Goldstein NA, Armfield DR, Kingsley LA, Borland LM, Allen GC, Post JC. Postoperative Complications After Tonsillectomy and Adenoidectomy in Children With Down Syndrome. Arch Otolaryngol Head Neck Surg. 1998;124(2):171-176. doi:10.1001/archotol.124.2.171
To compare the postoperative course and complications after tonsillectomy or tonsillectomy and adenoidectomy in children with Down syndrome (group 1) with the postoperative course and complications in children in a control group (group 2).
Retrospective review of medical records for the period January 1, 1986, through March 30, 1996.
Tertiary care children's hospital.
The study included 87 children in group 1 and 64 children in group 2 matched for age, sex, and year of surgery.
Tonsillectomy and adenoidectomy(group 1, 79 children; group 2, 57 children) and tonsillectomy (group 1, 8 children; group 2, 7 children).
Main Outcome Measures
Length of hospitalization and postoperative complications.
The length of hospitalization was significantly increased for the children in group 1 compared with that of children in group 2 (1.6 vs 0.80 days; P=.001, Mann-Whitney U test). Twenty-two children (25%) in group 1 required airway management or observation in the pediatric intensive care unit compared with no children in group 2 who required such care (P<.001, Fisher exact test). None of the children in either group required reintubation, continuous positive airway pressure, or tracheotomy. Respiratory complications requiring intervention were 5 times more likely in group 1 (22  vs 3 ; P<.001, Fisher exact test). The median time until intake of clear liquids and duration of intravenous therapy were significantly increased in group 1 compared with group 2 (5.0 vs 4.0 hours, P=.03; 23.5 vs 16.0 hours, P=.001, respectively; Mann-Whitney U test).
Although tonsillectomy and adenoidectomy can be performed safely in children with Down syndrome, the rate of postoperative respiratory complications is higher and the duration until adequate oral intake is resumed is longer. We therefore recommend that children with Down syndrome be admitted to the hospital overnight after undergoing tonsillectomy and adenoidectomy.
DOWN SYNDROME occurs in 1 in 800 live births and is the most common autosomal chromosomal disorder causing mental retardation.1 Most children with Down syndrome have trisomy 21 (95%), while 3% to 4% have an unbalanced translocation for all or part of chromosome 21. Down syndrome is characterized by aberrant craniofacial features, including microbrachycephaly, flat occiput, short neck, oblique palpebral fissures, epicanthal folds, Brushfield spots, flat nasal dorsum, small low-set auricles, stenotic ear canals, prominent furrowed tongue, and microdontia with fused teeth. Associated anomalies include congenital cardiac disease, gastrointestinal disease, hypotonia, delayed growth, developmental delays, hearing loss, cervical spine disorders, thyroid disease, and obesity.
Upper-airway obstruction and obstructive sleep apnea syndrome (OSAS) are increasingly recognized problems in children with Down syndrome.2,3 In fact, OSAS may contribute to the noncardiac pulmonary hypertension that occurs in these children.4,5 Predisposing factors for OSAS in Down syndrome include midfacial hypoplasia; micrognathia;6 narrow nasopharynx; small oral cavity; macroglossia; relative tonsil and adenoid hyperplasia; increased secretions; hypotonia of the palatal, lingual, and pharyngeal muscles; laryngotracheal abnormalities; and obesity.1,7 There is an increased incidence of chronic rhinosinusitis in children with Down syndrome. Recurrent and chronic tonsillitis also affect these children, as they do other children.
Tonsillectomy and adenoidectomy (T&A) may be required in children with Down syndrome for treatment of upper-airway obstruction, OSAS, recurrent or chronic tonsillitis, recurrent peritonsillar abscesses, dentofacial abnormalities, and, rarely, for malignant neoplasms, spontaneous tonsil hemorrhage, and refractory halitosis. Because of their multiple underlying medical problems, especially cardiac disease, increased incidence of OSAS, and developmental delays, we suspected that children with Down syndrome would have an increased rate of respiratory complications and a more protracted postoperative course. With increasing economic pressure to perform T&As on an ambulatory basis, high-risk patients who are inappropriate candidates for ambulatory surgery must be ascertained. Our aim was to perform a review of the postoperative course and complications after tonsillectomy or T&A in children with Down syndrome to determine whether these children are at high risk for postoperative complications.
