomography pulmonary angiogram (CTPA) or V/Q single photon emission computed tomography (SPECT) (Tables 52.2 and 52.3) [11, 12, 14]. Each imaging modality has unique advantages and limitations, and each may require additional testing to confirm a VTE (Table 52.4). For example, the physiologic increase in intravascular volume and cardiac output of 30–50% in pregnancy necessitates alterations to protocols during pregnancy regarding the dose and bolus timing of intravenous contrast to obtain optimal diagnostic accuracy during CTPA evaluation for a pulmonary embolism [16, 19, 20]. Without an increase in contrast concentration, increased venous return from the inferior vena cava creates a small area in which a dilutional effect occurs, causing artifact. Without decreasing the delay between contrast administration and imaging, the pulmonary vasculature contrast may not sufficiently be delineated and the number of non‐diagnostic studies increases. This leads to increased exposure of a mother and fetus to either additional doses of radiation, due to repeat or subsequent studies, or to a potential delay in diagnosis and treatment when VTE is present despite low initial clinical suspicion [19, 20]. Regardless of the diagnostic algorithm chosen, more than one test may be required to confirm diagnosis in some pregnant patients, and continued vigilance and treatment are warranted in a patient with an initial high pre‐test probability in whom an initial test is non‐diagnostic.

  1. 2. Are the Wells’ criteria of assessment of pretest probability of pulmonary embolism (assessment), useful in pregnancy (population)? Is there another formal assessment that may more accurately determine pretest probability in pregnancy (comparison/outcome)?

Search Strategy

  • PUBMED: “Wells’” AND “diagnosis of pulmonary embolism in pregnancy”.

Table 52.3 Proposed algorithms for evaluation and diagnosis of VTE in pregnancy

Group/Panel Diagnostic algorithms/recommendations
ACOG [11] 1) If DVT suspected, obtain compression ultrasonography (CUS)

2) If negative, and no pelvic involvement suspected → surveillance

If negative or equivocal and pelvic involvement suspected→further imaging

3) If additional imaging positive, treat, if negative → surveillance

4) If PE suspected, obtain V/Q scan or CTPA. CXR may be used as discriminator

to reduce likelihood of non‐diagnostic V/Q scan.
RCOG [9, 10] 1) In pregnant women with suspicion of VTE, initiate anticoagulant therapy until testing performed

2) Individual hospitals should have an agreed upon protocol for objective diagnosis of VTE during pregnancy

3) If DVT suspected, CUS should be performed. If negative and low suspicion, anticoagulation may be discontinued

4) When PE suspected, perform CXR. If normal, perform CUS. If both are normal and PE is still suspected, CTPA or V/Q scanning should be performed.

5) Alternate or repeat testing should be performed where V/Q scanning or CTPA are negative, but clinical suspicion is still high. Anticoagulation should be continued until PE definitively ruled out.

6) Women with suspected PE should be counseled that V/Q scanning has a slightly higher risk of childhood cancer compared to CTPA (1/280 000 vs. 1/1 000 000) but carries a slightly lower risk of maternal breast cancer (lifetime risk increased up to 13.6% with CTPA), and when feasible, women should be involved in the decision of which test to undergo, and informed consent given.

7) D‐Dimer should not be used in pregnancy
ASR/ATS [14] 1) Do not use D‐Dimer in pregnancy to rule out DVT

2) If DVT symptoms, perform compression ultrasonography (CUS)

– treat if positive, PERFORM additional testing if negative

3) If pregnant and with PE symptoms, but NO symptoms of DVT, perform studies of pulmonary vasculature rather than CUS

4) Use Chest X‐ray (CXR) for initial radiation‐producing imaging

5) Pregnant women with PE symptoms and normal CXR, perform lung scintigraphy rather than CTPA as next step

6) PE suspected and a nondiagnostic V/Q scan, further diagnostic testing suggested rather than clinical management alone

