 |
|
 |
|
 |
 |
|
 |
 |
 |
 |
 |
 |
 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
* Division of Pediatric Otolaryngology, State University of New York Downstate Medical Center, Brooklyn, New York
Division of Pediatric Pulmonology, State University of New York Downstate Medical Center, Brooklyn, New York
Division of Pediatric Cardiology, State University of New York Downstate Medical Center, Brooklyn, New York
|| Scientific Computing Center, State University of New York Downstate Medical Center, Brooklyn, New York
¶ Department of Communication Science and Disorders, University of Pittsburgh and the Department of Audiology and Communication Disorders, Children’s Hospital of Pittsburgh, Pittsburgh, Pennsylvania
# Department of Pediatric Otolaryngology, Allegheny General Hospital, Pittsburgh, Pennsylvania
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Methods. In a prospective, randomized, investigator-blinded, controlled trial, 59 otherwise healthy children (mean age: 6.3 years [3.0]; 31 boys, 28 girls) with a clinical diagnosis of OSA (clinical assessment score 
40) were recruited from the pediatric otolaryngology and pediatric pulmonary private offices and clinics of a tertiary care, academic medical center. A standardized assessment was performed on all patients, including history, physical examination, voice recording, tape recording of breathing during sleep, lateral neck radiograph, echocardiogram, and PSG. A clinical assessment score was assigned. Children with positive PSG (n = 27) were scheduled for T&A, whereas children with negative PSG (n = 29) were randomized to T&A (n = 15) or no surgery (n = 14). Children were reassessed in an identical manner at a planned 6-month follow-up.
Results. Follow-up was available for 21 patients with positive PSG, 11 patients with negative PSG randomized to T&A, and 9 nonsurgery patients. In the randomized subjects, the median reduction in clinical assessment score was 49 (range: 32–61) for the T&A patients as compared with 8 (range: –9 to 29) for the nonsurgery patients. Nine (82%) of the T&A patients were asymptomatic (clinical assessment score <20) compared with 2 (22%) of the nonsurgery patients.
Conclusion. Children with a positive clinical assessment of OSA but negative PSG have significant improvement after T&A as compared with observation alone, thus validating the clinician’s role in diagnosing upper airway obstruction.
Key Words: obstructive sleep apnea • polysomnography • tonsillectomy • adenoidectomy • sleep-disordered breathing • snoring
Abbreviations: OSA, obstructive sleep apnea • PSG, polysomnography • UARS, upper airway resistance syndrome • T&A, tonsillectomy and adenoidectomy • BMI, body mass index • RDI, respiratory disturbance index • ROC, receiver operating characteristic • AI, apnea index
Obstructive sleep apnea (OSA) was first described in children in the medical literature by Guilleminault et al1 in 1975. They and subsequent investigators described the clinical features found in children with overnight polysomnography (PSG) positive for OSA. The prevalence of pediatric OSA has been estimated to be between 1% and 3% in preschool and school-aged children.2 Additional work in pediatric sleep disorders has shown that sleep-disordered breathing is a continuum of severity from partial obstruction of the upper airway, producing snoring, to increased upper airway resistance syndrome (UARS) to continuous episodes of complete upper airway obstruction or OSA.3 Tonsillectomy and adenoidectomy (T&A) is successful in eliminating obstruction in 85% to 95% of otherwise normal children with OSA.4,5
Although the clinical effects of OSA in children have been well described, reports have documented the inaccuracy of predicting which children with histories and physical examinations suggestive of OSA will have positive PSG. In 7 trials, the accuracy of clinical evaluation of pediatric OSA in predicting positive sleep studies was poor, ranging from 30% to 85%.5–11 On the basis of the published studies and because only 20% to 30% of snoring children have positive PSG, the 1996 American Thoracic Society Consensus Committee recommended that PSG be obtained before T&A to differentiate primary snoring from OSA.12 Although the published studies suggest that clinical evaluation is inaccurate in diagnosing OSA in children, most of the studies used adult criteria for interpretation of the sleep studies, which are now recognized to be inappropriate for children. In addition, none of the studies considered the diagnosis of UARS in the evaluations, which requires esophageal pressure monitoring and is not routine in most centers. Therefore, it is likely that the number of children with significant sleep-disordered breathing was underestimated in these studies, as was the value of clinical assessment.13
Children with a clinical assessment suggestive of OSA but negative PSG are a treatment dilemma. PSG has been considered the "gold standard" diagnostic tool to determine which children with symptoms of upper airway obstruction would benefit from T&A,12 yet numerous previous reports have shown that children’s symptoms of upper airway obstruction improve after T&A regardless of whether apnea is documented by PSG.5,14–18 Reports have also documented improvement in children’s behavior and quality of life.19–22
Our objective was to determine whether otherwise healthy children with a positive clinical assessment of significant upper airway obstruction but with PSG negative for OSA have improvement in their clinical assessment score after T&A as compared with children who do not undergo surgery. If children with a clinical assessment of pediatric OSA but with negative PSG do not improve with observation alone, then we will validate the clinician’s role in diagnosing significant upper airway obstruction.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Children who were between 2 and 14 years of age and were suspected of having sleep-disordered breathing were recruited from the pediatric otolaryngology private office and clinic at the University Hospital of Brooklyn, SUNY Downstate Medical Center, and the otolaryngology and pediatric pulmonary clinics at the Kings County Hospital Center in Brooklyn, NY, from March 1999 through May 2001. Children were referred to the specialty offices by their primary care physicians for evaluation of snoring and nighttime breathing difficulties. All children were evaluated by the principal investigator and were required to have a clinical assessment score 40 to be included in the study (Table 1). Children with recognized craniofacial syndromes, neuromuscular disorders, or known cranial nerve palsies were excluded. The protocol was approved by the SUNY Downstate Medical Center Institutional Review Board, informed consent was obtained from the parents, and a consecutive sample was recruited.