The medical records of 87 children with Down syndrome (group 1) who underwent T&A or tonsillectomy at Children's Hospital of Pittsburgh, Pittsburgh, Pa, between January 1, 1986, and March 30, 1996, were identified by International Classification of Diseases, Ninth Revision, Clinical Modification8 codes. Of the 87, 64 were randomly matched for age, sex, and year of surgery with control patients (group 2) who did not have Down syndrome, underwent T&A or tonsillectomy, and were assigned an American Society of Anesthesiologists Physical Status of 1 by the anesthesia service. All medical records were reviewed to determine the admitting diagnosis, more remote and antecedent medical history, physical examination findings, results of ancillary studies, operative procedure performed, anesthetic complications, length of hospitalization, respiratory complications, required medical interventions, time after the operation until resumption of adequate oral intake, duration of intravenous therapy, duration of postoperative emesis, occurrence of postoperative bleeding, and the need for readmission. Before 1991, children were routinely admitted to the hospital for 1 night after T&A. By matching group 1 and group 2 by the year of surgery, we accounted for changes in established practices. All comparisons of dichotomous data were made with the Fisher exact test, which was especially appropriate because of the very low cell size for many of the 2×2 tables. Continuous variables were compared between groups by using the Mann-Whitney U test.
The demographics of our patient population are given in Table 1. There were no significant differences in age, sex, or race in groups 1 and 2. Clinically significant medical history data for the children are given in Table 2. Forty-two children in group 1 (48%) had congenital cardiac disease, with 28 (32%) requiring cardiac surgery compared with only 1 child in group 2 (2%) with congenital cardiac disease who did not require cardiac surgery. No child had clinical pulmonary hypertension or congestive heart failure at the time of adenotonsillar surgery, and all were considered, by their cardiologists or pediatricians, to be healthy enough to undergo surgery. Gastrointestinal disease, hypothyroidism, developmental delay, growth retardation, and hypotonia were, as expected, relatively common in the children in group 1. Only 3 children in group 1 had atlantoaxial instability. Other otolaryngologic manifestations frequently found in children with Down syndrome, including chronic otitis media, rhinosinusitis, recurrent croup, subglottic stenosis, and recurrent pneumonia, were also found in children in group 1. Two children in group 1 had tracheotomies in place; of these, 1 had a history of right vocal cord paralysis and subglottic stenosis and 1 had a history of choanal atresia and subglottic stenosis.
Of the 151 children in both groups, 136 (90.0%) had a tonsillectomy and a full or upper-half adenoidectomy (Table 3). An upper-half adenoidectomy was performed if the child had a submucous cleft palate or bifid uvula. A tonsillectomy alone was performed in children with recurrent tonsillitis without obstructive symptoms or in children who had previously undergone adenoidectomy. Upper-airway obstruction and OSAS were the indications for surgery in 82 children (94%) in group 1 and 36 children (56%) in group 2.
Of the children in group 1, 80 (92%) had a history of snoring, and 46 (53%) had a history of nighttime breathing pauses. In group 2, 52 children (81%) also had a history of snoring, and 15 (23%) had a history of nighttime breathing pauses. Percentile weights for age at the time of surgery were comparable between the 2 groups except that 18 children (21%) in group 1, but no children in group 2, weighed less than the 5th percentile for age. Preoperative nocturnal polysomnograms (NPSGs) were performed in 7 children (8%) in group 1 and only 1 child (2%) in group 2. In group 1, 5 studies demonstrated obstructive sleep apnea, the results of 1 study were normal, and the results of 1 study, which had been performed at an outside institution, were not recorded. The NPSG from the 1 child in group 2 demonstrated obstructive sleep apnea.
