Pulmonary aspiration of gastric contents during the perioperative period in infants and children may be associated with postoperative mortality or pulmonary morbidity. There has not been a recent determination of the frequency of this event and its outcomes in infants and children.
The authors prospectively identified all cases of pulmonary aspiration of gastric contents during the perioperative courses of 56,138 consecutive patients younger than 18 yr of age who underwent 63,180 general anesthetics for procedures performed in all surgical specialties from July 1985 through June 1997 at the Mayo Clinic.
Pulmonary aspiration occurred in 24 patients (1: 2,632 anesthetics; 0.04%). Children undergoing emergency procedures had a greater frequency of pulmonary aspiration compared to those undergoing elective procedures (1:373 vs. 1:4,544, P < 0.001). Fifteen of the 24 children who aspirated gastric contents did not develop respiratory symptoms within 2 h of aspiration, and none of these 15 developed pulmonary sequelae. Five of these nine children who aspirated and in whom respiratory symptoms developed within 2 h subsequently had pulmonary complications treated with respiratory support (P < 0.003). Three children were treated with mechanical ventilation for more than 48 h, but no child died of sequelae of pulmonary aspiration.
In this study population, the frequency of perioperative pulmonary aspiration in children was quite low. Serious respiratory morbidity was rare, and there were no associated deaths. Infants and children with clinically apparent pulmonary aspiration in whom symptoms did not develop within 2 h did not have respiratory sequelae.
PERIOPERATIVE pulmonary aspiration is an uncommon event in adults. [1,2]Olsson and colleagues found the frequency of perioperative pulmonary aspiration in adults who received care at the Karolinska Hospital in Stockholm to be approximately 1:2,600 during the period 1967-1983. Subsequently, Warner et al. reported the frequency in adults undergoing procedures at the Mayo Clinic from July 1985 to June 1991 to be 1:3,200. Several studies suggest that this problem occurs more frequently in infants and children. [1,3]Unfortunately, these retrospective studies of younger patients have not rigidly defined pulmonary aspiration, and this lack of a rigid definition may cause inaccuracies in calculating the frequency and determining outcomes of perioperative pulmonary aspiration in infants and children.
We evaluated the perioperative courses of infants, children, and adolescents who underwent 63,180 consecutive elective or emergency general anesthetics at a single institution during a 12-yr period. A uniform definition of pulmonary aspiration was used. The purpose of this study was to determine the frequency and clinical significance of pulmonary aspiration in these age groups.
From July 1985 through June 1997, 63,180 general anesthetic procedures were performed at the Mayo Clinic on 56,138 patients who were younger than 18 yr of age. These patients underwent 59,074 elective and 4,106 emergency surgical or diagnostic procedures that were performed in all surgical specialties. With institutional approval, we identified all cases of perioperative pulmonary aspiration in this pediatric population. Pulmonary aspiration was defined as occurring anytime from entry into the operating room until 2 h after the discontinuation of anesthesia. It was identified by either the presence of bilious secretions or particulate matter in the tracheobronchial tree or, in children who did not have their tracheobronchial airways directly examined after regurgitation, by postoperative chest roentgenogram with infiltrates not identified by preoperative roentgenography or physical examination. Cases that did not fulfill these criteria were excluded. The 2 h period was chosen based on our past epidemiologic survey of perioperative pulmonary aspiration in adults. 
All anesthesia providers and recovery room personnel recorded cases that fulfilled these criteria with entry into an extensive institutional perioperative database. The investigators reviewed this database monthly for the entry “pulmonary aspiration.” Based on previous experiences with this database, [2,4,5]this 1-month period was chosen to gain a maximum yield within a time period that provided accurate recall by personnel involved in these cases. Personnel who recorded episodes of pulmonary aspiration were individually interviewed to confirm that the episode fulfilled these criteria for perioperative pulmonary aspiration. They also were asked guided questions about specifics of each occurrence. Two providers were interviewed for each occurrence. The first provider interviewed was the resident, the nurse anesthetist, or the student nurse anesthetist who was directly providing the patient's care. The second provider interviewed was the anesthesiologist who was supervising the anesthetic care. Both had full access to all medical records associated with the patient.
