- Email: [email protected]
The effect of polysomnography on pediatric adenotonsillectomy postoperative management
The effect of polysomnography on pediatric adenotonsillectomy postoperative management ANTHONY A. RIEDER,
MD,
and VALERIE FLANARY,
MD,
Milwaukee, Wisconsin
OBJECTIVE: We retrospectively investigated the effect and predictability of preoperative polysomnography (PSG) on the postoperative course of younger pediatric patients undergoing adenotonsillectomy. STUDY DESIGN AND SETTING: A retrospective chart review was performed for patients 3 years of age and younger who had undergone adenotonsillectomy between July 1997 and July 2002 at the Children’s Hospital of Wisconsin. RESULTS: Two hundred eighty-two patients were identified. Forty-three patients had preoperative PSG. No correlation between the severity of PSG results and postoperative course was identified. CONCLUSIONS: The role of PSG in upper airway obstruction and OSA remains controversial. This study suggests that although the complication rate may be higher in this younger population, these complications do not appear to have a large impact on their length of stay. SIGNIFICANCE: This study suggests that the 3-yearsand-younger group, in the absence of other comorbidities, can safely undergo adenotonsillectomy without undergoing preoperative PSG. EBM raing: C. (Otolaryngol Head Neck Surg 2005;132: 263-7.)
S
leep-disordered breathing is a common diagnosis in the pediatric population. Patients generally present with symptoms such as loud snoring, respiratory pauses, and mouth breathing. Physical examination often shows enlarged tonsil or adenoid tissue. In these children, adenotonsillectomy is often curative. In most cases, the decision for surgery is established on the basis of clinical history and physical exam. The gold standard for diagnosis of sleep-disordered breathing is polysomnogFrom the Division of Pediatric Otolaryngology and Department of Otolaryngology and Communication Sciences, Medical College of Wisconsin, Milwaukee, WI. Presented at the Annual Meeting of the Academy of Otolaryngology–Head and Neck Surgery, Orlando, FL. September 22, 2003. Reprint requests: Valerie Flanary, MD, PO Box 1997, 9000 West Wisconsin Avenue, Milwaukee, WI 53201; e-mail, [email protected]. 0194-5998/$30.00 Copyright © 2005 by the American Academy of Otolaryngology–Head and Neck Surgery Foundation, Inc. doi:10.1016/j.otohns.2004.10.007
raphy (PSG). However, PSG is costly, time limiting, and often not readily available.1 Furthermore, its value as a preoperative tool is controversial with regard to which patient groups require or benefit from preoperative PSG.2 Several studies have addressed adenotonsillectomy and have identified populations at high risk for postoperative complications. These populations include children with craniofacial disorders, failure to thrive, neurological impairment, Down syndrome, obstructive sleep apnea, and aged 3 years or younger.3-7 Preoperative PSG and postoperative overnight observation has been recommended for these patients. Polysomnography has also been recommended for differentiating primary snoring from obstructive sleep apnea.8,9 In addition, tonsillectomy is often postponed in children aged 3 or younger because of reported increases in risks. Some authors believe that polysomnographic results are predictive of postoperative complications and postoperative course.3-5,7 This study attempts to retrospectively review the postoperative course in children 3 years of age and younger and to correlate this course with the severity of obstruction by diagnosis on preoperative PSG. We hypothesize that the severity of sleepdisordered breathing or upper airway obstruction (UAO) identified by PSG does not correlate well with the severity of postoperative complications and postoperative course. MATERIALS AND METHODS A search by procedure code identified all children who underwent tonsillectomy or adenotonsillectomy at The Children’s Hospital of Wisconsin between July 1997 and July 2002. This data set was then limited to those children 3 years of age and younger who had procedures performed by attending surgeons or by a resident with attending supervision. The Children’s Hospital Institutional Review Board granted approval for this study. Data were extracted from each patient’s medical chart and included the following: preoperative PSG results, admitting diagnosis, surgeon, medical comorbidities, length of stay (LOS), postoperative complications, and associated interventions. In addition, postoperative progress notes were reviewed to identify the general postoperative course for each patient as well as any information that could be attributed to the medical decision-making process. Patients were grouped on the 263
Otolaryngology– Head and Neck Surgery February 2005
264 RIEDER and FLANARY
Table 1. Age distribution Age (mo) 0-24 25-36 37-47 Total Average age (mo)
Total study group
Sleep study
No sleep study
67 141 74 282 30.2
18 20 5 43 25.9
49 121 69 239 31.1
basis of whether or not they had undergone preoperative PSG. The average LOS was used as an objective indicator for evaluating the postoperative course in subject groups. RESULTS During the 5-year period described, there were 4,057 adenotonsillectomy or tonsillectomy-alone procedures performed. The charts for 305 children aged 3 years or younger were reviewed. Of these, 23 children were excluded from the study for the following reasons: incomplete chart (n ⫽ 8), procedure miscoded (n ⫽ 2), additional procedures performed (n ⫽ 5), or procedure performed during period of acute illness (n ⫽ 8). The remaining 282 sets of patient data were included in the statistical analysis. The average age was 30 months. Table 1 summarizes the age distribution. UAO was identified as the primary indication for surgery in 271 of the patients. Sleep studies, including complete 16-channel (n ⫽ 38), 4-channel (n ⫽ 3), or hardcopy pulse-oximetry (n ⫽ 2), were performed preoperatively in 43 patients. The average age of patients who had a preoperative PSG was 25.9 months, compared with 31.1 months in the non-PSG group (P ⬍ 0.0001). Twenty-seven patients were identified as having obstructive sleep apnea (OSA) with a respiratory disturbance index (RDI) that ranged from 1 to 52. There was no statistical difference in the average LOS for patients with OSA when the subjects were grouped by using their RDI as a measure of OSA severity. The average LOS for patients with OSA was 1.57 days. The discharge criteria for patients typically requires that the patient not require any supplemental oxygen and be able to maintain oxygen saturations of ⬎90% without periods of sustained desaturation attributed to UAO. Upper airway resistance syndrome (UARS) was identified in 11 patients. UARS is abnormally increased upper airway resistance during sleep that leads to increased respiratory effort and sleep fragmentation without classically defined apneic or hypopneic episodes or notable declines in oxygen saturation.10 The average LOS for patients with UARS (2.2 days) was longer when compared with the average
LOS stay (1.57 days) for patients with OSA (P ⫽ 0.0448). The remaining sleep studies without a calculated RDI or UARS were classified as normal, mildly abnormal, or significantly abnormal by the pulmonologist reading the study. The average LOS for each age distribution was also calculated for each age group in both the non-PSG and PSG groups. The average LOS for all comparable age groups was less in the non-PSG group, with the shortest average LOS occurring in the 37- to 47-month age distribution in both the non-PSG and PSG groups. The average LOS for the non-PSG group was 1.06 days, compared with 1.98 days in the PSG group (P ⬍ 0.0001). The age distribution and average LOS for both groups are listed in Table 2. Complications were identified in 62 patients (21.9%). For this study, a complication was defined as any documented desaturation event (regardless of SaO2 nadir) or any other event or complication that required medical intervention or readmission or that prolonged the patient’s length of hospital stay. Comparisons were made between patients without PSG (n ⫽ 239) and those with PSG results (n ⫽ 43). Complications occurred in 43 patients (17.9%) in the non-PSG group and in 19 patients (44.2%) in the PSG group. The most common complication identified was oxygen desaturation in both groups, representing 65.1% of the total complications in the non-PSG group and 57.9% in the PSG group. Overall, desaturation events represented a 13.8% complication rate (n ⫽ 39) for the entire study population. Complications for each group are listed in Table 3. Additionally, Table 4 lists the total number of complications for the 62 patients identified with respect to the age distribution of the complication. The highest complication rate (32.8%) was found in the 24-monthsand-younger age group, with the lowest complication rate (8.1%) occurring in the 37- to 47-month age group. There was no statistical difference in the mean age of the patients who had complications in the PSG group (26.5 months) compared with the case of the non-PSG group (27.2 months), with a P value equal to 0.7279. The average LOS for patients with complications in each group was also calculated and listed in Table 4. The average LOS for patients with complications in the non-PSG group was 1.74 days, compared with 2.63 days in the PSG group. A trend toward a longer average LOS in the PSG group was identified but the difference did not reach statistical significance (P ⫽ 0.0997). Comorbidities and associated medical conditions were present in 88 patients. The most common conditions are listed in Table 5. Fifty-seven (23.8%) of the patients who did not have a preoperative sleep study were found to have comorbidities, compared with 31
Otolaryngology– Head and Neck Surgery Volume 132 Number 2
RIEDER and FLANARY 265
Table 2. Average length of stay Sleep study group Age distribution (mo)
No. of patients
Average length of stay (d)
No. of patients
Average length of stay (d)
P value
18 20 5 43
2.0 2.1 1.4 1.98
49 121 69 239
1.3 1.1 .81 1.06
0.0048 0.0018 0.0126 0.0001
0-24 25-36 37-47 Total
Table 3. Complications
Complication Desaturation event Poor oral intake Pneumonia/atelectasis Seizure Bleeding Death Total no. of patients with complications
No sleep study group
Group with sleep study (n ⴝ 43)
Group without sleep study (n ⴝ 239)
11 (26%) 6 (14%) 0 0 1 (2%) 1 (2%) 19 (44%)
28 (12%) 10 (4%) 3 (1%) 1 (0.4%) 1 (0.4%) 0 43 (18%)
(72%) patients in the group that had undergone a preoperative sleep study. Eleven (25.5%) of 43 patients with complications in the non-PSG had comorbidities, compared with 14 (73.6%) of 19 patients with complications in the PSG group. DISCUSSION Sleep-disordered breathing is common in the pediatric population, with a reported prevalence between 4% and 11%.11,12 UAO secondary to adenotonsillar hypertrophy is the primary cause of this sleep-disordered breathing and has replaced recurrent tonsillitis as the most common indication for adenotonsillectomy in children.13 Adenotonsillectomy is considered the primary surgical treatment option for children with obstructive sleep disorders, having been shown to improve sleep, breathing, PSG results, enuresis, and patient quality of life.14,15 Several studies have compared the clinical diagnosis of OSA with formal PSG.16,17 These studies were completed before the recognition of UARS and before a clear consensus on a PSG definition of pediatric sleep apnea; therefore, the accuracy of making the clinical diagnosis of OSA remains in question when compared with PSG. In clinical practice, the decision to perform an adenotonsillectomy is often made without the benefit of a diagnostic sleep study, and the decision to perform surgery is based on clinical history and exam findings. This is consistent with the findings of our study in which