|
15 seconds, 5 to 15 seconds, <5 seconds, and none. School/daytime performance was recorded as poor (left back or failed 1 subject), below average (C/D), average (B/C), and above average (A/B). Recurrent tonsillitis was recorded as 6 to 10 episodes per year, 4 to 5 episodes per year, 2 to 3 episodes per year, and 0 to 1 episode per year. An episode of tonsillitis required a physician visit and treatment with an antibiotic. The physical examination included the patient’s height, weight, blood pressure, presence or absence of mouthbreathing and hyponasality, degree of adenoid facies (open mouth, long face, mandibular hypoplasia scored as severe, mild, absent), tonsil size, and the ability to fog a mirror with nasal breathing (none, poor, fair, well). Body mass index (BMI) was calculated by dividing the child’s weight in kilograms by the square of the height in meters. Hyponasality was scored as present or absent by review of a 2- to 5-minute audiotape recording of patient conversation by a speech pathologist who was blinded to patient group. Tonsil size was graded as the reduction in pharyngeal luminal diameter: 1+, 0% to 25%; 2+, 26% to 50%; 3+, 51% to 75%; and 4+, 76% to 100% as described by Brodsky.23
An audiotape recording of the child’s breathing while asleep (sleep tape) was performed during the PSG by the technician, who was instructed to record the child’s breathing at its worst. At least 20 minutes of the tape was evaluated by the principal investigator and scored as no apnea, snoring without apneic pauses (pauses in breathing of at least 5 seconds) or a struggling sound; moderate apnea, snoring with 5- to 10-second apneic pauses or a struggling sound; or severe apnea, snoring with 10-second apneic pauses and a struggling sound.
The lateral neck radiographs and echocardiograms were obtained by standard techniques. The lateral neck radiograph was graded according to the degree of adenoid hypertrophy: normal adenoid pad, mild adenoid hypertrophy, moderate adenoid hypertrophy, and severe adenoid hypertrophy. The presence or absence of pulmonary hypertension was assessed on the echocardiogram by estimating the pulmonary artery systolic pressure from the tricuspid regurgitation jet using the Bernoulli equation. The pulmonary artery systolic pressure was considered abnormal when it was >30 mm Hg.24 The lateral neck radiographs and echocardiograms were interpreted by clinicians who were blinded to patient group.
Clinical Assessment Score
Each child was assigned a clinical assessment score (Table 1). Symptoms highly suggestive of OSA (pauses, gasping, sleeping with neck extended, daytime sleepiness) contributed more than those that were nonspecific (morning headache, poor school performance, rhinorrhea). The more frequent the symptom, the higher the score. Enuresis was scored only for children aged 4 and above. A score of 2 was given to all children younger than 4. Similarly, height, BMI, and blood pressure were compared with standard percentiles for age and gender, and those at the extremes contributed more than those closer to the norm.25,26 The more severe the adenotonsillar hypertrophy seen on physical examination or demonstrated by the lateral neck radiograph, the higher the score. A high sleep tape score or an echocardiogram demonstrating pulmonary hypertension contributed heavily to the clinical assessment score. In addition to the overall score, each child received a separate score for symptoms, sleep tape, and echocardiogram (score A; Table 1) and physical examination and lateral neck radiograph (score B; Table 1). Surgical removal of the tonsils and adenoids automatically lowered score B in the surgical patients, whereas improvement in score A reflected only clinical improvement.