General patient conditions previously reported to be associated with increased frequency of pulmonary aspiration were noted during the interviews. These included age, gender, and presence of congenital and acquired cardiac, gastroenterologic, renal, neurologic, metabolic, and pulmonary diseases and malformations. [6-9]The absence or presence of congenital or acquired diseases and malformations and their severity were rated using the ASA physical status classification scale. Specific conditions that have been reported to predispose to pulmonary aspiration also were identified. These included pregnancy, oral intake of solid foods within 6 h of anesthesia, opioid administration within 4 h of anesthesia, gastrointestinal obstruction, depressed consciousness, acute physical injury, previous esophageal surgery or dysfunction, and lack of coordination of swallowing and respiration. 
Specific questions related to the perioperative period were asked during the interviews. The timing of aspiration was identified. Providers described the characteristics and measurement or estimate of volume and pH of aspirated material, when possible. The choice, performance, and reasons for specific airway management techniques at induction of anesthesia, during intraoperative care, and at the discontinuation of anesthesia were noted.
During the 2 h after discontinuation of anesthesia, anesthesia, recovery room, and postoperative nursing personnel noted the presence and timing of any new cough or wheeze. They also recorded any arterial hemoglobin desaturation while the patient was breathing room air when the measured pulse oximetry (SpO(2)) was < 90% less than the preoperative value or the alveolar-arterial oxygen tension was > 300 mmHg in patients in whom the trachea remained intubated.
The investigators reviewed all abnormal chest roentgenography results performed within 24 h of completion of the anesthetic in patients who were noted in the database as having pulmonary aspiration. The findings on the roentgenograms were compared to any preoperative films and to the recorded preoperative physical examination. Other data related to the extended postoperative period (> 2 h after conclusion of the surgery) included disposition, use of intensive care services, ventilatory support, and pulmonary outcomes. These data were collected by individual medical record review. The Mayo Clinic uses a unit medical record system, and the complete history of every patient, including outpatient and inpatient data, is available for review. 
A comparison of observed frequency of pulmonary aspiration in elective and emergency cases was performed using Fisher's exact test. This same test was used to compare the number of children who aspirated but in whom respiratory symptoms did not develop within 2 h of the discontinuation of anesthesia to the number of children who aspirated and in whom respiratory symptoms developed during this time period. The effect of age and ASA physical status as a continuous variable on the frequency of pulmonary aspiration was assessed using logistic regression. P < 0.05 was considered significant for all comparisons.
There were 56,138 infants, children, and adolescents younger than 18 yr of age who underwent 63,180 general anesthetics. Of these, 39,172 patients (62%) were male. More than one third of the procedures were performed on children 5 yr of age or younger, and 5,746 (8.7%) were younger than 1 yr of age. Anesthesia for elective procedures comprised 59,074 (93.5%) of all cases. There were 4,106 emergency cases.
Frequency and Timing of Pulmonary Aspiration
The overall frequency of perioperative pulmonary aspiration was 1:2,632 anesthetics (0.04%). Pulmonary aspiration was identified by the presence of bilious secretions in the tracheobronchial airways of all 24 patients. No additional patients were identified by secondary database queries. Emergency procedures were associated with a greater frequency of aspiration than were elective procedures (1:373 vs. 1:4,544 anesthetics; P < 0.001). There was no significant difference in frequency of pulmonary aspiration across the different ages (P = 0.38) or ASA physical status classes (P = 0.14). Seven of 55,616 (1:7,945) ASA physical class I and II children undergoing elective procedures aspirated.
Pulmonary aspiration occurred at various times in the perioperative period, although 21 of 24 instances of aspiration occurred at the start of the case (Table 1). In two instances, the patients vomited immediately before induction of anesthesia, but their cases were continued because of the emergency nature of their surgical processes. The majority of the 24 aspirations occurred during induction of anesthesia, but either before pre-planned laryngoscopy and tracheal intubation (12) or during laryngoscopy (7).