Rigid bronchoscopy was performed in 6 children (7%) in group 1 (Table 3); of these 6 children, 4 required interval bronchoscopy (2 had tracheotomies, 1 had a history of subglottic stenosis, and 1 had a history of tracheal stenosis), 1 had a history of recurrent croup, and 1 had OSAS and suspected multilevel airway obstruction that required complete airway evaluation. One child in group 1 with suspected gastroesophageal reflux disease underwent bronchoscopy and esophagoscopy at the time of the T&A. Tonsillectomies were performed by electrocautery dissection in 143 children (94.7%), while blunt and sharp dissection were used in the remainder. Adenoidectomy was performed by curettage. The anesthetic technique was variable, but children received a combination of inhalational agents, opioid or barbiturate narcotics, muscle relaxants, anticholinergics, antiemetics, and benzodiazepines. Antibiotic prophylaxis for subacute bacterial endocarditis was administered to all children with congenital cardiac disease. This study included medical records for operations performed before the routine use of intraoperative corticosteroids to prevent oropharyngeal edema and shorten the time until oral intake is resumed.
Data about the treatment and hospital course for the children studied are given in Table 4. There were no significant differences in operative or anesthesia time between groups. The estimated amount of intraoperative blood loss was significantly higher in group 2.
Anesthetic complications occurred in 7 children (8%) in group 1 compared with none in group 2 (P=.02). The complications were all respiratory, and the most common complication was postextubation stridor. The indication for surgery for all 7 children with anesthetic complications was OSAS or upper-airway obstruction. None had preoperative cardiac or pulmonary disease, and none had a history of recurrent croup or subglottic stenosis. Three of the 4 children with postextubation stridor were intubated with an endotracheal tube of the appropriate size for age as calculated by a standard formula,9 while the endotracheal tube size used for 1 child was not recorded in the anesthetic record. An endotracheal tube of the appropriate size should allow some leaking around the tube at 20 to 25 cm of water of airway pressure. Results of a gas leak test were recorded in only 1 anesthetic record. One child with postextubation stridor had undergone bronchoscopy and esophagoscopy for evaluation of gastroesophageal reflux at the time of the T&A. All anesthetic complications were successfully managed with supplemental oxygen, positioning of the child, positive pressure ventilation, nebulized racemic epinephrine, intravenous corticosteroids, or naloxone hydrochloride. Two of the 7 children were admitted to the pediatric intensive care unit (PICU) for 1 night after surgery, and all but 1 were hospitalized for 1 or 2 nights. The condition of 1 child with postextubation stridor responded to nebulized racemic epinephrine, and the child was discharged from the same-day surgery unit.
The length of hospitalization was significantly longer in group 1 than in group 2. A significantly higher number of the children in group 1 (25%) required either airway management or observation in the PICU compared with the children in group 2. The mean±SD length of stay in the PICU was 1.4±1.1 days. Of the 22 children admitted to the PICU, 11 (50%) were electively admitted for observation, while 11 (50%) required admission because of intraoperative or postoperative complications. Significantly more children in group 2 than in group 1 were discharged from the same-day surgery unit.
Recovery from incisional pain was longer in group 1 than in group 2. The median duration of intravenous therapy, median time until intake of clear liquids, and median time until out of bed were all significantly longer in group 1 than in group 2. The mean duration of intravenous morphine sulfate administration and mean duration of postoperative emesis were also longer in group 1, but the differences were not statistically significant. Although there was an increased rate of readmission for the treatment of dehydration in children in group 1, the difference was not statistically significant. No significant difference was found in the rate of postoperative bleeding between the 2 groups.