7) PE suspected and abnormal CXR, use CTPA as next imaging modality rather than V/Q scan

Table 52.4 Comparison of imaging modalities used to diagnose VTE in pregnancy

Test Advantages Disadvantages
Duplex compression sonography (Doppler)

  • No radiation exposure
  • Treatment similar for PE/DVT

  • Cannot detect clots isolated to the pelvic vessels
  • May not detect thrombi that have already embolized
Computed tomography pulmonary angiogram (CTPA)

  • Relatively rapid diagnosis
  • Lower radiation exposure if single photon emission CT (SPECT) available
  • Can provide alternative diagnosis (pneumonia, pulmonary edema)

  • Some radiation exposure necessary for test
  • Physiologic changes in pregnancy may result in artifact
Ventilation/perfusion scanning

  • Lower radiation exposure to patient and fetus
  • Reasonably high sensitivity/specificity

  • Diagnostic accuracy limited in patients with active or obstructive lung disease due to low specificity (asthma, chronic obstructive pulmonary disease (COPD), pulmonary edema)
  • Diagnosis less reliable if pre‐test probability moderate or low and results indeterminate.

Several clinical prediction scoring systems have been proposed to estimate pretest probability of a VTE based on patient characteristics, signs, and symptoms (Table 52.5). Scoring systems validated with sufficient numbers of patients include the Modified Geneva Score [20], Wells Criteria [21], and the Pulmonary Embolism Rule‐out Criteria (PERC) or Charlotte score [22]. The Wells criteria include factors dependent upon a clinician’s implicit judgment about whether a diagnosis other than pulmonary embolism (PE) is less likely than PE, and therefore is not based solely on objective criteria [21]. The PERC score, when initially developed, relied solely on objective variables [22]. It was initially validated in a very low‐risk population, and studies in emergency department and internal medicine populations showed that its scoring system alone, or in combination with the revised Geneva score, may not sufficiently exclude patients at risk for pulmonary embolism in high‐risk populations without further testing [23]. Indeed, the most critical value in clinical prediction scoring is in the ability to create an accurate pre‐test probability, and no single scoring system without imaging is sufficient to rule out completely VTE.

Table 52.5 Comparison of clinical scoring systems

Scoring system Finding Points Score/probability
Revised Geneva21 Age > 65 years 1 0–3 Low Probability
4–10 Intermediate
>/= 11 High Probability

Previous DVT/PE 3

Surgery (under general anesthesia) or lower limb fracture within one month 2

Active malignancy (within 1 yr) 2

Unilateral lower limb pain 3

Hemoptysis 2

Heart rate 75–94 beats min−1 3

HR >/= 95 beats min−1 5

Pain on lower limb deep venous palpation and unilateral edema 4
Modified Well’s Criteria22 Suspected DVT 3.0 0–2 Low risk
3–6 Moderate risk
>6 High risk

Alternative diagnosis less likely than PE 3.0

Heart rate > 100 beats/min 1.5

Immobilization/surgery in previous four weeks 1.5

Previous DVT/PE 1.5

Hemoptysis 1.0

Malignancy 1.0
Pulmonary Embolism Rule‐out Criteria (PERC or Charlotte rule)23 Age < 50 All criteria must be negative If negative, likelihood of PE so low that D‐Dimer testing not useful (<1.8% likelihood of PE)