only 15% of the patients (n ⫽ 43) had a formal sleep study before their surgery. Sleep studies may also include nocturnal or hard copy to evaluate OSA. When used in children suspected of having OSAS, these studies have shown a high predictive value.22 Sleep studies are typically ordered for patients with other comorbidities or for those patients considered to be at an increased risk for surgery when their clinical diagnosis is unclear. In this study, 72% of the patients with sleep studies had other comorbidities or medical conditions, compared with 24% of the patients without sleep studies. PSG has also been advocated as method to predict which children are at risk of postoperative respiratory complications after adenotonsillectomy and to assist with postoperative planning and monitoring. Several studies have looked at the risk factors associated with postoperative complications.3-5,7 All of these studies identify age either younger than 2 years or younger than 3 years as a significant risk factor for postoperative respiratory compromise. Other risk factors identified include craniofacial anomalies, failure to thrive, hypotonia, morbid obesity, apnea or hypopnea index greater than 5 (or 10), and preoperative oxygen saturation nadir. These studies were all performed retrospectively and used different inclusion and exclusion criteria for patients and various definitions and classifications for their complications. The complication rate after adenotonsillectomy in children reported in several series ranges from 0 to 32%, with airway issues or compromise representing 0 to 16% of these complications.6,18-20 Those studies that were limited to younger children (3 years of age and younger) showed a rate of airway complications ranging from 0 to 5%.6,18,19 Our study compares similarly, with an overall complication rate of 22%, but our rate of airway complications, based on our classification, was higher with a rate of 14% (n ⫽ 39). It is difficult to compare the complication rates among studies because of the definitions used. For this study, a complication was defined as any documented desaturation event (regardless of SaO2 nadir) or any other event or complication that required medical intervention or
Otolaryngology– Head and Neck Surgery February 2005
266 RIEDER and FLANARY
Table 4. Age distribution of complications by group Total study group Age distribution (mo)
Group with sleep studies
Group without sleep studies
No. of patients
Fraction of total complications (%)
No. of patients
Average length of stay (d)
No. of patients
Average length of stay (d)
22 34 6 62
22/67 (32.8) 34/141 (24.1) 6/74 (8.1)
7 10 2 19
2.28 3.00 2.00 2.63
15 24 4 43
1.87 1.75 1.25 1.74
0-24 25-36 37-47 Total
Table 5. Comorbidities Comorbidity Asthma/Reactive airway disease Neurologic impairment (Down syndrome, developmental delay, cerebral palsy) Prematurity Craniofacial anomaly or syndrome Neurologic impaired swallow Laryngomalacia Platelet dysfunction Seizure disorder Other (recurrent croup, cardiac anomalies, recurrent pneumonia, reflux)
No. of patients 25 18 10 9 8 6 6 4 10
readmission or prolonged the patient’s length of hospital stay. Some studies have identified minor and major complications depending upon the intervention required,7 whereas other studies have attempted to identify complications based on the degree of airway obstruction or oxygen saturation nadir.4,5 This makes it difficult to compare complication rates among these studies to the results found in this study. Furthermore, the significance of postoperative desaturations is controversial. Theoretically, a decrease in oxygen saturation may increase the postoperative patient risk; however, no formal studies support this conclusion. In this study, we evaluated the severity of sleepdisordered breathing or UAO identified on preoperative PSG to determine how well this correlated with a patient’s postoperative course; this was done by using average LOS as an indicator. There was no statistical difference in the average LOS for patients who were identified to have OSA when the severity of their OSA was compared by using their RDI. However, when comparing the patients with OSA to those with UARS, there was a statistically significant difference in the average LOS for these patients. The average LOS for OSA patients was 1.5 days, compared with 2.2 days in the UARS patients (P ⫽ 0.0424). UARS has generally been considered to be a more mild form of sleepdisordered breathing as compared with OSA; therefore, it is difficult to explain these findings. Additionally, the
patients who underwent preoperative PSG had a longer average LOS (1.98 days) compared with the patients in the non-PSG group (1.06 days). This was statistically significant (P ⬍ 0.0001). This indicates that patients who undergo preoperative sleep studies are likely at an increased risk postoperatively on the basis of factors such as their underlying medical condition or comorbidities. This is consistent with the findings of this study in that the PSG group had a higher percentage of comorbidities (72%) when compared with the non-PSG group (24%). We recognize the limitations of a retrospective review in that it is not always possible to determine the rationale for all of the postoperative decisions that were made for each subject. This also made it difficult to accurately classify complications, because the oxygen saturation nadir and duration for desaturation events was not always recorded and may only be represented by a comment in the medical chart, such as “desats noted”. Second, the surgical technique and adjuvant therapies, such as steroids, anesthetic methods, and postoperative analgesia were not controlled for and are not addressed by this study. Additionally, we did not review the complication rate at this institution for