Abstract
Maternal morbidity and mortality caused by aspiration of gastric contents in anesthetized parturients with an unprotected airway was recognized by early pioneers in the use of anesthesia to facilitate childbirth. However, because of changes in anesthesia management in the past century, the aspiration of gastric contents is an increasingly rare complication of childbirth. Parturients, however, continue to remain at higher risk in the setting of general anesthesia, both during induction of, and emergence from, general anesthesia. Because of the physiologic changes of pregnancy, pregnant women have an increased risk for difficult mask ventilation and tracheal intubation, and oxygen desaturation occurs more quickly in pregnant than in nonpregnant patients. Anesthesia providers should consider administrating aspiration prophylaxis with a nonparticulate antacid, histamine-2 receptor antagonists, and metoclopramide to women at increased risk for aspiration. Fasting guidelines in preparation for elective surgery mimic those for nonpregnant patients. Gastric emptying is not delayed by pregnancy, and clearance of liquid occurs within 2 hours at term gestation in both normal weight and obese parturients. Current guidelines allow oral fluid intake during labor. Ingestion of solid food increases gastric volume without improving obstetric outcomes and should be discouraged in labor.
Keywords
Aspiration, Mandelson’s syndrome, Prophylaxis, Nil per os, Fasting during labor
Chapter Outline
History, 671
Incidence, Morbidity, and Mortality, 671
Gastroesophageal Anatomy and Physiology, 672
Esophagus, 672
Gastrointestinal Motility, 673
Gastric Secretion, 674
Ingestion of Food, 674
Effects of Pregnancy on Gastric Function, 675
Risk Factors for Aspiration Pneumonitis, 675
Pathophysiology, 677
Clinical Course, 678
Treatment, 678
Prophylaxis, 681
Preoperative Oral Fluid Administration, 682
Choice of Anesthesia, 682
Antacids, 682
Histamine-2 Receptor Antagonists, 683
Proton-Pump Inhibitors, 683
Metoclopramide, 683
Sellick Maneuver and Induction of Anesthesia, 684
Recommendations for Cesarean Delivery, 684
Gastric Ultrasonography, 685
Oral Intake during Labor, 686
History
In 1848, Sir James Simpson first suggested aspiration as a cause of death during anesthesia. Hannah Greener, a 15-year-old given chloroform for a toenail extraction, became cyanotic and “sputtered” during the anesthetic. A “rattling in her throat” then developed, and she soon died. Her physician administered water and brandy by mouth. Simpson contended that it was the aspiration of water and brandy, and not the adverse effects from the chloroform, that caused her death. In 1940, Hall published a report of 15 cases of aspiration, 14 of which occurred in mothers receiving inhalation anesthesia for a vaginal or cesarean delivery. Among the 14 obstetric cases, 5 mothers died.
Subsequently, Curtis Mendelson, in a landmark paper, reported a series of animal experiments that clearly described the clinical course and pathology of pulmonary acid aspiration. In the same paper, Mendelson audited 44,016 deliveries at the New York Lying-In Hospital between 1932 and 1945. He identified 66 (0.15%) cases of aspiration, of which the aspirated material was recorded in 45 cases; 40 mothers aspirated liquid, and 5 aspirated solid food. Importantly, no mothers died from aspirated liquid, but 2 mothers died from asphyxiation caused by the aspiration of solid food. At this time general anesthesia usually involved the inhalation of ether, often as Mendelson observed, by “a new and inexperienced intern.” Mendelson therefore advocated (1) the withholding of food during labor, (2) the greater use of regional anesthesia, (3) the administration of antacids, (4) the emptying of the stomach before administration of general anesthesia, and (5) the competent administration of general anesthesia. This advice became the foundation of obstetric anesthesia practice during subsequent decades.
Incidence, Morbidity, and Mortality
Maternal mortality from pulmonary aspiration of gastric contents has declined to almost negligible levels in the past 3 decades ( Fig. 28.1 ). This decline can probably be attributed to the following factors: (1) the greater use of neuraxial anesthesia; (2) the use of antacids, histamine-2 (H 2 ) receptor antagonists, and/or proton-pump inhibitors; (3) the use of rapid-sequence induction of general anesthesia; (4) an improvement in the training of anesthesia providers; and (5) the establishment and enforcement of nil per os (NPO) policies. Arguably, the common use of neuraxial analgesic/anesthetic techniques, both during labor and for cesarean delivery, is the single most important factor in this remarkable decline in maternal mortality from pulmonary aspiration.