Although the clinical assessment score has not been previously validated against PSG, the overall score was developed from the clinical assessment used in the first author’s previous prospective study of the clinical assessment of pediatric OSA.7 Clinical features were weighted more or less heavily depending on the factor’s association with OSA according to the statistical analysis performed during the previous data review. The current score also included an assessment of hyponasality as a measure of adenotonsillar hypertrophy27 and an objective measurement of adenoid size from the lateral neck film. The highest possible score was 164. Children with a score 40 were considered to have OSA, children with a score 20 but <40 were considered to have moderate symptoms of upper airway obstruction but not apnea, and children with a score <20 were considered to be asymptomatic. These subdivisions were based on the collective opinion of 2 pediatric otolaryngologists (Dr Goldstein and Dr Post) and a pediatric pulmonologist as well as review of the published studies.
Interobserver Reliability
Clinical assessments were performed by the principal investigator for every child. Additional assessments were performed by 1 or 2 independent examiners, a pediatric pulmonary fellow, and an otolaryngology resident for 25 patients. The additional assessments were limited to the 25 items scored by the examining physician and not independent evaluators (hyponasality, sleep tape, lateral neck radiograph, echocardiogram) or the nursing staff (height, weight, blood pressure). The initial assessments were performed before the PSG. The follow-up assessments were performed by the investigators who were blinded to whether the child had had surgery. Evaluation of tonsil size was performed as the last part of the physical examination after the remainder of the clinical assessment had been recorded. For determining interrater reliability for coding the presence or absence of hyponasality, 20 conversational samples were randomly selected and rated by a second examiner, who was blinded to the treatment status of the child. Subject-to-subject agreement for the presence or absence of hyponasality between the 2 raters was .90 (18 of 20).
PSG
PSG was performed at the SUNY Downstate Pediatric Pulmonary Function, Exercise, and Sleep Physiology Laboratories and consisted of respiratory rate, pulse rate, pulse oximetry, inductive plethysmography of the chest and abdomen and the mathematical sum of the 2 for respiratory effort, and oronasal airflow from a loose mask or oral and nasal thermistors. Obstructive apnea was defined as the cessation of oronasal airflow with continued respiratory effort for at least 2.5 times the typical breath interval, and obstructive hypopnea was defined as a decrease in amplitude of oronasal airflow of at least 50% with no decrease in respiratory effort for the same duration. PSG was considered positive for OSA when the number of obstructive apneas plus hypopneas per hour of sleep (respiratory disturbance index [RDI]) was at least 5 or at least 10% of the night was spent with oxygen saturation <90%. The 6-month follow-up PSG was analyzed by a pulmonologist who was blinded to whether the patient had had surgery.
Randomization
Randomization was performed by using a computerized list generated by the biostatistician in blocks of 2 to ensure equal size treatment groups. T&A was performed at the University Hospital of Brooklyn or the Kings County Hospital Center by a pediatric otolaryngologist who was not 1 of the investigators.
Sample Size Estimation
Because clinical assessment of OSA has been shown to have a positive predictive value of 50%,7 we predicted that half of the patients enrolled would have positive PSG. Because T&A relieves symptoms of upper airway obstruction, we expected that 10% of the PSG– T&A children would remain symptomatic at the 6-month follow-up (overall clinical assessment score 20) and that 90% would be asymptomatic (overall clinical assessment score <20). We estimated that 80% of children who had negative PSG and did not undergo surgery would remain symptomatic at the 6-month follow-up and that 20% would be asymptomatic. Assuming 
of .05 and power of .9, we estimated that we would need a total of 22 children with negative PSG to find a significant difference between the patients who had negative PSG and underwent T&A and those who did not. Of the 22 with negative PSG, 11 would be randomized to the surgery group and 11 to the no surgery group. Therefore, we needed a total of 44 patients for 22 to have negative PSG. Assuming that 10 (23%) patients would drop out, a total of 54 children were initially recruited. When the dropout rate after the recruitment of the initial 54 children was 31%, an additional 5 patients were recruited to help achieve an adequate sample size.
Statistical Methods
The initial clinical assessment scores and change in clinical assessment scores were compared between the PSG– T&A and PSG– nonsurgery groups using the Wilcoxon rank sum test; the initial and final PSG parameters between the PSG– T&A and PSG– nonsurgery groups were also compared using the Wilcoxon test. The proportion of patients with final clinical assessment score <20 and <40 in the PSG– T&A and PSG– nonsurgery groups were compared using Fisher exact test. Comparisons of age and length of follow-up between the PSG– T&A and PSG– nonsurgery groups were performed using the Wilcoxon test; comparison of the gender and race distributions were compared using Fisher exact test. Interrater agreement for the clinical assessment score was measured by intraclass correlations using 2-way random effects models. A mixed linear model was used to test for differences among observers in mean clinical assessment scores. The positive predictive value of the total clinical assessment score in predicting a positive PSG was calculated. The sensitivity, specificity, and positive and negative predictive values of the sleep tape score in predicting a positive PSG were calculated, and Fisher exact test was used to determine their significance. A subset of items from the clinical assessment score (snoring, pauses, gasping, neck extension, daytime sleepiness, adenoid facies, sleep tape) were analyzed by logistic regression to determine the utility of the items in predicting a positive PSG. The area under the receiver operating characteristic (ROC) curve was used as a measure of predictive utility for logistic regression analysis. P < .05 was considered statistically significant. SAS (SAS Institute, Cary, NC) software was used for data analysis.