all age groups other than those 3 years of age and younger; instead, we used the reported complication rates from other studies for comparison regarding older age groups. CONCLUSION This study suggests that although the complication rate may be higher in this younger population, the nature of the complications does not appear to have a large impact on their LOS. This study also suggests that the 3-years-and-younger group, in the absence of other comorbidities, can safely undergo adenotonsillectomy without undergoing preoperative PSG, provided that close attention is paid to postoperative clinical findings while patients are being monitored. Furthermore, the cost, time, and expertise necessary for interpretation of the data have all affected the relative utility of PSG, especially in the pediatric population. The average cost of overnight PSG at a qualified institution is between
Otolaryngology– Head and Neck Surgery Volume 132 Number 2
$600 and $2,800.21 Although home studies are valuable in the evaluation of adults, young children most often require overnight stays at centers where the study is being performed. Given this information, it would be impractical and not clinically justified to consider performing PSG on all pediatric patients before adenotonsillectomy. Higher risk as indicated by the increased LOS for certain patient groups with comorbidities does appear to exist, and we would reserve the use of preoperative PSG for these patients and for patients with ambiguous clinical history. The data in this study are inconclusive with regard to which medical conditions or comorbidities, and which immediate postoperative findings, would mandate an overnight inpatient stay. We believe that this question can only be answered with a prospective study that controls for the previously mentioned surgical variables and clearly defines the timing, degree, and duration of respiratory complications and their interventions. Preoperative PSG findings had little influence on the postoperative management of patients in this study. Postoperative management decisions are primarily based on the postoperative clinical findings present for each patient. Patients are managed individually on the basis of their postoperative course or established protocol, regardless of the PSG recommendations or predictions. REFERENCES 1. Leach J, Olson J, Hermann J, et al. Polysomnographic and clinical findings in children with obstructive sleep apnea. Arch Otolaryngol Head Neck Surg 1992;118:741-4. 2. Messner AH. Evaluation of obstructive sleep apnea by polysomnography prior to pediatric adenotonsillectomy. Arch Otolaryngol Head Neck Surg 1999;125:353-6. 3. Biavati MJ, Manning SC, Phillips DL. Predictive factors for respiratory complications after tonsillectomy and adenoidectomy in children. Arch Otolaryngol Head Neck Surg 1997;123:517-21. 4. McColley SA, April MM, Carroll JL, et al. Respiratory compromise after adenotonsillectomy in children with obstructive sleep apnea. Arch Otolaryngol Head Neck Surg 1992;118:940-3. 5. Rosen GM, Muckle RP, Mahowald MW, et al. Postoperative
RIEDER and FLANARY 267
6. 7. 8.
9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20.
21. 22.
respiratory compromise in children with obstructive sleep apnea syndrome: can it be anticipated? Pediatrics 1994;93:784-8. Slovik Y, Tal A, Tarasiuk A, et al. Complications of adenotonsillectomy in children with OSAS younger than 2 years of age. Int J Pediatr Otorhinolaryngol 2003;67:847-51. Wilson K, Lakheeram I, Morielli A, et al. Can assessment for obstructive sleep apnea predict post adenotonsillectomy respiratory complications. Anesthesiology 2002;96:313-22. Section on Pediatric Pulmonology, Subcommittee on Obstructive Sleep Apnea Syndrome. Clinical practice guideline: diagnosis and management of childhood obstructive sleep apnea. Pediatrics 2002;109:704-12. American Thoracic Society. Standards and indications for cardiopulmonary sleep studies in children. Am J Respir Crit Care Med 1996;153:866-78. Guilleminault C, Chowdhuri S. Upper airway resistance syndrome is a distinct syndrome. Am J Respir Crit Care Med 2000;161:1412-6. Messner A, Pelayo R. Pediatric sleep-related breathing disorders. Am J Otolaryngol 2000;21:98-107. Guilleminault C, Pelayo R. Sleep-disordered breathing in children. Ann Med 1998;30:350-6. Ross AT, Kazahaya K, Lawrence WC. Revisiting outpatient tonsillectomy in young children. Otolaryngol Head Neck Surg 2003;128:326-31. Goldstein NA, Fatima M, Campbell TF, et al. Child behavior and quality of life before and after tonsillectomy and adenoidectomy. Arch Otolaryngol Head Neck Surg 2002;128:770-5. Serres LM, Derkay C, Sie K, et al. Impact of adenotonsillectomy on quality of life in children with obstructive sleep disorders. Arch Otolaryngol Head Neck Surg 2002;128:489-96. Goldstein NA, Sculerati N, Walsleben JA, et al. Clinical diagnosis of pediatric obstructive sleep apnea validated by polysomnography. Otolaryngol Head Neck Surg 1994;111:611-7. Wang RC, Elkins TP, Keech D, et al. Accuracy of clinical evaluation in pediatric obstructive sleep apnea. Otolaryngol Head Neck Surg 1998;118:69-73. Berkowitz RG, Zalzal GH. Tonsillectomy in children under 3 years of age. Arch Otolaryngol Head Neck Surg 1990;116: 685-6. Helmus C. Tonsillectomy and adenoidectomy in the one and two-year-old child. Laryngoscope 1979;89:1764-71. Postma DS, Folsom F. The case for an outpatient “approach” for all pediatric tonsillectomies and/or adenoidectomies: a 4-year review of 1419 cases at a community hospital. Otolaryngol Head Neck Surg 2002;127:101-8. Leach J, Olson J, Hermann J, et al. Polysomnographic and clinical findings in children with obstructive sleep apnea. Arch Otolaryngol Head Neck Surg 1992;118:741-4. Brouillette RT, Morielli A, Leimanis A, et al. Nocturnal pulse oximetry as an abbreviated testing modality for pediatric obstructive sleep apnea. Pediatric 2000;105:405-12.