The reported incidence of aspiration pneumonitis depends on the criteria used for making the diagnosis. The relative risk for aspiration in pregnant versus nonpregnant women can best be estimated from comparisons within single-study populations. Olsson et al. reported an overall incidence of aspiration of 1 in 2131 in the general population undergoing anesthesia and 1 in 661 in women undergoing cesarean delivery (i.e., a threefold higher aspiration risk). In two other surveys related to aspiration (one a retrospective review of 172,334 consecutive patients undergoing general anesthesia and the other a review of 133 cases of aspiration from the Australian Anaesthetic Incident Monitoring Study [AIMS]), there were no cases of pulmonary aspiration in women undergoing either elective or emergency cesarean delivery. However, in the latter two studies, emergency surgery was a significant predisposing factor for aspiration; this finding may be relevant for the practice of obstetric anesthesia, given that many obstetric surgical procedures are performed on an urgent or emergency basis. The AIMS study also implicated obesity as a significant risk factor for aspiration.
Morbidity and mortality associated with aspiration vary according to (1) the physical status of the patient, (2) the type and volume of aspirate, (3) the therapy administered, and (4) the criteria used for making the diagnosis. Since 1952, organizations in the United Kingdom have published detailed triennial reports on all maternal deaths. Data from these reports, now administered by the body Mothers and Babies—Reducing Risk through Audits and Confidential Enquiries across the UK (MBRRACE-UK), indicate that death from pulmonary aspiration in obstetrics is vanishingly rare (see Fig. 28.1 ). In the MBRRACE-UK reports from 2009 to 2015, there were no reported maternal deaths from aspiration. Prior reports from 1994 to 2008 identified three maternal deaths from aspiration ; one was an obese parturient, the second was a mother anesthetized 3 days after delivery, and the third was a woman with a placenta previa who required an emergency cesarean delivery after eating a full meal and aspirated on emergence from general anesthesia. Although the number of general anesthetics administered to parturients during this 20-year period (1994 to 2014) is not known, there were approximately 13.8 million deliveries, indicating that the mortality rate from aspiration was less than 1 in 4.6 million deliveries.
Data on pulmonary aspiration in obstetrics in the United States are less comprehensive. The Serious Complication Registry (SCORE) from the Society for Obstetric Anesthesia and Perinatology review of 307,000 deliveries from 30 U.S. centers between 2004 and 2009 reported a failed tracheal intubation rate of 1 in 533, but no related cases of pulmonary aspiration. Data from the U.S. Centers for Disease Control and Prevention’s Pregnancy Mortality Surveillance System suggest that before 1990, aspiration was the most common cause of anesthesia-related maternal death in the United States. It was calculated that at that time there were 17 deaths related to general anesthesia for every 1 death related to regional anesthesia, although it is not clear from these data what proportion of deaths are attributable to aspiration. By the early 1990s, this ratio had improved to 6 to 1. By 2002, death rates for both general and regional anesthesia were similar, likely due in part to a decrease in the risk for aspiration in cases in which general anesthesia was employed. However, mortality statistics are generally a poor predictor of maternal morbidity; several studies have indicated that perioperative aspiration is associated with important morbidity in obstetric patients ; thus all possible measures must be taken to prevent pulmonary aspiration in obstetric patients.
Gastroesophageal Anatomy and Physiology
Esophagus
In adults, the esophagus is approximately 25 cm long and the esophagogastric junction is approximately 40 cm from the incisor teeth. In humans, the proximal one-third of the esophagus is composed of striated muscle, but the distal end contains only smooth muscle. Muscular sphincters at both ends are normally closed. The cricopharyngeal or upper esophageal sphincter prevents the entry of air into the esophagus during respiration, and the gastroesophageal or lower esophageal sphincter prevents the reflux of gastric contents. The lower esophageal sphincter is characterized anatomically and manometrically as a 3-cm zone of specialized muscle that maintains tonic activity. The end-expiratory pressure in the sphincter is 8 to 20 mm Hg above the end-expiratory gastric pressure. The lower esophageal sphincter is kept in place by the phrenoesophageal ligament, which inserts into the esophagus approximately 3 cm above the diaphragmatic opening ( Fig. 28.2 ). The lower esophageal sphincter is not always closed; transient relaxations occur that account for the gastroesophageal reflux that healthy subjects experience.