HR < 100

SaO2 > 94%

No unilateral leg swelling

No recent trauma/surgery

No hemoptysis

No hormone use

No Prior PE/DVT

None of the above‐mentioned scoring systems were developed specifically for the pregnant population, nor have they been prospectively validated in pregnancy. The PERC score was derived from logistic regression of 21 independent clinical variables and 3.7% of the 3148 patients from whom clinical variables were analyzed were pregnant [22]. In development of the revised Geneva score, 1% (n = 10) of the derivation population were either pregnant or post‐partum [20]. Interestingly, although pregnant patients were excluded from the derivation of the Wells criteria [24], only the modified Wells criteria has been validated retrospectively in a single‐institution cohort of pregnant patients [25]. In this study, use of the modified Wells score demonstrated a sensitivity of 90% and specificity of 100%, among 125 patients included over a five‐year period. Negative CTPA results were considered equivalent to the absence of PE. By using immediate CT results rather than more long‐term endpoints, such as the lack of diagnosis of PE/DVT or initiation of anticoagulation within a three‐month follow‐up period, patients with false‐negative CT results or who developed a subtle PE within a short time period after imaging might be missed using the scoring system alone. Additionally, 22 patients (18%) were lost to follow‐up. The retrospective nature of this study precludes firm conclusions regarding the safety of use of such a scoring when used as the authors intended – to avoid unnecessary imaging, treatment or hospitalization. Nonetheless, the findings in this study suggest that use of a clinical prediction score can aid in developing a reasonable accurate pre‐test probability prior to imaging the pregnant or postpartum patients.

  1. 3. In pregnant patients admitted to the hospital (population), are the number of venous thrombotic events reduced (outcome) when using a “care bundle” or “Early Obstetrical Warning System” compared to routine care (comparison)?
  2. Search Strategy

    • PUBMED: pregnancy AND bundles.
    • PUBMED: “Early Warning Systems” AND “Pregnancy”.
    • PUBMED: “reduction in thromboembolism in pregnancy”.
    • Cochrane Database: “Thromboembolism AND pregnancy”.
    • Hand‐searching: references listed in the articles obtained.

One approach to preventing morbidity and mortality from thromboembolism includes utilizing protocols for prophylaxis for patients at risk. Interestingly, in the Cochrane review of Prophylaxis for venous thromboembolic disease in pregnancy and the early postnatal period, reviewers found insufficient evidence to guide clinical decision‐making [26]. The reviewers attributed this to a lack of reporting of maternal mortality in any of the studies reviewed. Additionally, a majority of studies were relatively small, with only 2592 women included in the 16 trials that were included in the review. The authors’ conclusion was that absent evidence from randomized controlled trials that can identify the best means for prophylaxis, practitioners must rely on consensus‐derived clinical guidelines, such as those produced or endorsed by the Royal College of Obstetricians and Gynecologists (RCOG) [9, 10], the National Institute for Health and Care Excellence (NICE) [27], the American College of Chest Physicians [12], or other international organizations such as the ACOG [11].

Thromboprophylaxis guidelines were created by the Swedish Society of Obstetrics and Gynecology in 1998, and a prospective study was designed to evaluate the efficacy of the use of LMWH for thromboprophylaxis in women with a prior VTE [28]. Over the five‐year study period, 326 women who were prescribed Lovenox for prophylaxis were compared to 1000 controls. With thromboprophylaxis in accordance with the Swedish guidelines, the investigators identified an estimated 88% reduction in the relative risk for recurrent VTE in pregnancy.

These findings are similar to those found in the United Kingdom (UK), whereby thromboprophylaxis guidelines developed in response to the decades old quality and safety initiative “Saving Mothers’ Lives: Confidential Enquiry into Maternal Deaths,” has led to the sharpest decline in maternal mortality in the UK since 1985 [29]. In a review of nearly 1.5 million pregnancies in one of the largest US hospital systems, pulmonary embolism was the causative factor in 9 (10%) of the 95 deaths identified, and thromboembolism was identified as one of the most accessible targets to effectively aim systematic efforts to reduce maternal mortality [7]. After implementation of a protocol for universal use of sequential compression devices at the time of cesarean delivery, (intervention) and reevaluation of the maternal mortality rate within this same system three years after this intervention, the authors found an 86% reduction in maternal mortality from post‐cesarean thromboembolism [30].