MD,
and VALERIE FLANARY,
MD,
Milwaukee, Wisconsin
OBJECTIVE: We retrospectively investigated the effect and predictability of preoperative polysomnography (PSG) on the postoperative course of younger pediatric patients undergoing adenotonsillectomy. STUDY DESIGN AND SETTING: A retrospective chart review was performed for patients 3 years of age and younger who had undergone adenotonsillectomy between July 1997 and July 2002 at the Children’s Hospital of Wisconsin. RESULTS: Two hundred eighty-two patients were identified. Forty-three patients had preoperative PSG. No correlation between the severity of PSG results and postoperative course was identified. CONCLUSIONS: The role of PSG in upper airway obstruction and OSA remains controversial. This study suggests that although the complication rate may be higher in this younger population, these complications do not appear to have a large impact on their length of stay. SIGNIFICANCE: This study suggests that the 3-yearsand-younger group, in the absence of other comorbidities, can safely undergo adenotonsillectomy without undergoing preoperative PSG. EBM raing: C. (Otolaryngol Head Neck Surg 2005;132: 263-7.)
S
leep-disordered breathing is a common diagnosis in the pediatric population. Patients generally present with symptoms such as loud snoring, respiratory pauses, and mouth breathing. Physical examination often shows enlarged tonsil or adenoid tissue. In these children, adenotonsillectomy is often curative. In most cases, the decision for surgery is established on the basis of clinical history and physical exam. The gold standard for diagnosis of sleep-disordered breathing is polysomnogFrom the Division of Pediatric Otolaryngology and Department of Otolaryngology and Communication Sciences, Medical College of Wisconsin, Milwaukee, WI. Presented at the Annual Meeting of the Academy of Otolaryngology–Head and Neck Surgery, Orlando, FL. September 22, 2003. Reprint requests: Valerie Flanary, MD, PO Box 1997, 9000 West Wisconsin Avenue, Milwaukee, WI 53201; e-mail, [email protected]. 0194-5998/$30.00 Copyright © 2005 by the American Academy of Otolaryngology–Head and Neck Surgery Foundation, Inc. doi:10.1016/j.otohns.2004.10.007
raphy (PSG). However, PSG is costly, time limiting, and often not readily available.1 Furthermore, its value as a preoperative tool is controversial with regard to which patient groups require or benefit from preoperative PSG.2 Several studies have addressed adenotonsillectomy and have identified populations at high risk for postoperative complications. These populations include children with craniofacial disorders, failure to thrive, neurological impairment, Down syndrome, obstructive sleep apnea, and aged 3 years or younger.3-7 Preoperative PSG and postoperative overnight observation has been recommended for these patients. Polysomnography has also been recommended for differentiating primary snoring from obstructive sleep apnea.8,9 In addition, tonsillectomy is often postponed in children aged 3 or younger because of reported increases in risks. Some authors believe that polysomnographic results are predictive of postoperative complications and postoperative course.3-5,7 This study attempts to retrospectively review the postoperative course in children 3 years of age and younger and to correlate this course with the severity of obstruction by diagnosis on preoperative PSG. We hypothesize that the severity of sleepdisordered breathing or upper airway obstruction (UAO) identified by PSG does not correlate well with the severity of postoperative complications and postoperative course. MATERIALS AND METHODS A search by procedure code identified all children who underwent tonsillectomy or adenotonsillectomy at The Children’s Hospital of Wisconsin between July 1997 and July 2002. This data set was then limited to those children 3 years of age and younger who had procedures performed by attending surgeons or by a resident with attending supervision. The Children’s Hospital Institutional Review Board granted approval for this study. Data were extracted from each patient’s medical chart and included the following: preoperative PSG results, admitting diagnosis, surgeon, medical comorbidities, length of stay (LOS), postoperative complications, and associated interventions. In addition, postoperative progress notes were reviewed to identify the general postoperative course for each patient as well as any information that could be attributed to the medical decision-making process. Patients were grouped on the 263
Otolaryngology– Head and Neck Surgery February 2005
264 RIEDER and FLANARY
Table 1. Age distribution Age (mo) 0-24 25-36 37-47 Total Average age (mo)
Total study group
Sleep study
No sleep study
67 141 74 282 30.2
18 20 5 43 25.9
49 121 69 239 31.1
basis of whether or not they had undergone preoperative PSG. The average LOS was used as an objective indicator for evaluating the postoperative course in subject groups. RESULTS During the 5-year period described, there were 4,057 adenotonsillectomy or tonsillectomy-alone procedures performed. The charts for 305 children aged 3 years or younger were reviewed. Of these, 23 children were excluded from the study for the following reasons: incomplete chart (n ⫽ 8), procedure miscoded (n ⫽ 2), additional procedures performed (n ⫽ 5), or procedure performed during period of acute illness (n ⫽ 8). The remaining 282 sets of patient data were included in the statistical analysis. The average age was 30 months. Table 1 summarizes the age distribution. UAO was identified as the primary indication for surgery in 271 of the patients. Sleep studies, including complete 16-channel (n ⫽ 38), 4-channel (n ⫽ 3), or hardcopy pulse-oximetry (n ⫽ 2), were performed preoperatively in 43 patients. The average age of patients who had a preoperative PSG was 25.9 months, compared