Gastrointestinal Motility
Differences in fasting and fed patterns of gut motility are firmly established. During fasting, the main component of peristalsis is the migrating motor complex (MMC). Each MMC cycle lasts 90 to 120 minutes and comprises four phases: Phase I has little or no electrical spike activity and thus no measurable contractions; phase II has intermittent spike activity; phase III has spikes of large amplitude and is associated with strong contractile activity; and phase IV is a brief period of intermittent activity leading back to phase I. The MMC first appears in the lower esophageal sphincter and stomach, followed by the duodenum, and finally the terminal ileum, at which time a new cycle begins in the lower esophageal sphincter and stomach. The phase of the MMC at the time of administration of certain drugs can affect absorption and thereby the onset of therapeutic effect. Eating abolishes the MMC and induces a pattern of intermittent spike activity that appears similar to that in phase II. The duration of the fed pattern is determined both by the calorie content and the type of nutrients in the meal.
The stomach, through the processes of receptive relaxation and gastric accommodation, can accept 1.0 to 1.5 L of food before intragastric pressure begins to increase. The contraction waves that propel food into the small intestine begin in the antrum. The pylorus closes midway through the contraction wave, allowing some fluid to exit into the duodenum but causing the remaining fluid to move retrograde toward the body of the stomach. The jet of fluid that exits the pylorus contains primarily liquid and fine particles. Large particles that lag behind are caught in the retrograde flow of fluid, which assists in their disintegration. Therefore, the manner by which individual components of a meal pass through the stomach depends on the particle size and the viscosity of the suspension. Small particles and fluids exit the stomach faster than larger particles. The outlet of the stomach—the pylorus—limits outflow by means of both its chronic tone and its anatomic position. The pylorus is higher than the most dependent portion of the stomach in both the supine and standing positions.
Gastric Secretion
In one day, the stomach produces as much as 1500 mL of highly acidic fluid containing the proteolytic enzyme pepsin. Normal individuals can produce a peak acid output of 38 mmol/h. Acid is secreted at a low basal rate of approximately 10% of maximal output, even when the stomach is empty. There is diurnal variation in this basal rate of gastric acid secretion, with the lowest and highest outputs occurring in the morning and evening, respectively.
The stomach lining has two types of glands: pyloric and oxyntic. The pyloric glands contain chief cells, which secrete pepsinogen, the precursor for pepsin. The oxyntic glands contain the oxyntic cells, which secrete hydrochloric acid. Water molecules and carbon dioxide in the oxyntic cells combine to form carbonic acid, which dissociates into hydrogen ions and bicarbonate. The bicarbonate leaves the cell for the bloodstream, and the hydrogen ions are actively exchanged for potassium ions in the canaliculi connecting with the lumen of the oxyntic gland. The secretions of the oxyntic cell can contain a hydrochloric acid concentration as great as 160 mmol/L (pH 0.8). Proton-pump inhibitors (PPIs) block the hydrogen ion pump on the canaliculi to decrease acid production.
The pylorus contains G cells, which secrete gastrin into the bloodstream when stimulated by the vagus nerve, stomach distention, tactile stimuli, or chemical stimuli (e.g., amino acids, certain peptides). Gastrin binds to gastrin receptors on the oxyntic cell to stimulate the secretion of hydrochloric acid. Acetylcholine binds to muscarinic (M 1 ) receptors on the oxyntic cell to cause an increase in intracellular calcium ion concentration, which results in hydrochloric acid secretion. Histamine potentiates the effects of both acetylcholine and gastrin by combining with H 2 receptors on the oxyntic cell to increase the intracellular cyclic adenosine monophosphate concentration, leading to a dramatic increase in the production of acid. H 2 -receptor antagonists (e.g., ranitidine, famotidine) prevent histamine’s potentiation of acid production ( Fig. 28.3 ).
