Critical Care in Pregnancy

Recent case series suggest that between 0.1% and 0.8% of obstetric patients are admitted to a traditional intensive care unit (ICU) (1–9). Among this population, the risk of death ranges from 2% to 11%. Although survival is better than that in an unselected population, mortality is substantially higher than the maternal mortality ratio in the developed world. In addition, approximately 1–2% of pregnant women receive critical care outside of a traditional ICU but within a specialized obstetric care unit (10, 11). Therefore, overall estimates suggest that 1–3% of pregnant women require critical care services in the United States each year, approximately 40,000– 120,000 women (based on 4 million births per year) (12).


Background
Recent case series suggest that between 0.1% and 0.8% of obstetric patients are admitted to a traditional intensive care unit (ICU) (1)(2)(3)(4)(5)(6)(7)(8)(9). Among this population, the risk of death ranges from 2% to 11%. Although survival is better than that in an unselected population, mortality is substantially higher than the maternal mortality ratio in the developed world. In addition, approximately 1-2% of pregnant women receive critical care outside of a traditional ICU but within a specialized obstetric care unit (10,11). Therefore, overall estimates suggest that 1-3% of pregnant women require critical care services in the United States each year, approximately 40,000-120,000 women (based on 4 million births per year) (12).

Techniques for Critical Care
Competence in a number of core procedural skills is necessary for any physician practicing in critical care. Except in very large maternity centers, there is little opportunity to develop or maintain these skills, although alternative means may be available. Some techniques can be performed under supervision in other settings (eg, airway techniques under anesthesiologists' supervision in the operating room), others can be rehearsed through electronic resources or online (electrocardiogram interpretation), and still others lend themselves to task training through medical simulation.

Organization of Critical Care
In an open ICU, any physician can write orders or perform procedures; management or consultation by a qualified intensive care physician is not mandatory. In a closed ICU, only the critical care attending physician or house staff can write orders and manage patient care. The hybrid or transitional model allows all physicians to write orders but requires an on-site critical care physician to provide consultation, conduct rounds, or comanage all patient care. Additional terminology classifies ICU physician staffing as high intensity (closed ICU or mandatory intensivist consultation) or low intensity (optional intensivist) (13). High-intensity ICU physician staffing is associated with lower ICU mortality, lower hospital mortality, and decreased length of stay in both the ICU and the hospital compared with low-intensity ICU physician staffing (13). Although there are limited data specifically addressing the critical care obstetric patient, it seems reasonable that these findings would apply to this population as well (14, 15).

Critical Care in Pregnancy
Critical care in pregnancy is a field that remains unevenly researched. Although there is a body of evidence to guide many recommendations in critical care, limited research specifically addresses obstetric critical care. The purpose of this document is to review the available evidence, propose strategies for care, and highlight the need for additional research. Much of the review will, of necessity, focus on general principles of critical care, extrapolating where possible to obstetric critical care.
Critical care requires a multidisciplinary approach to achieve the best outcomes (16). The usual ICU team includes physicians, nurses, pharmacists, and respiratory therapists. In critical care obstetrics, the team also should include obstetricians or maternal-fetal medicine subspecialists, obstetric nurses, and neonatologists.
The obstetrician transferring a patient to an ICU must be familiar with the types of units available at the institution, such as a general medical-surgical ICU or a specialty unit for cardiothoracic or neurologic or neurosurgical care as well as the role of the obstetrician within each unit (17).
There are three levels of adult critical care described by the American College of Critical Care Medicine, with Level I delivering the highest level of care. Alternatives to traditional ICU care include intermediate-care or high-dependency units such as post-ICU step-down units, telemetry units for cardiac patients, and units for patients requiring long-term ventilator use (10,11,18).

