Respiratory failure is an uncommon complication of pregnancy, affecting only 0.1% of pregnant women; however, it is one of the main indications for intensive care unit (ICU) admission of pregnant and peri-partum women. Pregnancy is a unique physiologic state in women’s lives, and physicians should be familiar with it. Given the low incidence of respiratory failure during pregnancy, there is limited scientific evidence about the best strategies for this population.
Respiratory failure is an uncommon complication of pregnancy, affecting only 0.1% of pregnant women; however, it is one of the main indications for intensive care unit (ICU) admission of pregnant and peri-partum women.1 Pregnancy is a unique physiologic state in women’s lives, and physicians should be familiar with it. Given the low incidence of respiratory failure during pregnancy, there is limited scientific evidence about the best strategies for this population. The existing literature is scarce, primarily composed of case reports and expert opinion. This chapter is intended to describe a practical approach to acute respiratory failure and the application of the most important mechanical ventilatory principles during pregnancy, taking into account haemodynamic support, issues of drug safety and monitoring of the fetoplacental unit. Only general and basic concepts of mechanical ventilation are included in this chapter: further and comprehensive discussion of mechanical ventilation is out of the scope of this chapter.
Respiratory Physiology During Pregnancy: Relevant Issues for Mechanical Ventilation
Pregnancy is characterized by progressive and sequential cardiovascular adaptations, reaching maximum adaptation at term, followed by a regression to the non-pregnancy parameters by approximately six weeks post-partum. Several anatomic adaptations are relevant to airway management. These changes consist of pregnancy-induced weight gain, including breast enlargement, respiratory tract mucosal oedema and capillary engorgement of nasal and oropharyngeal mucosa and laryngeal tissues. The diaphragm is displaced upward 4 cm, decreasing functional residual capacity by 10–25% at term. In addition, mechanical compression from the enlarged uterus and a rise in circulating progesterone is associated with delayed gastric emptying and increased residual gastric volume, which increases the risk of aspiration of gastric contents during intubation. Thus, every obstetric airway should be considered a difficult airway (see Chapter 4 for physiological adaptations in pregnancy and Chapter 24 for airway management in pregnancy).
Chapter 4 describes the changes in pulmonary physiology occurring in pregnancy, which may affect oxygen transport. Several variables, including maternal haematocrit, arterial oxygen content and uterine artery perfusion contribute to fetal oxygenation and therefore need to be taken in consideration. Another important parameter to consider is the difference in haemoglobin between adult and fetus. The fetal or neonatal oxyhaemoglobin dissociation curve is shifted to the left as a result of fetal haemoglobin and lower levels of 2,3-diposphoglycerate (2,3-DPG). This facilitates oxygen uptake and transport to the fetus, which has a much lower PO2 (25–30 mmHg) (Figure 25.1). In contrast, there is rightward shift of the maternal oxyhaemoglobin dissociation curve caused by an increase in 2,3-DPG in the erythrocytes.
Figure 25.1 The maternal and fetal oxygen-dissociation curve. The oxygen-dissociation curve of human blood is depicted above. For fetal haemoglobin, the normal curve is shifted to the left, while in the mother it is the opposite.
General Concepts Regarding Mechanical Ventilation and Indications for Mechanical Ventilation During Pregnancy and Puerperium
The modes of mechanical ventilation (MV) are commonly defined by four elements determining the phases of the respiratory cycle (Figure 25.2):
Trigger phase: initiates a breath
Inspiration phase: initiates when a given flow or pressure is generated by the ventilator
The target (or controlled) phase: is the pressure or flow that will be maintained until the inspiration ends
The cycling phase determines the end of the inspiratory phase.
Figure 25.2 Flow-time tracing demonstrating phases of the respiratory cycle. 1. Triggering and initiation of inspiration; 2. Inspiratory phase; 3. Cycling phase (end of inspiration); 4. Expiratory phase.
According to interactions between patient and ventilator, MV modes are also classified as:
Fully controlled: when trigger and cycling are time controlled, the target variable is reached passively in a time basis, and the patient does not actively contribute to breath.