The occurrence of the postoperative respiratory complications of upper-airway obstruction and arterial oxygen desaturation to less than 90% was significantly more frequent in group 1 than in group 2(Table 5). The mean±SD number of oxygen desaturations to less than 90% per child was 4.1±6.6 (range, 1.0-30.0; P <.001). Oxygen desaturation to less than 70%, hypoventilation, and bradycardia occurred infrequently in both groups. Respiratory complications requiring intervention occurred significantly more often in group 1 than in group 2. The most frequent interventions were supplemental oxygen therapy, positioning of the child, the insertion of a nasal airway, treatment with nebulized racemic epinephrine, and administration of intravenous corticosteroids. In group 1, the mean±SD duration for various interventions was as follows: supplemental oxygen, 25.7±35.7 hours (range, 1.0-120.0 hours); positioning, 9.7±10.7 hours (range, 1.0-43.0 hours); and nasal airway placement, 10.4±17.1 hours (range, 1.0-48.0 hours). One child in group 1 remained intubated for 19.0 hours after surgery, but this elective intubation was planned preoperatively. No child in either group required reintubation, continuous positive airway pressure, or tracheotomy during the postoperative period. Of the children in group 1 with respiratory complications, mean±SD hospital stay was 2.4±2.7 days (range, 0.0-10.0 days), with 11 children requiring admission to the PICU for a mean±SD of 1.8±1.5 days (range, 1.0-5.0 days). Only 1 child in group 1 required continued intervention once discharged from the hospital. Because of persistent obstruction and arterial oxygen desaturation documented by NPSG 6 days after T&A, a 6-year-old girl required home nighttime supplemental oxygen therapy for 1 month until her symptoms resolved.
A total of 24 children in group 1 required intervention for anesthetic or postoperative respiratory complications because 5 of these children exhibited both types of complications. These children did not constitute a younger group; the mean±SD age was 5.8±3.5 years (range, 2.0-14.7 years). No significant differences were found in sex, race, admitting diagnosis, surgical procedure, medical history, obstructive symptoms, or results of preoperative studies between these 24 children and the entire group 1.
Our results demonstrated that children in group 1 had a significantly higher rate of postoperative complications compared with children in group 2 after T&A or tonsillectomy. The respiratory complications of acute upper-airway obstruction, arterial oxygen desaturation, and postextubation stridor occurred frequently in children in group 1. Twenty-two (25%) children in group 1 were admitted to the PICU, half electively for observation and half because of perioperative respiratory complications, while no children in group 2 required admission to the PICU. The children in group 1 experienced significant delay in resuming postoperative oral intake. The need for care in the PICU and the delay in resuming oral intake significantly lengthened the hospital stay for children in group 1.
Children with Down syndrome have multiple predisposing factors for the development of complications after tonsil and adenoid surgery. The high incidence of OSAS in the children with Down syndrome places the children at a higher risk for postoperative respiratory compromise. McColley et al10 reported that in 23% of children with polysomnographically proved OSAS, severe respiratory compromise developed after T&A and required intervention; the respiratory compromise involved intermittent or continuous oxygen saturation of 70% or less, hypercapnia, or both. Rosen et al11 reported that in 27% of children with OSAS documented by NPSG, postoperative respiratory compromise developed after tonsillectomy, adenoidectomy, or both, with symptoms ranging from oxygen desaturation to less than 80% to respiratory failure. At our institution, children with obstructive symptoms do not routinely have a preoperative NPSG. The prevalence of OSAS in our study groups is unknown because only 7 children in group 1 and 1 in group 2 underwent an NPSG. Yet, according to parental reports, more than 90% of the children in group 1 snored and more than 50% had a history of nighttime breathing pauses.
The midfacial and mandibular hypoplasia, narrow nasopharynx, macroglossia, increased secretions, and hypotonia associated with Down syndrome contribute to upper-airway obstruction, especially after the induction of anesthesia. Intubation and airway management may be difficult because of the craniofacial abnormalities and associated laryngotracheal abnormalities. The prevention of hyperextension and hyperflexion during laryngoscopy, intubation, and surgical procedures is important because of the high incidence of atlantoaxial instability.12 Children with Down syndrome often have elevated pulmonary vascular resistance in response to cardiac disease or hypoxemia secondary to upper-airway obstruction. Abnormal central nervous system control of breathing has been demonstrated in children with Down syndrome and contributes to the occurrence of sleep apnea.4 As demonstrated by our study, children with Down syndrome have a high incidence of lower and upper respiratory infections, which contribute to perioperative respiratory compromise.