with 31.1 months in the non-PSG group (P ⬍ 0.0001). Twenty-seven patients were identified as having obstructive sleep apnea (OSA) with a respiratory disturbance index (RDI) that ranged from 1 to 52. There was no statistical difference in the average LOS for patients with OSA when the subjects were grouped by using their RDI as a measure of OSA severity. The average LOS for patients with OSA was 1.57 days. The discharge criteria for patients typically requires that the patient not require any supplemental oxygen and be able to maintain oxygen saturations of ⬎90% without periods of sustained desaturation attributed to UAO. Upper airway resistance syndrome (UARS) was identified in 11 patients. UARS is abnormally increased upper airway resistance during sleep that leads to increased respiratory effort and sleep fragmentation without classically defined apneic or hypopneic episodes or notable declines in oxygen saturation.10 The average LOS for patients with UARS (2.2 days) was longer when compared with the average
LOS stay (1.57 days) for patients with OSA (P ⫽ 0.0448). The remaining sleep studies without a calculated RDI or UARS were classified as normal, mildly abnormal, or significantly abnormal by the pulmonologist reading the study. The average LOS for each age distribution was also calculated for each age group in both the non-PSG and PSG groups. The average LOS for all comparable age groups was less in the non-PSG group, with the shortest average LOS occurring in the 37- to 47-month age distribution in both the non-PSG and PSG groups. The average LOS for the non-PSG group was 1.06 days, compared with 1.98 days in the PSG group (P ⬍ 0.0001). The age distribution and average LOS for both groups are listed in Table 2. Complications were identified in 62 patients (21.9%). For this study, a complication was defined as any documented desaturation event (regardless of SaO2 nadir) or any other event or complication that required medical intervention or readmission or that prolonged the patient’s length of hospital stay. Comparisons were made between patients without PSG (n ⫽ 239) and those with PSG results (n ⫽ 43). Complications occurred in 43 patients (17.9%) in the non-PSG group and in 19 patients (44.2%) in the PSG group. The most common complication identified was oxygen desaturation in both groups, representing 65.1% of the total complications in the non-PSG group and 57.9% in the PSG group. Overall, desaturation events represented a 13.8% complication rate (n ⫽ 39) for the entire study population. Complications for each group are listed in Table 3. Additionally, Table 4 lists the total number of complications for the 62 patients identified with respect to the age distribution of the complication. The highest complication rate (32.8%) was found in the 24-monthsand-younger age group, with the lowest complication rate (8.1%) occurring in the 37- to 47-month age group. There was no statistical difference in the mean age of the patients who had complications in the PSG group (26.5 months) compared with the case of the non-PSG group (27.2 months), with a P value equal to 0.7279. The average LOS for patients with complications in each group was also calculated and listed in Table 4. The average LOS for patients with complications in the non-PSG group was 1.74 days, compared with 2.63 days in the PSG group. A trend toward a longer average LOS in the PSG group was identified but the difference did not reach statistical significance (P ⫽ 0.0997). Comorbidities and associated medical conditions were present in 88 patients. The most common conditions are listed in Table 5. Fifty-seven (23.8%) of the patients who did not have a preoperative sleep study were found to have comorbidities, compared with 31
Otolaryngology– Head and Neck Surgery Volume 132 Number 2
RIEDER and FLANARY 265
Table 2. Average length of stay Sleep study group Age distribution (mo)
No. of patients
Average length of stay (d)
No. of patients
Average length of stay (d)
P value
18 20 5 43
2.0 2.1 1.4 1.98
49 121 69 239
1.3 1.1 .81 1.06
0.0048 0.0018 0.0126 0.0001
0-24 25-36 37-47 Total
Table 3. Complications
Complication Desaturation event Poor oral intake Pneumonia/atelectasis Seizure Bleeding Death Total no. of patients with complications
No sleep study group
Group with sleep study (n ⴝ 43)
Group without sleep study (n ⴝ 239)
11 (26%) 6 (14%) 0 0 1 (2%) 1 (2%) 19 (44%)
28 (12%) 10 (4%) 3 (1%) 1 (0.4%) 1 (0.4%) 0 43 (18%)
(72%) patients in the group that had undergone a preoperative sleep study. Eleven (25.5%) of 43 patients with complications in the non-PSG had comorbidities, compared with 14 (73.6%) of 19 patients with complications in the PSG group. DISCUSSION Sleep-disordered breathing is common in the pediatric population, with a reported prevalence between 4% and 11%.11,12 UAO secondary to adenotonsillar hypertrophy is the primary cause of this sleep-disordered breathing and has replaced recurrent tonsillitis as the most common indication for adenotonsillectomy in children.13 Adenotonsillectomy is considered the primary surgical treatment option for children with obstructive sleep disorders, having been shown to improve sleep, breathing, PSG results, enuresis, and patient quality of life.14,15 Several studies have compared the clinical diagnosis of OSA with formal PSG.16,17 These studies were completed before the recognition of UARS and before a clear consensus on a PSG definition of pediatric sleep apnea; therefore, the accuracy of making the clinical diagnosis of OSA remains in question when compared with PSG. In clinical practice, the decision to perform an adenotonsillectomy is often made without the benefit of a diagnostic sleep study, and the decision to perform surgery is based on clinical history and exam findings. This is consistent with the findings of our study in which