Partially supported or assisted: if combination of ventilator assistance and patient effort occurs in the same cycle.
Unassisted: when the inspiratory flow is generated entirely by the patient’s respiratory muscles.
In general, conventional mechanical ventilation modes are:
CMV: continuous mandatory ventilation, in volume (VC-CMV) or in pressure (PC-CMV)
IMV: intermittent mandatory ventilation
SIMV: synchronized intermittent mandatory ventilation
PSV: pressure support ventilation
CPAP: Continuous positive airway pressure
Other unconventional modes of MV are:
PRVC: Pressure-regulated volume control (which delivers pressure-targeted breaths, varying from breath to breath to reach a target volume)
APRV: airway pressure release ventilation
PAV: proportional assist ventilation
NAVA: neurally adjusted ventilatory assist
All these modes (conventional and unconventional) have some variation in terms of the level of interaction between patient and ventilator, from a fully controlled trigger and cycling mode, to partially supported or assisted (a combination of ventilator assistance and patient effort). Initial ventilator settings may differ, depending on the critical illness leading to respiratory failure, and on the severity of respiratory failure. Additionally, current recommendations for tidal volume when setting the ventilator are based on predicted body weight (based on height) and not actual body weight because lung size does not increase with the increase in body weight seen during pregnancy.
Oxygen delivery can be impaired for conditions that worsen cardiac output, arterial oxygen content, or both. Indications for MV in obstetric patients are similar to those in the general population, including airway protection and inadequate oxygenation or ventilation (adjusted to the physiological changes of pregnancy).2 The limited literature published on this topic suggest that respiratory failure and the need for ventilator support may be due to obstetric causes such as pre-eclampsia, or to non-obstetric causes.3
There are no studies evaluating a conventional MV mode over another during pregnancy, so the decision on MV modes is more related to the physician’s specific knowledge and expertise. Table 25.1 describes the suggested oxygenation goals and parameters for pregnant women according to the main causes of respiratory failure described above. Maintaining a PaO2 value greater than 70 mmHg but not greater than 120 mmHg would be desirable, avoiding hyperoxaemia, which could result in worse maternal outcomes. Although oxygen therapy has been considered part of intrauterine fetal resuscitation, recent studies suggest that caution should be exercised when using high concentrations of maternal oxygenation.4 A theoretic model based on animal studies showed that a maternal drop in O2 saturation from 95% to 88% resulted in desaturation of umbilical venous blood. However, it is important to remember that information derived from sheep may not be applicable to humans.5 The limited information published on the methods and settings of mechanical ventilation in pregnant patients suggest that settings are similar to those used in the general population, aiming to maintain desirable physiological variables in obstetrics. Given the limited data on oxygenation and carbon dioxide goals available for the pregnant patient, a similar approach to that used in the non-pregnant patient is appropriate, recognizing that:
Hypoxaemia may be harmful to the fetus
Hypocapnia and alkalosis reduces placental perfusion
The effects of hypercapnia on the fetus are unknown and risks need to be weighed against benefits.