Admission to Intensive Care
Because ICU beds are a scarce resource, ICU admission should be reserved for those patients who are likely to benefit. Most obstetric patients will be admitted under the objective parameters triage model. In this model, specific criteria trigger ICU admission, regardless of diagnosis. These triggers were achieved by consensus, in response to the review by the Joint Commission (formerly the Joint Commission on Accreditation of Healthcare Organizations) and are acknowledged to be largely arbitrary. They include specific abnormalities in vital signs, laboratory values, and imaging and physical findings. Admission criteria for nonpregnant patients are listed in the box "Objective Parameters Model for Admission of Nonpregnant Patients to an Intensive Care Unit." A description of changes to normal laboratory values during pregnancy is provided in the box "Key Laboratory Values That Are Different in Pregnancy."

Considerations in Transfer
The care of any pregnant woman requiring ICU services should be managed in a facility with obstetric adult ICU and neonatal ICU capability. Standard guidelines for perinatal transfer have been published by the American College of Obstetricians and Gynecologists and the American Academy of Pediatrics and follow federal Emergency Medical Treatment and Labor Act guidelines (19). These guidelines generally recommend antenatal rather than neonatal transfer and describe the responsibilities of the referring and receiving hospitals. In the event that maternal transport is unsafe or impossible, alternative arrangements for neonatal transport may be necessary. In patients where imminent delivery is expected, transfer should be held until after delivery.
The minimal monitoring required for a critically ill patient during transport includes continuous pulse oximetry and electrocardiography as well as regular assessment of vital signs (19, 20). All critically ill patients must have secure venous access before transport. Patients who already have arterial or central lines or other invasive monitoring devices in place should have those monitored as well. Women who are mechanically ventilated must have the endotracheal tube position confirmed and secured before transport and must be assessed for adequacy of oxygenation and ventilation.
There are no data to guide obstetric monitoring during transport of the critically ill obstetric patient. Fetal and tocodynamic monitoring during transport is likely feasible but of unproven utility (21). In some circumstances, identifying fetal compromise in transit may allow for advance preparation for intervention, including delivery, by the receiving institution. Simple measures, such as left uterine displacement and supplemental oxygen, should be applied routinely during transport.

What are the findings in a pregnant or postpartum woman that might prompt ICU admission?
Patients should be transferred to an ICU if they need circulatory or pulmonary support. An obstetric service should adopt site-specific guidelines for transfer based on the level of care required. These guidelines also should define and distinguish between levels of care that can be provided on the labor floor or, if applicable, the obstetric intermediate care unit.
Hemorrhage and hypertension are the most common causes of ICU admission in obstetric patients (1-7, 9, 22-36). Most of these patients typically require relatively simple interventions, monitoring, and supportive care.
Approximately 20-30% of obstetric ICU patients have nonobstetric causes for an ICU admission, such as sepsis (1,5,7,11). Early goal-directed therapy for sepsis should not be delayed until the admission to the ICU but should begin as soon as septic shock is diagnosed (37)(38)(39). The patient should be stabilized, intravenous access maintained, urine output and fluid volume managed, and antibiotics started for treatment of sepsis.
Broad-spectrum antibiotic therapy should be started within 1 hour of the diagnosis of severe sepsis or septic shock (37 the neonatologist also will need to advocate for bringing together the critically ill mother and her baby. Multidisciplinary care is essential for the critically ill obstetric patient. When a pregnant patient is transferred to the ICU, members of the care team should assess the anticipated course of her condition or disease, including possible complications, and set parameters for delivery, if appropriate. The plan should be clear to the medical team and to the patient's family, and to the patient herself if she is able to understand. Because the risk-benefit calculation for a given intervention may change as pregnancy progresses, it is important to reevaluate the care plan on a regular basis. The plan for delivery should be made long before delivery is imminent, and it must include decisions about preferred location for delivery, preferred mode of delivery (vaginal versus cesarean), need for analgesia or anesthesia, and access to pediatricians. It also must include an alternative plan or set of plans in the event that the original plan cannot be followed.
If postpartum ICU admission is necessary, the patient and her family may have questions regarding the obstetric events that precipitated transfer even when obstetric care has been optimal. Anger, dissatisfaction, or legal action often follow a perceived bad outcome; in the case of a postpartum complication or condition requiring critical care, the obstetrician may bear the brunt of these questions. Although a full discussion about disclosure and review of adverse events is beyond the scope of this document, resources are available to assist the obstetrician through this stressful time (40).