only 15% of the patients (n ⫽ 43) had a formal sleep study before their surgery. Sleep studies may also include nocturnal or hard copy to evaluate OSA. When used in children suspected of having OSAS, these studies have shown a high predictive value.22 Sleep studies are typically ordered for patients with other comorbidities or for those patients considered to be at an increased risk for surgery when their clinical diagnosis is unclear. In this study, 72% of the patients with sleep studies had other comorbidities or medical conditions, compared with 24% of the patients without sleep studies. PSG has also been advocated as method to predict which children are at risk of postoperative respiratory complications after adenotonsillectomy and to assist with postoperative planning and monitoring. Several studies have looked at the risk factors associated with postoperative complications.3-5,7 All of these studies identify age either younger than 2 years or younger than 3 years as a significant risk factor for postoperative respiratory compromise. Other risk factors identified include craniofacial anomalies, failure to thrive, hypotonia, morbid obesity, apnea or hypopnea index greater than 5 (or 10), and preoperative oxygen saturation nadir. These studies were all performed retrospectively and used different inclusion and exclusion criteria for patients and various definitions and classifications for their complications. The complication rate after adenotonsillectomy in children reported in several series ranges from 0 to 32%, with airway issues or compromise representing 0 to 16% of these complications.6,18-20 Those studies that were limited to younger children (3 years of age and younger) showed a rate of airway complications ranging from 0 to 5%.6,18,19 Our study compares similarly, with an overall complication rate of 22%, but our rate of airway complications, based on our classification, was higher with a rate of 14% (n ⫽ 39). It is difficult to compare the complication rates among studies because of the definitions used. For this study, a complication was defined as any documented desaturation event (regardless of SaO2 nadir) or any other event or complication that required medical intervention or
Otolaryngology– Head and Neck Surgery February 2005
266 RIEDER and FLANARY
Table 4. Age distribution of complications by group Total study group Age distribution (mo)
Group with sleep studies
Group without sleep studies
No. of patients
Fraction of total complications (%)
No. of patients
Average length of stay (d)
No. of patients
Average length of stay (d)
22 34 6 62
22/67 (32.8) 34/141 (24.1) 6/74 (8.1)
7 10 2 19
2.28 3.00 2.00 2.63
15 24 4 43
1.87 1.75 1.25 1.74
0-24 25-36 37-47 Total
Table 5. Comorbidities Comorbidity Asthma/Reactive airway disease Neurologic impairment (Down syndrome, developmental delay, cerebral palsy) Prematurity Craniofacial anomaly or syndrome Neurologic impaired swallow Laryngomalacia Platelet dysfunction Seizure disorder Other (recurrent croup, cardiac anomalies, recurrent pneumonia, reflux)
No. of patients 25 18 10 9 8 6 6 4 10
readmission or prolonged the patient’s length of hospital stay. Some studies have identified minor and major complications depending upon the intervention required,7 whereas other studies have attempted to identify complications based on the degree of airway obstruction or oxygen saturation nadir.4,5 This makes it difficult to compare complication rates among these studies to the results found in this study. Furthermore, the significance of postoperative desaturations is controversial. Theoretically, a decrease in oxygen saturation may increase the postoperative patient risk; however, no formal studies support this conclusion. In this study, we evaluated the severity of sleepdisordered breathing or UAO identified on preoperative PSG to determine how well this correlated with a patient’s postoperative course; this was done by using average LOS as an indicator. There was no statistical difference in the average LOS for patients who were identified to have OSA when the severity of their OSA was compared by using their RDI. However, when comparing the patients with OSA to those with UARS, there was a statistically significant difference in the average LOS for these patients. The average LOS for OSA patients was 1.5 days, compared with 2.2 days in the UARS patients (P ⫽ 0.0424). UARS has generally been considered to be a more mild form of sleepdisordered breathing as compared with OSA; therefore, it is difficult to explain these findings. Additionally, the
patients who underwent preoperative PSG had a longer average LOS (1.98 days) compared with the patients in the non-PSG group (1.06 days). This was statistically significant (P ⬍ 0.0001). This indicates that patients who undergo preoperative sleep studies are likely at an increased risk postoperatively on the basis of factors such as their underlying medical condition or comorbidities. This is consistent with the findings of this study in that the PSG group had a higher percentage of comorbidities (72%) when compared with the non-PSG group (24%). We recognize the limitations of a retrospective review in that it is not always possible to determine the rationale for all of the postoperative decisions that were made for each subject. This also made it difficult to accurately classify complications, because the oxygen saturation nadir and duration for desaturation events was not always recorded and may only be represented by a comment in the medical chart, such as “desats noted”. Second, the surgical technique and adjuvant therapies, such as steroids, anesthetic methods, and postoperative analgesia were not controlled for and are not addressed by this study. Additionally, we did not review the complication rate at this institution for all age groups other than those 3 years of age and younger; instead, we used the reported complication rates from other studies for comparison regarding older age groups. CONCLUSION This study suggests that although the complication rate may be higher in this younger population, the nature of the complications does not appear to have a large impact on their LOS. This study also suggests that the 3-years-and-younger group, in the absence of other comorbidities, can safely undergo adenotonsillectomy without undergoing preoperative PSG, provided that close attention is paid to postoperative clinical findings while patients are being monitored. Furthermore, the cost, time, and expertise necessary for interpretation of the data have all affected the relative utility of PSG, especially in the pediatric population. The average cost of overnight PSG at a qualified institution is between