|Goals of mechanical ventilation: PaO2 >70 mmHg or SO2 >95%; PaCO2: 28–32 mmHg; HCO3: 18–22 mmHg|
|Parameter||General causes inducing respiratory failure|
|Acute respiratory distress syndrome||Pulmonary embolism||Pulmonary oedema related to pre-eclampsia||Near-fatal asthma|
|Plateau pressure||~30 cmH2O and driving pressure <15 cmH2O||<30 cmH2O and driving pressure <15 cmH2O||<30 cmH2O and driving pressure <15 cmH2O||<30 cmH2O and driving pressure <15 cmH2O|
|Tidal volume||6 ml/kg||Mild hyperventilation||6–8 ml/kg||6–8 ml/kg, with a prolonged expiratory time and decreasing respiratory rate|
|Positive end-expiratory pressure (PEEP)||Adjusted to achieve optimum oxygenation (PaO2 >95%)||Usually not necessary to titrate because the lack of response of hypoxaemia to PEEP levels||Often helpful to improve cardiac output, but trying to avoid under filling in preload-dependent patients||The minimum required to avoid maternal haemodynamic impact secondary to dynamic hyperinflation|
|Permissive hypercapnia||<60 mmHg||Avoid as possible, because of worsening in right ventricle afterload||N/A||Usually a PCO2 <60 mmHg|
|Non-invasive mechanical ventilation (NIVM)||Not recommended||Not recommended||Recommended||Recommended on a case-by-case basis|
However, given a decrease in chest wall compliance caused by the enlarging uterus and breast enlargement, usual airway pressures may fail to produce appropriate tidal volumes at the conventional limit of 30 cmH2O. Thus, even higher (<35 cmH2O) plateau pressure (Pplat) might be appropriate and safe, but with a concomitant driving pressure (Pplat – PEEP) <15 cmH2O. Common lung-protective strategies, such as permissive hypercapnia, have not been validated in the obstetric population, but are commonly used in practice. However, animal models of hypercapnia have shown adequate fetal and neonatal tolerance,6 and has also been reported in isolated case reports of fetal tolerance to hypercapnia in humans.7 Furthermore, maternal transcutaneous partial pressure of carbon dioxide (tcPCO2) values in pregnant women not in labour have also shown that fetuses exposed to hypercapnia have a better APGAR and parameters suggesting higher neonatal cerebral oxygenation compared to babies exposed to maternal hypocapnia.8
Acute Respiratory Failure During Pregnancy: Prevalence and Epidemiology
Respiratory failure is a syndrome that occurs when one or both functions of the respiratory system (i.e. oxygenation and carbon dioxide elimination) fail. Thus, respiratory failure is classified as either hypoxaemic or hypercapnic. Hypoxaemic respiratory failure is characterized by an arterial partial pressure of oxygen (PaO2) less than 60 mmHg with a normal or low arterial partial pressure of carbon dioxide (PaCO2). On the other hand, an elevated PaCO2 (probably more than 40 mmHg in pregnancy, given the physiological respiratory alkalosis) characterizes hypercapnic respiratory failure. Clinically, this is recognized as tachypnoea, tachycardia, intercostal muscle retraction, accessory muscle use, diaphoresis and paradoxical breathing. It is estimated that respiratory failure affects 0.1–0.2% of pregnancies.9 Below is a brief outline of suggested goals and ventilatory parameters in selected conditions associated with respiratory failure during pregnancy. Management of the more common conditions associated with respiratory failure in pregnant women are explained in detail in Chapters 7, 9, 10, 14 and 18.
Pulmonary Thromboembolic Disease
The challenge of treating pregnant women with massive pulmonary embolism is even greater than in the general population. In particular, maternal bleeding is a major concern with anticoagulation, and pregnancy is thought to be a relative contraindication to thrombolysis. However, a recent systematic review of thrombolysis, embolectomy and extracorporeal membrane oxygenation (ECMO) identified 127 cases during pregnancy and the puerperium, and found thrombolysis was effective in most of the cases in which it was tried, albeit with a high risk of maternal bleeding; the risk of major bleeding was 17.5% when thrombolysis was given antepartum and 58.3% when it was used post-partum.10 In addition to clot-burden reduction strategies and anticoagulation, optimization of right ventricle (RV) function by decreasing RV preload, providing inotropic support and improving RV afterload are the cornerstones of management. RV dysfunction due to pulmonary embolism during pregnancy should include a RV-protective approach, involving three main components: reducing lung stress by limitation of Pplat and driving pressure, improving oxygenation to reverse hypoxic pulmonary vasoconstriction and reducing hypercapnia. For this, trying to maintain a mild hyperventilation and non-PEEP-dependent oxygenation would help to improve pulmonary hypertension.
Acute Respiratory Distress Syndrome
Acute respiratory distress syndrome (ARDS) is a diffuse inflammatory lung injury characterized by acute-onset, non-cardiogenic pulmonary oedema with hypoxaemia and partial pressure of oxygen to inspired oxygen ratio <300.11 The incidence of ARDS during pregnancy has been estimated just recently to be 36.5 cases per 100 000 pregnancies.12 In pregnant patients, non-obstetric causes for ARDS include sepsis due to pyelonephritis, pneumonia, intracerebral infection and haemorrhage. Obstetric causes include severe pre-eclampsia, amniotic fluid embolism and septic abortion. This wide range suggests variation in the criteria used to define the syndrome. The criteria concerning ARDS have been re-examined and a recent consensus (Berlin criteria)11 has proposed the following definition:
Timing of onset occurs within one week of a known insult or worsening respiratory symptoms.