How should care be organized when a laboring patient needs critical care?
If a laboring patient requires critical care services, it is important to determine the optimal setting for her care. If the fetus is previable or the duration of ICU services is anticipated to be lengthy, the labor floor is unlikely to be the best option. However, during active labor, the labor unit may be the best choice if adequate maternal support can be provided. Advantages of vaginal delivery in the ICU include the availability of critical care interventions and staff. Disadvantages include lack of space to conduct a vaginal delivery and to accommodate pediatric personnel and equipment, and unfamiliarity of critical care personnel with obstetric interventions and management. Factors that will affect this decision include the degree of patient instability, interventions required, staffing and expertise available, anticipated duration of ICU stay, and probability of delivery.
Delivery in the ICU is associated with an increased likelihood of operative vaginal delivery. In part, this is obtained but should not delay initiating antibiotics. Fetal resuscitation in utero through maternal oxygen therapy and circulatory support is preferable to cesarean delivery for nonreassuring fetal heart rate patterns in most cases.
Approximately 75% of obstetric ICU patients admitted to the ICU are postpartum (5,6,25). This may be due to specific postpartum causes, such as postpartum hemorrhage, or to ascertainment bias; obstetricians may be less willing to transfer or intensivists less willing to accept a patient whose fetus must still be considered in care planning. The process of transport itself is risky for a critically ill patient, who requires ongoing monitoring and maintenance while in transport. In a viable pregnancy, fetal monitoring during transport between obstetric units and the ICU may be prudent, especially if the patient is in labor (20).
Admission to the ICU is dependent on levels of care available on a local level. Patients needing the following procedures should be treated in a critical care unit: 1. Respiratory support, including airway maintenance and endotracheal intubation 2. Treatment of pneumothorax 3. Cardiovascular support, including treatment with pressors 4. Pulmonary artery catheterization (insertion, maintenance, and interpretation)

What is the obstetrician-gynecologist's role in the transfer of a patient to a critical care unit?
When obstetric patients are transferred to the ICU, the obstetrician's role will depend on the ICU model (open or closed) and the patient's status (antepartum or postpartum). Regardless of the primary caregiver, patient care decisions must be made collaboratively between the intensivist, obstetrician, and neonatologist, and should involve the patient, her family, or both. Obstetric input in the care of the postpartum ICU patient may include evaluation of vaginal or intraabdominal bleeding, evaluation of obstetric sources of infection, duration of specific therapies, such as magnesium for eclampsia prophylaxis, and feasibility of breastfeeding, especially compatibility of various medications with breastfeeding. There may be issues related to surgical interventions, including reexploration of the abdomen, or reclosure of abdominal or vaginal incisions. Under some circumstances, the obstetrician and Copyright ª by The American College of Obstetricians and Gynecologists. Published by Wolters Kluwer Health, Inc.
Unauthorized reproduction of this article is prohibited.
membranes and multiple gestations. Corticosteroids also may be considered for pregnant women starting at 23 weeks of gestation who are at risk of preterm delivery within 7 days, irrespective of membrane status (43, 44). Administration of betamethasone may be considered in pregnant women between 34 0/7 weeks and 36 6/7 weeks of gestation at imminent risk of preterm birth within 7 days, and who have not received a previous course of antenatal corticosteroids (44). Antenatal corticosteroids are not contraindicated in an ICU setting, even in the setting of sepsis (43). Pregnancy often modifies drug effects or serum levels. Drugs that cross the placenta may have fetal effects; for example, sedative or parasympatholytic drugs can affect the fetal heart rate tracing. However, necessary medications should not be withheld from a pregnant woman because of fetal concerns. Additionally, necessary imaging studies should not be withheld out of potential concern for fetal status, although attempts should be made to limit fetal radiation exposure during diagnostic testing.
Fetal surveillance often is used when a pregnant patient is admitted to the ICU. Because fetal heart rate monitoring reflects uteroplacental perfusion and maternal acid-base status, changes in baseline variability or the new onset of decelerations may serve as an early warning system for derangements in maternal end-organ status. Changes in fetal monitoring should prompt reassessment of maternal mean arterial pressure, hypoxia, acidemia, or compression of the inferior vena cava by the gravid uterus. Correction of these factors may result in improvement of the tracing and every attempt should be made at intrauterine fetal resuscitation.