Outcomes of Pulmonary Aspiration
All 24 patients who aspirated survived at least 1 month after their procedures, and none died within 1 yr from sequelae of their perioperative pulmonary aspiration. In 15 of these patients (63%), signs or symptoms associated with pulmonary aspiration did not develop within 2 h of the aspiration or discontinuation of the anesthetic. None of these 15 patients required intensive care or respiratory support, nor did pulmonary complications develop. Seven of these 15 patients were originally scheduled to undergo procedures on an outpatient basis, and 6 were discharged from the hospital on the day of their procedures. Postoperative surveillance of at least 1 week for each of these outpatients did not identify any respiratory problems.
Of nine patients who did develop signs or symptoms associated with pulmonary aspiration within 2 h of aspiration or discontinuation of the anesthetic, one or more changes of a new cough or wheeze developed in six, a decrease in SpO(2) less than 90% of the preoperative value while the patient was breathing room air occurred in three, and radiological evidence of pulmonary aspiration developed in two. Although three of these patients were originally scheduled undergo procedures on an outpatient basis, none were discharged on the day of their procedures. Five of the nine symptomatic patients required intensive care and respiratory support. In contrast, none of the 15 patients who remained asymptomatic after their procedures required respiratory support (P < 0.003).
Three patients required postoperative mechanical support for more than 48 h. A 6-month-old boy with multiple gastrointestinal malformations aspirated during laryngoscopy and wheezing and radiological changes developed immediately after the aspiration. He required ventilatory support for 18 days and supplemental oxygen for an additional 3 weeks. A 2-yr-old boy with a mesenteric tear and bowel ischemia aspirated during a rapid-sequence induction and immediately an SpO(2) less than 85% developed, despite full ventilatory support. Escherichia coli pneumonia eventually developed and he required mechanical ventilation for 14 days. A 16-yr-old girl with a high blood alcohol level, recent ingestion of solids, and severe head trauma aspirated during laryngoscopy. Respiratory distress syndrome developed that necessitated ventilatory support for 33 days.
Clinically important pulmonary aspiration of gastric contents in healthy children younger than 18 yr of age during the immediate perioperative period is uncommon. In 55,616 elective procedures and general anesthetics performed on ASA physical status I and II infants, children, and adolescents, only 1 in 8,000 patients exhibited any clinical evidence of pulmonary aspiration. This rate is nearly identical to the rate of perioperative pulmonary aspiration in healthy adults who underwent elective procedures at our institution during a similar period. None of these aspirations in healthy infants and children resulted in serious pulmonary complications. Even in young patients with congenital and acquired diseases and malformations previously described to predispose them to an increased risk of pulmonary aspiration, aspiration occurred infrequently and morbidity was low. More importantly, none of these young patients with documented pulmonary aspiration died of pulmonary complications.
No pulmonary sequelae developed in any of our patients with a documented pulmonary aspiration and who remained asymptomatic for cough or wheeze and did not have new hypoxia while breathing room air or radiological abnormalities within 2 h of aspiration or discontinuation of anesthesia. In contrast, 55% of our patients with any of these symptoms within 2 h required supplemental oxygen to maintain an SpO(2) of more than 90% or ventilatory support, or both. Strikingly similar results were found previously in our study of the frequency and outcomes of perioperative pulmonary aspiration in adults. Based on these findings and our previous results, we developed a standard approach for managing the disposition of children and adults in whom pulmonary aspiration has occurred. We send ambulatory patients home and inpatients to a regular nursing unit if they remain asymptomatic after 2 h of observation. Symptomatic patients who do not require ventilatory support within 2 h of aspiration are sent to a regular nursing unit and closely observed if they have maintained an SpO(2) of more than 90% while breathing room air or supplemental oxygen for 2 h after aspiration. All others are admitted for observation or therapy to an intensive care area.
The majority of our infants and children who aspirated during the immediate perioperative period had a bowel obstruction or ileus. The proportion of our patients who aspirated and had a bowel obstruction or ileus was especially high in children younger than 3 yr of age. Ten of 11 infants and toddlers (91%) who aspirated had one of these problems compared to only 2 of 13 children older than 3 yr of age (15%). There were 381 children 3 years of age or younger and 348 children older than 3 yr of age who had bowel obstructions or ileus during this study period. This finding suggests that we were unable to manage the airway and prevent regurgitated gastric contents from being aspirated in infants and toddlers as well as we could in older children.