Chest imaging revealing bilateral opacities not fully explained by effusions, lobar collapse or nodules as observed on chest X-ray or chest tomography
Origin of the radiographic opacities and respiratory failure not fully explained by cardiac failure or volume overload
Oxygenation deficits: Mild: PaO2/FiO2 range 201–300 with PEEP or CPAP 5 cmH2O; Moderate: PaO2/FiO2 range 101–200 with PEEP or CPAP 5 cmH2O; Severe: PaO2/FiO2 <100 with PEEP or CPAP 5 cmH2O.
The Berlin criteria are important in identifying patients at a higher risk because they grade severity according to indices of oxygenation with PEEP application. The case mortality for ARDS in pregnancy has been reported to range from 11–50%. A recent nationwide inpatient sample from the US reported a case mortality for pregnant patients undergoing mechanical ventilation for ARDS of 9–14%.12 Mortality in this cohort depended on the cause and duration of mechanical ventilation, and identified several risk factors associated with higher in-hospital mortality, including renal failure, puerperal infection, septic obstetric emboli and influenza.12
Pregnant women require a higher level of oxygen supply for the maintenance of fetal demands.13 Recent evidence from the non-obstetric population suggests that PaO2/FiO2 >150 mmHg is more likely to be associated with improved outcomes, and thus, a value of PaO2/FiO2 >150 mmHg might be reasonable in obstetric scenarios. Lung-protective parameters for obstetrics patients are basically similar to those for the non-obstetric population. PEEP is often titrated to the level needed to achieve the optimum oxygenation that would be desirable in pregnancy (SpO2 >95%). However, recent data suggest no benefit from a high (~15 cmH2O) vs a low PEEP (~9 cmH2O) strategy in terms of clinical outcomes. Other recommendations for the management of these patients includes the use of conservative fluid volumes, given its impact on ventilator-free days and a decreased length of ICU stay.14
Asthma affects 8% of pregnant women.15 One-third of pregnant asthmatics experience a worsening of their asthma that may progress to a critical asthma syndrome. Progressive hypoxaemia, respiratory acidosis, maternal fatigue, alteration in mental status and other causes for increased work of breathing despite maximal bronchodilator therapies and non-invasive ventilation are indications for intubation and mechanical ventilation. As expected, only a few case reports describing MV settings in asthmatic women during pregnancy are available. In general, ensuring a moderate tidal volume (6–10 ml/kg), with a low respiratory rate and permissive hypercapnia, trying to avoid and minimize air-trapping (also known as dynamic hyperinflation or auto-PEEP) and using prolonged expiratory times, may remedy the dyssynchrony between the pregnant patient and the ventilator, and ameliorate volutrauma and barotrauma.16 In terms of ventilator support, non-invasive ventilation may be an alternative under the right clinical circumstances.17
Non-invasive Mechanical Ventilation: Possible Implications
Non-invasive ventilation (NIV) is a good option for short-term ventilatory support (24–48 hours), avoiding the potential complications of endotracheal intubation and associated sedation. This strategy is recommended only in patients with an adequate level of consciousness, since there is a risk of aspiration. Either NIV or continuous positive airway pressure (CPAP) are recommended as first choice in patients with cardiogenic pulmonary oedema admitted to an ICU.18 Moreover, NIV decreases endotracheal intubation rates and hospital mortality in acute hypoxaemic respiratory failure due to cardiogenic pulmonary oedema in non-pregnant patients,19 and also in pre-eclamptic patients.20,21 Aspiration and precipitous deterioration are concerns with use of NIV in pregnant women, and given the paucity of data, it should probably only be used by teams with experience in this intervention.