The Institute for Healthcare Improvement (IHI) defines the term “bundle” as a way to “describe a collection of processes needed to effectively care for patients undergoing particular treatments with inherent risks.” The goal is to “bundle” together several scientifically grounded, essential elements in the care process essential to improving clinical outcomes. Ideally, bundles are relatively straightforward and uncomplicated, and limited to ensure that the bundle can be feasibly carried out [31]. Another critical component of a bundle is the concept that all components work synergistically and must be completed for it to be effective [31]. In other words, bundles work by an all or nothing approach. The Council on Patient Safety in Women’s Healthcare is a consortium of organizations across women’s healthcare including ACOG, the American Board of Obstetrics and Gynecology (ABOG), the American Academy of Family Physicians, the American College of Nurse Midwives, among others. This consortium uses the term “bundle” to include a collection of materials such as checklists, protocols, educational materials and reporting systems targeted toward a particular morbidity, designed to use a comprehensive treatment approach [32].

The thromboprophylaxis bundle endorsed by the Council on Patient Safety in Women’s Healthcare includes development of national and local tools to recognize and prevent VTE in every patient, protocols for every unit, such as utilizing standardized recommendations for mechanical and pharmacologic prophylaxis and therapy. Additionally standardized recommendations for timing of use of prophylaxis with neuraxial anesthesia, as well as monitoring process metrics and protocol compliance are recommended [32]. The goal is to encourage use of these bundles throughout the United States by the end of 2016, and long‐term data regarding local and regional compliance and outcomes are pending [33]. Some potential limitations to large‐scale data collection is the documented need to tailor guidelines to the needs and capabilities of local systems and populations, the potential for variation in compliance amongst individuals and groups and limitations inherent in large‐scale data collection [34].

Another approach to reducing adverse outcomes when prophylaxis is contraindicated, otherwise not feasible, or fails, is to facilitate early detection and treatment of thromboembolism. Multiple audits and enquiries into root causes of preventable maternal deaths have shown that human failure, specifically failure to recognize the severity of a patient’s condition, failure to act, or communication failure, contributed significantly to the outcome [7, 8, 3537]. Early warning systems, designed to alert care providers to pathologic physiologic parameters that may precede critical illness have been used in fields outside of obstetrics, such as in general medicine [38, 39] and pediatrics wards. A MEOWS was originally proposed in the UK in 2007 as a result of the Confidential Enquiries into Maternal Deaths as a means to systematically improve early identification of women at highest risk for severe morbidity or death [40]. The initially proposed MEOWS allowed any care provider, such as a bedside nurse, to identify abnormal physiologic parameters including: pulse oxygenation, respiratory rate, blood pressure, urine output, level of consciousness, based on color coded values [41]. Values that registered in a “yellow” zone were slightly abnormal, and “red” values were markedly abnormal. To be effective such early warning systems cannot rely upon scoring alone. Instead, mechanisms to encourage appropriate action when abnormal findings are identified must be implemented along with such warning systems. This may include implementing “triggers”; in the above MEOWS example, one “red” or two “yellow” values for any one patient were designed to trigger an action, such as a physician being called to the bedside [41].

In a large, multi‐center pilot study conducted in the United States, a Maternal Early Warning Trigger Tool (MEWT) was developed and implemented in 6 of 29 hospitals in a large hospital system [42]. The MEWT was designed to address the four most common causes of maternal morbidity: hemorrhage, preeclampsia, sepsis, and cardiopulmonary dysfunction. During the 13‐month study period, the MEWT was used in 93.4% of all patients at the study sites. There were 32 patients who were screened for ICU admission at these sites, and of those, 31 required admission. There was a noted 5.5% increase in ICU admission at participating sites, compared to a 8% decrease in ICU admission at nonparticipating sites, but a significant reduction in Center for Disease Control (CDC)‐defined severe maternal morbidity and composite maternal morbidity (18.4% and 13.6% respectively) [42].

Jul 19, 2020 | Posted by in GYNECOLOGY | Comments Off on 52: Obstetric emergencies
Premium Wordpress Themes by UFO Themes