Why this difference? First, the effectiveness of the lower esophageal sphincter to inhibit gastric regurgitation may be less in young children than in adults. The competence of the lower esophageal sphincter and the motor function of the esophagus in infants are generally not well developed and may not prevent gastroesophageal reflux or vomiting during crying or abdominal straining. [12-14]The functions of the sphincter and esophagus improve with age, and the frequency of gastroesophageal reflux decreases toward a rate seen in adults by the third birthday of most children. [15,16]Second, infants and young children often have distended stomachs and generate high intragastric pressures during gagging or straining. These very young patients swallow moderate amounts of air when they cry or suck on pacifiers. This intragastric air, usually present even in infants who have fasted, and the ability of infants to generate very high intraabdominal pressures during crying and gagging may increase the opportunity to regurgitate gastric contents. Third, we speculate that the ability to effectively apply pressure to the cricoid cartilage during rapid-sequence inductions or when gagging occurs during inhalation or routine intravenous inductions may be more difficult in infants and toddlers than in older children. Although Salem et al. [17,18]repeatedly noted the efficacy of cricoid pressure to prevent gastric insufflation in children, there are no studies that show its ability to decrease the frequency of gastric regurgitation in infants. Fourth, there are wide-ranging opinions about the efficacy of airway management techniques for infants who have a “full stomach” or gastrointestinal disease. For example, in a 1994 survey of the anesthetic management of a theoretical case of a 4-week-old infant undergoing pyloromyotomy, only two thirds of British consultant anesthetists responded that they would use an intravenous induction technique, and less than half would use cricoid pressure during this anesthetic induction. The lack of good outcomes data for any association between pulmonary aspiration and various airway management techniques in infants may lead to nonstandardized use of techniques and contribute to inconsistent outcomes.
Our findings do not support the conclusions of previous studies that suggest that pediatric patients compared to adult patients have an increased risk of perioperative pulmonary aspiration. In a retrospective study of this complication at the Karolinska Hospital during the years 1967 to 1970 and 1975 to 1983, Olsson et al. found the frequency of perioperative pulmonary aspiration in children 0-9 yr of age to be 1:1,163 (0.09%). This rate was approximately 2.5 times greater than the rate they found in adolescents and adults. In contrast, in a large prospective study of anesthesia-related complications in 461 French medical institutions during the years 1978 to 1982, Tiret et al. found that only 1:10,060 children 0-14 yr of age (0.01%) aspirated during or within 24 h of anesthesia. Why this great disparity in findings between these two studies and ours? First, the methodologies of the studies were very different. Second, the definitions of perioperative pulmonary aspiration varied considerably and reporting sensitivities probably differed. Third, our patient populations probably were quite different. For example, the Mayo Clinic pediatric surgical practice has a high proportion of patients who are referred for complex elective procedures. Most emergency procedures are performed on the small proportion of our practice that is composed of local children. Therefore, our rate of performance of emergency procedures that have a higher frequency of perioperative pulmonary aspiration may be lower than it is in the other reported populations.
Nearly all cases of pulmonary aspiration in our study occurred in patients who gagged or coughed during airway manipulation or during induction of anesthesia in which paralysis with muscle relaxants was either not present or was insufficient to prevent a gag or cough. Unlike our experience with adult patients in whom one third of pulmonary aspirations occurred at the end of anesthetic induction, we found only one of every eight aspirations in these children to occur at the discontinuation of anesthesia. Although we do not routinely suction the stomachs of tracheally intubated children during their anesthetics at our institution, we generally extubate the tracheas of children, especially those who are very young, when they are awake. Having our youngest children awake with good gag reflexes before tracheal extubation may account for our low proportion of pulmonary aspirations that occurred after the surgical procedures.
In summary, this study found no serious morbidity from pulmonary aspiration in the immediate perioperative period in more than 55,000 elective procedures and general anesthetics in ASA physical status classification I and II patients younger than 18 yr of age. Most children who aspirated were undergoing emergency procedures, and nearly all children younger than 3 yr of age who aspirated had either a bowel obstruction or ileus preoperatively. Children who have clinically apparent pulmonary aspiration but in whom symptoms do not develop within 2 h of aspiration or discontinuation of anesthesia appear to be unlikely to have respiratory sequelae.