Ingestion of Food
When a meal is eaten, the mechanisms that control the secretion of gastric juice and the motility and emptying of the stomach interact in a complex manner to coordinate the functions of the stomach. The response to eating is divided into three phases: cephalic, gastric, and intestinal. Chewing, tasting, and smelling cause an increase in the vagal stimulation of the stomach, which in turn increases gastric acid production. This represents the cephalic phase of digestion. In this phase, gastric acid output increases to approximately 55% of peak output. The gastric phase begins with the release of gastrin. Gastric acid secretion depends on antral distention, vagal activity, gastrin concentration, and the composition of the meal. Gastric acid secretion during a mixed-composition meal increases to approximately 80% of peak acid output. The intestinal phase begins with the movement of food into the small intestine and is largely inhibitory. Hormones (e.g., gastrin, cholecystokinin, secretin) and an enterogastric reflex further modulate gastric acid secretion and motility depending on the composition and volume of the food in the duodenum. This inhibition of gastric emptying by food in the duodenum enables the duodenal contents to be processed before more material is delivered from the stomach.
After the ingestion of a meal, gastric emptying depends on (1) the premeal volume, (2) the volume ingested, (3) the composition of the meal, (4) the size of the solids, (5) the amount of gastric secretion, (6) the physical characteristics of the stomach contents entering the duodenum, and (7) patient position. A mixture of liquids and solids passes through the stomach much more slowly than liquids alone. Gastric emptying is slowed by high lipid content, high caloric load, and large particle size. Thus, predicting an exact time for the passage of liquids and solids through the stomach is very difficult. For non-nutrient liquids (e.g., normal saline), the gastric volume decreases exponentially with respect to time. In one study, 90% of a 150-mL saline meal given to fasting adults in the sitting position passed through the stomach in a median time of 14 minutes; however, in adults in the left lateral position, the median time for gastric emptying was 28 minutes. In a subject eating a 400-mL meal of steak, bread, and vanilla ice cream, 800 mL of gastric juice was secreted ( Fig. 28.4 ). Consequently, the volume in the stomach remained high for almost 2 hours despite early, rapid emptying. These studies indicate that the volume and composition of the test meal, the resulting gastric secretions, and even patient positioning can impact gastric emptying and residual gastric content.
Effects of Pregnancy on Gastric Function
Gastroesophageal reflux, resulting in heartburn, is a common complication of late pregnancy. Pregnancy compromises the integrity of the lower esophageal sphincter; it alters the anatomic relationship of the esophagus to the diaphragm and stomach, raises intragastric pressure, and in some women limits the ability of the lower esophageal sphincter to increase its tone. Progesterone, which relaxes smooth muscle, probably accounts for the inability of the lower esophageal sphincter to increase its tone. Lower esophageal pH monitoring has shown a higher incidence of reflux in pregnant women at term, even in those who are asymptomatic, than in nonpregnant controls. Therefore, at term gestation the pregnant woman who requires anesthesia should be regarded as having an incompetent lower esophageal sphincter. These physiologic changes return to their prepregnancy levels by 48 hours after delivery.
Serial studies assessing gastric acidity during pregnancy have proved difficult to perform because they require repeated placement of nasogastric tubes. However, in the most comprehensive study of gastric acid secretion during pregnancy, basal and histamine-augmented gastric acid secretion was measured in 10 controls and 30 pregnant women equally distributed throughout the three trimesters of pregnancy. No significant differences in basal gastric acid secretion were seen between the pregnant and nonpregnant women.
A variety of techniques have been used to study gastric emptying during pregnancy and labor ( Table 28.1 ). Overall, the data suggest that pregnancy does not significantly alter the rate of gastric emptying. In addition, gastric emptying has not been found to be delayed in either obese or nonobese term pregnant women who ingested 300 mL of water after an overnight fast. However, management of obese parturients should take into account the possible presence of other associated problems in this group of patients (e.g., hiatal hernia or difficult airway). Gastric emptying appears to be normal in early labor but becomes delayed as labor advances ; the cause is uncertain. Pain is known to delay gastric emptying, but even when labor pain is abolished with epidural analgesia using a local anesthetic alone, the delay still occurs. Parenteral opioids cause a significant delay in gastric emptying, as do bolus doses of epidural and intrathecal opioids. Continuous epidural infusion of low-dose local anesthetic with fentanyl does not appear to delay gastric emptying until the total dose of fentanyl exceeds 100 µg.