Otolaryngology– Head and Neck Surgery Volume 132 Number 2
$600 and $2,800.21 Although home studies are valuable in the evaluation of adults, young children most often require overnight stays at centers where the study is being performed. Given this information, it would be impractical and not clinically justified to consider performing PSG on all pediatric patients before adenotonsillectomy. Higher risk as indicated by the increased LOS for certain patient groups with comorbidities does appear to exist, and we would reserve the use of preoperative PSG for these patients and for patients with ambiguous clinical history. The data in this study are inconclusive with regard to which medical conditions or comorbidities, and which immediate postoperative findings, would mandate an overnight inpatient stay. We believe that this question can only be answered with a prospective study that controls for the previously mentioned surgical variables and clearly defines the timing, degree, and duration of respiratory complications and their interventions. Preoperative PSG findings had little influence on the postoperative management of patients in this study. Postoperative management decisions are primarily based on the postoperative clinical findings present for each patient. Patients are managed individually on the basis of their postoperative course or established protocol, regardless of the PSG recommendations or predictions. REFERENCES 1. Leach J, Olson J, Hermann J, et al. Polysomnographic and clinical findings in children with obstructive sleep apnea. Arch Otolaryngol Head Neck Surg 1992;118:741-4. 2. Messner AH. Evaluation of obstructive sleep apnea by polysomnography prior to pediatric adenotonsillectomy. Arch Otolaryngol Head Neck Surg 1999;125:353-6. 3. Biavati MJ, Manning SC, Phillips DL. Predictive factors for respiratory complications after tonsillectomy and adenoidectomy in children. Arch Otolaryngol Head Neck Surg 1997;123:517-21. 4. McColley SA, April MM, Carroll JL, et al. Respiratory compromise after adenotonsillectomy in children with obstructive sleep apnea. Arch Otolaryngol Head Neck Surg 1992;118:940-3. 5. Rosen GM, Muckle RP, Mahowald MW, et al. Postoperative
RIEDER and FLANARY 267
6. 7. 8.
9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20.
21. 22.
respiratory compromise in children with obstructive sleep apnea syndrome: can it be anticipated? Pediatrics 1994;93:784-8. Slovik Y, Tal A, Tarasiuk A, et al. Complications of adenotonsillectomy in children with OSAS younger than 2 years of age. Int J Pediatr Otorhinolaryngol 2003;67:847-51. Wilson K, Lakheeram I, Morielli A, et al. Can assessment for obstructive sleep apnea predict post adenotonsillectomy respiratory complications. Anesthesiology 2002;96:313-22. Section on Pediatric Pulmonology, Subcommittee on Obstructive Sleep Apnea Syndrome. Clinical practice guideline: diagnosis and management of childhood obstructive sleep apnea. Pediatrics 2002;109:704-12. American Thoracic Society. Standards and indications for cardiopulmonary sleep studies in children. Am J Respir Crit Care Med 1996;153:866-78. Guilleminault C, Chowdhuri S. Upper airway resistance syndrome is a distinct syndrome. Am J Respir Crit Care Med 2000;161:1412-6. Messner A, Pelayo R. Pediatric sleep-related breathing disorders. Am J Otolaryngol 2000;21:98-107. Guilleminault C, Pelayo R. Sleep-disordered breathing in children. Ann Med 1998;30:350-6. Ross AT, Kazahaya K, Lawrence WC. Revisiting outpatient tonsillectomy in young children. Otolaryngol Head Neck Surg 2003;128:326-31. Goldstein NA, Fatima M, Campbell TF, et al. Child behavior and quality of life before and after tonsillectomy and adenoidectomy. Arch Otolaryngol Head Neck Surg 2002;128:770-5. Serres LM, Derkay C, Sie K, et al. Impact of adenotonsillectomy on quality of life in children with obstructive sleep disorders. Arch Otolaryngol Head Neck Surg 2002;128:489-96. Goldstein NA, Sculerati N, Walsleben JA, et al. Clinical diagnosis of pediatric obstructive sleep apnea validated by polysomnography. Otolaryngol Head Neck Surg 1994;111:611-7. Wang RC, Elkins TP, Keech D, et al. Accuracy of clinical evaluation in pediatric obstructive sleep apnea. Otolaryngol Head Neck Surg 1998;118:69-73. Berkowitz RG, Zalzal GH. Tonsillectomy in children under 3 years of age. Arch Otolaryngol Head Neck Surg 1990;116: 685-6. Helmus C. Tonsillectomy and adenoidectomy in the one and two-year-old child. Laryngoscope 1979;89:1764-71. Postma DS, Folsom F. The case for an outpatient “approach” for all pediatric tonsillectomies and/or adenoidectomies: a 4-year review of 1419 cases at a community hospital. Otolaryngol Head Neck Surg 2002;127:101-8. Leach J, Olson J, Hermann J, et al. Polysomnographic and clinical findings in children with obstructive sleep apnea. Arch Otolaryngol Head Neck Surg 1992;118:741-4. Brouillette RT, Morielli A, Leimanis A, et al. Nocturnal pulse oximetry as an abbreviated testing modality for pediatric obstructive sleep apnea. Pediatric 2000;105:405-12.