Drugs Used During Mechanical Ventilation in Pregnant Woman, from Intubation to Sedation, and Safety Issues
Long-term sedation is sometimes necessary during mechanical ventilation in pregnant women, and the effects of sedatives in this setting are of major concern for clinicians. Table 25.2 describes sedatives and neuromuscular blockers commonly used in pregnant ICU patients. The principal factors allowing medications to cross the placenta include a molecular weight of less than 400 daltons, high lipid solubility, non-ionized state and low protein binding. However, most drugs transfer via passive diffusion along a concentration gradient. As pregnancy advances, the placenta expands its surface area, allowing greater drug transfer. See also Chapter 21, Respiratory Drug Therapy in Pregnancy.
|Agents||US FDA recommendation||Dose||Comments|
|Sulfate fentanyl||25–100 mg/h||Maternal use of fentanyl may cause withdrawal symptoms and respiratory depression in the newborn infant|
|Remifentanil||0.5 mg/kg bolus followed by infusion at 0.5–2 mg/kg/min||Use during pregnancy may cause neonatal opioid withdrawal syndrome|
|Cisatracurium||There have been no demonstrated adverse effects in the fetus or the newborn infant|
|Vecuronium||Bolus of 0.1 mg/kg followed by infusion at 1–2 mg/kg/min|
Critical care guidelines suggest that prolonged sedation strategies using non-benzodiazepine sedatives (either propofol or dexmedetomidine, US Food and Drug Administration (FDA) class B and C, respectively, during pregnancy) may be preferred over sedation with benzodiazepines (either midazolam or lorazepam) to improve clinical outcomes in mechanically ventilated adult ICU patients. A recent report issued from the FDA, advised that ‘repeated or lengthy use (>3 hours), of general anaesthetic and sedation drugs during surgeries or procedures in children younger than 3 years or in pregnant women during their third trimester may affect the development of children’s brains’.22 This warning implicated agents such as benzodiazepines (specifically midazolam) and propofol. The warning was based on concerns about the exposure of the fetal brain to these agents when synaptogenesis is occurring, late in gestation up to early infancy.22 There has been some resistance to this warning, since the studies used were based on animal experimentation, and many of the pre-clinical studies were using doses far larger than those used in clinical practice.23,24 Further research is needed to clarify the scope of this warning in the setting of long-term sedation in a pregnant woman. The only other drugs in clinical use for sedation and anaesthesia that may not cause neurodegenerative damage in animal models are dexmedetomidine and opioids, and although there is considerably less experience compared with the usual agents, it might represent an alternative choice in pregnant patients. No dosing adjustment is recommended in pregnant patients, and no delirium scale has been validated in obstetrics, so similar detecting and monitoring for delirium is applied, using the most valid and reliable delirium monitoring tools in adult ICU patients. These include the Confusion Assessment Method for the ICU (CAM-ICU) and the Intensive Care Delirium Screening Checklist (ICDSC).
There is evidence supporting the use of neuromuscular-blocking agents (NMBAs) in theseverely acute hypoxaemic non-pregnant population, but there is a lack of recommendation in pregnant patients. However, these medications have been used extensively in pregnant patients. All NMBAs or their metabolites, except for cisatracurium, cross the placental barrier. Previous studies performed during caesarean section have suggested a low passage of NMBAs to the fetal circulation.25 However, there are no studies describing fetal concentrations of NMBAs for long-term infusions, and its safety during the first trimester is unknown since animal studies have reported none (rocuronium and pancuronium) or minimal teratogenic effect (atracurium). In vitro evidence has suggested a potential toxic effect during organogenesis.26,27 In addition, vecuronium and atracurium have been shown to have residual clinical effects in the newborn (abnormal neurobehavioural adaptive capacity) and should be avoided near delivery.28 Since transplacental passage of NMBAs will increase with higher injected doses and with longer infusion-to-delivery intervals, using the lowest possible dose of NMBAs with a low umbilical/maternal venous (UV/MV) ratio and short duration of action appears to be the safest choice.29 Suggested doses of sedatives and NMBAs are described in Table 25.3, along with the FDA safety recommendation for pregnancy and the potential transfer to the developing fetus.