Method of Assessment | Study | Study Period and Subjects | Gastric Emptying |
---|---|---|---|
Radiographic | Hirsheimer et al. (1938) | Labor (10 subjects) | Delay in 2 subjects |
La Salvia and Steffen (1950) | Third trimester and labor | Third trimester: no delay Third trimester + opioids: marked delay Labor: slight delay Labor + opioids: marked delay | |
Crawford (1956) | Labor (12 subjects) | Delay in 1 subject | |
Large-volume test meal | Hunt and Murray (1958) | Serial study Small numbers Second and third trimesters, postpartum | No change |
Double-sampling test meal | Davison et al. (1970) | Third trimester and labor | Labor: delay, with altered pattern of emptying |
Epigastric impedance | O’Sullivan et al. (1987) | Nonpregnant controls, third trimester, 60 minutes postpartum | No delay |
Applied potential tomography | Sandhar et al. (1992) | Sequential study 10 mothers: 37–40 weeks’ gestation, 2–3 days postpartum, 6 weeks postpartum | No delay |
Acetaminophen absorption | Nimmo et al. (1975) | Labor with intramuscular opioids Postpartum 2–5 days | Labor: No delay Labor + opioids: marked delay Postpartum: no delay |
Nimmo et al. (1977) | Labor | Labor: slight delay Labor + epidural analgesia (no opioid): slight delay | |
Simpson et al. (1988) | Nonpregnant controls, 8–11 weeks’ gestation, 12–14 weeks’ gestation | 8–11 weeks: no delay 12–14 weeks: delay | |
Macfie et al. (1991) | Nonpregnant controls, first, second, and third trimesters | No delay in any trimester | |
Geddes et al. (1991) | Postcesarean delivery Epidural fentanyl 100 µg | Delay | |
Gin et al. (1991) | Postpartum: day 1 and day 3, 6 weeks | No delay | |
Wright et al. (1992) | Labor with epidural bolus: (1) bupivacaine 0.375%; (2) bupivacaine 0.375% + fentanyl 100 µg | Epidural opioids: delay | |
Whitehead et al. (1993) | Nonpregnant controls, first, second, and third trimesters Postpartum: 2, 18–24, and 24–48 hours | Pregnancy: No change Postpartum: 2 hours: delay 18–24 hours: no delay 24–48 hours: no delay | |
Ewah et al. (1993) | Labor with epidural infusion: (1) bupivacaine 0.25%; (2) bupivacaine 0.25% + fentanyl 50 or 100 µg, or diamorphine 2.5 or 5 mg | Epidural opioids: delay | |
Levy et al. (1994) | Nonpregnant controls, 8–12 weeks’ gestation | Delay | |
Stanley et al. (1995) | Second and third trimesters and 8 weeks postpartum | No delay | |
Zimmermann et al. (1996) | Labor with epidural infusion: (1) bupivacaine 0.125%; (2) bupivacaine 0.125% + fentanyl 2 µg/mL | No delay | |
Porter et al. (1997) | Labor with epidural infusion: (1) bupivacaine 0.125%; (2) bupivacaine 0.125% + fentanyl 2.5 µg/mL | Epidural fentanyl total: < 100 µg: no delay > 100 µg: delay | |
Kelly et al. (1997) | Labor with neuraxial bolus: (1) epidural bupivacaine 0.375%; (2) epidural bupivacaine 0.25% + fentanyl 50 µg; (3) intrathecal bupivacaine 2.5 mg + fentanyl 25 µg | Epidural fentanyl: no delay Intrathecal fentanyl: delay | |
Real-time ultrasonography | Carp et al. (1992) | Nonpregnant controls, third trimester | No delay |
Chiloiro et al. (2001) | Serial study in 11 women: first and third trimesters, 4–6 months postpartum | No delay | |
Real-time ultrasonography and acetaminophen absorption | Wong et al. (2002) | Third trimester Cross-over study | 50-mL or 300-mL water: no delay Faster gastric emptying with 300 mL |
Wong et al. (2007) | 10 obese parturients Third trimester Cross-over study | 50-mL or 300-mL water: no delay |