Children with Down syndrome have a higher incidence of laryngotracheal stenosis than the general population. Jacobs et al13 found laryngeal or tracheal stenosis in 2% of 518 children with Down syndrome. Miller et al14 found that children with Down syndrome represented a high proportion (4%) of the children undergoing laryngotracheal reconstructions at their institution. Sherry15 reported that postextubation stridor developed in 38% of the children with Down syndrome after cardiac surgery, and de Jong et al16 reported that postextubation stridor developed in 33% after cardiac surgery, with a 6% incidence of subglottic stenosis. In the present study, a 5% incidence of postextubation stridor was found after adenotonsillar surgery, even though endotracheal tubes of the appropriate size were used. Care must be taken to use an endotracheal tube of the appropriate size or smaller than would be expected for age in children with Down syndrome because these children are often smaller than expected for age, their airways may be smaller, and they may have unsuspected laryngotracheal stenosis. The proper size should be confirmed by a gas leak test.
Previous studies have documented postoperative respiratory difficulty in children with Down syndrome. Beilin et al17 reported that in 9 (14%) of 63 children with Down syndrome undergoing facial reconstruction, postoperative respiratory compromise developed; 5 required insertion of a nasal airway to maintain airway patency, and 3 with stridor and 1 with subglottic stenosis required reintubation. Kobel et al18 reported that 13 (13%) of 100 children with Down syndrome undergoing general anesthesia for a variety of procedures had postoperative respiratory compromise; of these children, 10 had difficulty maintaining an airway, 2 had postextubation stridor, and 1 had marked respiratory depression.
In addition to postoperative respiratory complications, the children with Down syndrome in our study had a significantly increased time until the resumption of oral intake. Delayed postoperative oral intake, in addition to dehydration due to preoperative fasting and operative blood loss, could contribute to perioperative discomfort and morbidity, especially in small children with limited fluid reserves. Dehydration in the rare case is also life threatening. Delays in the resumption of adequate oral intake have been reported in very young children undergoing T&A.19,20 Children with Down syndrome who are prematurely discharged from the hospital after T&A, before the resumption of adequate oral intake, are at risk for dehydration.
Several reviews of postoperative complications after T&A have documented an increased rate in subsets of children with additional medical conditions. Richmond et al21 reported that of 784 children undergoing tonsillectomy, adenoidectomy, or both, 26 had Down syndrome, other types of congenital disorders, cerebral palsy, seizures, or mental retardation. Of the 26 children, 7 (27%) had a major respiratory complication involving admission to the PICU, insertion of a nasal airway, or reintubation compared with 10 (1.3%) of the entire group. Tom et al20 found an increased incidence of postoperative airway obstruction in 40 children with Down syndrome, other craniofacial anomalies, cardiac disease, failure to thrive, or pulmonary disease. Gerber et al22 reported that the risk for respiratory compromise involving oxygen desaturation to less than 90%, an obstructive breathing pattern, or respiratory distress that occurred more than 2 hours after surgery and required intervention was 8 times greater in 9 children with chromosomal anomalies and nearly 5 times greater in 9 children with neuromuscular disorders.
Our review of the medical records of 87 children with Down syndrome and 64 children in a control group demonstrated that after T&A, children with Down syndrome are at significantly increased risk for postoperative respiratory complications and delayed resumption of oral intake. Although T&A can be performed safely in children with Down syndrome, we recommend inpatient hospitalization with overnight measurement of pulse oximetry and intravenous hydration until the resumption of adequate postoperative oral intake. Observation in an intensive care setting with capabilities for cardiopulmonary resuscitation and intubation should also be considered for children with Down syndrome and suspected OSAS, because these children are at risk for acute upper-airway obstruction, arterial oxygen desaturation, and respiratory failure.
Accepted for publication August 1, 1997.
Presented at the annual meeting of the American Society of Pediatric Otolaryngology, Scottsdale, Ariz, May 15, 1997.
We thank Susan Strelinski for assistance with data entry.
Reprints: Nira A. Goldstein, MD, Department of Pediatric Otolaryngology, Children's Hospital of Pittsburgh, 3705 Fifth Ave, Pittsburgh, PA 15213.