Is intraoperative fetal monitoring needed for a pregnant patient?
Although there are no data to allow for a specific recommendation regarding fetal monitoring for nonobstetric surgery, it is important for physicians to obtain obstetric consultation before performing nonobstetric surgery. Obstetricians are uniquely qualified to discuss aspects of maternal physiology and anatomy that may affect intraoperative maternal-fetal well-being. The decision to use fetal monitoring should be individualized and, if used, may be based on gestational age, type of surgery, and facilities available. Ultimately, each case warrants a team approach (anesthesia, obstetrics, and surgery) for optimal maternal and fetal safety.

When is perimortem cesarean delivery appropriate?
Although there are no clear guidelines regarding perimortem cesarean delivery, fetal survival is unlikely if because patients with translaryngeal intubation cannot close the glottis to push; therefore, an assisted second stage of labor may be required. In addition, ICU patients often have underlying cardiac or neurologic processes for which an assisted second stage of labor is recommended. Adequate analgesia is required, although assessment of pain may be complicated by altered mental status or intubation. Regional analgesia is preferred but may not be possible because of coagulopathy, hemodynamic instability, or difficulties with patient positioning. Parenteral narcotics can be used instead of regional analgesia but are less effective in preventing pain; suboptimally treated pain may result in hemodynamic derangements that must be anticipated and treated.
Cesarean delivery in the ICU is complex and has significant disadvantages compared with procedures performed in a traditional operating room. These disadvantages include inadequate space for anesthetic, surgical, and neonatal resuscitation equipment and attendant personnel unfamiliar with the operation. In addition, ICUs have the highest rates of health care-associated infections in a hospital, so the risk of nosocomial infection with drugresistant organisms is increased (41, 42). Cesarean delivery in the ICU should be restricted to cases in which transport to the operating room or delivery room cannot be achieved safely or expeditiously, or to a perimortem procedure.

Are there special fetal considerations in the care of a pregnant woman in a critical care setting (eg, assessment of gestational age, fetal monitoring, or complications related to medications)?
Establishment of gestational age is crucial to determine whether the fetus is of gestational age sufficient to ensure a good chance of survival after birth. When possible, prenatal care records should be obtained to establish the most accurate dating criteria. In the event that gestational age remains uncertain, prompt ultrasound evaluation should establish the best possible estimate with documentation of the potential range of uncertainty.
Use of obstetric medications may pose particular challenges in the critically ill patient; known side effects must be carefully monitored and risk-benefit ratios should be assessed in each individual situation. Examples of common drug-related side effects include tachycardia and decreased blood pressure with beta-agonists, effects on platelet function and renal perfusion with indomethacin, and negative inotropic effects on cardiac function with magnesium. A single course of corticosteroids is recommended for pregnant women between 24 0/7 weeks and 33 6/7 weeks of gestation, including those with ruptured Copyright ª by The American College of Obstetricians and Gynecologists. Published by Wolters Kluwer Health, Inc.
Unauthorized reproduction of this article is prohibited.
e152 Practice Bulletin Critical Care in Pregnancy

OBSTETRICS & GYNECOLOGY
The care of any pregnant woman requiring ICU services should be managed in a facility with obstetric adult ICU and neonatal ICU capability.
Necessary medications should not be withheld from a pregnant woman because of fetal concerns.
Necessary imaging studies should not be withheld out of potential concern for fetal status, although attempts should be made to limit fetal radiation exposure during diagnostic testing.

Proposed Performance Measure
Percentage of pregnant or postpartum patients in the ICU who have documented involvement of an obstetriciangynecologist ore than 15-20 minutes have passed since the loss of maternal vital signs. There are insufficient data on which to base conclusions regarding the appropriateness of cesarean delivery when efforts at resuscitation have failed. Based on isolated case reports, cesarean delivery should be considered for both maternal and fetal benefit approximately 4 minutes after a woman has experienced cardiopulmonary arrest in the third trimester (45, 46).

The following conclusions are based on good and consistent scientific evidence (Level A):
Pregnancy changes normal laboratory values and physiologic parameters.
Approximately 75% of obstetric ICU patients are admitted to the unit postpartum.
Hemorrhage and hypertension are the most common causes of admission from obstetric services to intensive care.

The following recommendations are based on limited or inconsistent scientific evidence (Level B):
Cesarean delivery in the ICU should be restricted to cases in which transport to the operating room or delivery room cannot be achieved safely or expeditiously, or to a perimortem procedure.
Treatment of sepsis should not await admission to an ICU but should begin as soon as septic shock is diagnosed.

The following recommendations and conclusions are based primarily on consensus and expert opinion (Level C):
High-intensity ICU physician staffing is associated with lower ICU mortality rates, lower hospital mortality rates, and decreased length of stay in both the ICU and a hospital, compared with models in which intensivist consultation is optional.
Decisions about care for a pregnant patient in the ICU should be made collaboratively with the intensivist, obstetrician, specialty nurses, and neonatologist.
Copyright ª by The American College of Obstetricians and Gynecologists. Published by Wolters Kluwer Health, Inc.
Unauthorized reproduction of this article is prohibited.
he MEDLINE database, the Cochrane Library, and the American College of Obstetricians and Gynecologists' own internal resources and documents were used to conduct a lit er a ture search to lo cate rel e vant ar ti cles pub lished be tween January 1985-January 2008. The search was re strict ed to ar ti cles pub lished in the English lan guage. Pri or i ty was given to articles re port ing results of orig i nal re search, although re view ar ti cles and com men tar ies also were consulted. Ab stracts of re search pre sent ed at sym posia and sci en tif ic con fer enc es were not con sid ered adequate for in clu sion in this doc u ment. Guide lines pub lished by or ga ni za tions or in sti tu tions such as the Na tion al In sti tutes of Health and the Amer i can Col lege of Ob ste tri cians and Gy ne col o gists were re viewed, and ad di tion al studies were located by re view ing bib liographies of identified articles. When re li able research was not available, expert opinions from ob ste tri cian-gynecologists were used. Studies were reviewed and evaluated for qual i ty ac cord ing to the method outlined by the U.S. Pre ven tive Services Task Force:

I
Evidence obtained from at least one prop er ly de signed randomized controlled trial. II-1 Evidence obtained from well-designed con trolled tri als without randomization. II-2 Evidence obtained from well-designed co hort or case-control analytic studies, pref er a bly from more than one center or research group. II-3 Evidence obtained from multiple time series with or with out the intervention. Dra mat ic re sults in un controlled ex per i ments also could be regarded as this type of ev i dence. III Opinions of respected authorities, based on clin i cal ex pe ri ence, descriptive stud ies, or re ports of ex pert committees. Based on the highest level of evidence found in the data, recommendations are provided and grad ed ac cord ing to the following categories: Level A-Recommendations are based on good and consis tent sci en tif ic evidence. Level B-Recommendations are based on limited or in consis tent scientific evidence. Level C-Recommendations are based primarily on consen sus and expert opinion.