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
|
|
The intraclass correlations between the principal investigator and the 2 independent examiners for the clinical assessment score were .80 and .65. The mean total score for the principal investigator (50.7 [12.2], n = 60) and the independent examiners (51.9 [15.1], n = 26 and 49.0 [12.8], n = 29) did not differ significantly (P = .194). An additional child who underwent a clinical assessment but was not entered into the study because his clinical assessment score was <40 was included in these analyses.
The initial and change total clinical assessment score, score A and score B, are presented in Table 4 for the 41 children who were included in the final analysis. The median initial total score was 77 with a range of 49 to 114 for the PSG+ group. The median initial total score was 64 with a range of 42 to 77 for the PSG– T&A group and 50 with a range of 40 to 64 for the PSG– nonsurgery group. Although there was no significant difference in the initial score A between the PSG– T&A and PSG– nonsurgery groups, the initial total score was significantly lower in the PSG– nonsurgery group (P = .034) and the difference in score B approached significance (P = .074). The median total change score (initial total score minus final total score) was 59 with a range of 26 to 106 for the PSG+ children. The total change score was 49 with a range of 32 to 61 for the PSG– T&A children and 8 with a range of –9 to 29 for the PSG– nonsurgery children. The total change score, score A change score, and score B change score all were significantly higher in the PSG– T&A than the PSG– nonsurgery children (P = .001, .002, and .007, respectively).
|
|
95th percentile was found in 14 (34%) children: 9 in the PSG+ group, 3 in the PSG– T&A group, and 2 in the PSG– nonsurgery group. At follow-up evaluation, BMI 95th percentile was found in 17 (41%) children: 11 in the PSG+ group, 4 in the PSG– T&A group, and 2 in the PSG– nonsurgery group. The BMI of all 3 children with positive PSG at final evaluation was 95th percentile. BMI was 
5th percentile for only 2 (5%) children at both initial and final evaluations: 1 in the PSG– T&A group and 1 in the PSG– nonsurgery group.
|
95th percentile in 4 children and between the 90th and 95th percentile in 4 children. Three of these children also had a BMI 95th percentile. The hypertension had resolved by follow-up evaluation in all children, but 3 (7%) additional children (all in the initial PSG+ group) developed systemic hypertension. It was 95th percentile in 1 patient and between the 90th and 95th in the other 2 children. All 3 children had negative PSG at final evaluation, although 2 had a BMI 95th percentile. Only 1 child had an echocardiogram suggestive of pulmonary hypertension. The child, in the PSG+ group, had a peak velocity of tricuspid regurgitation jet of 24 to 36 mm Hg, estimating right ventricular pressures of 29 to 41 mm Hg on his initial echocardiogram. The estimated pressure gradient was 19.4 mm Hg, estimating normal right ventricular pressures on his follow-up echocardiogram, and his follow-up sleep study was normal.
The positive predictive value of the initial total clinical assessment score for predicting a positive PSG was 48% (27 of 56). Of the 27 children with positive PSG at their initial evaluation, 10 had sleep tapes scored as severe apnea, 13 had sleep tapes scored as moderate apnea, 3 had sleep tapes negative for apnea, and 1 did not have a sleep tape. Of the 15 children in the PSG– T&A group, initial sleep tapes were scored as severe apnea in 2 children, moderate apnea in 7 children, and negative for apnea in 6 children. Of the 14 children in the PSG– nonsurgery group, initial sleep tapes were scored as severe apnea in 1 child, moderate apnea in 4 children, and negative for apnea in 9 children. Thus, the sleep tape had a sensitivity of 88% (23 of 26), a specificity of 52% (15 of 29), a positive predictive value of 62% (23 of 37), and a negative predictive value of 83% (15 of 18) in predicting a positive PSG (Phi coefficient = .43, P = .002, Fisher exact test). Of the 3 children with positive PSG at the final evaluation, 1 had a sleep tape scored as severe apnea (initial PSG+ group), 1 had a sleep tape scored as moderate apnea (PSG– nonsurgery group), and 1 had a sleep tape negative for apnea (initial PSG+ group). Of the 38 patients with negative PSG at final evaluation, 3 had sleep tapes scored as moderate apnea (all in the PSG– nonsurgery group), and 35 had sleep tapes scored as negative for apnea.
Logistic regression using 7 items from the clinical assessment score (snoring, pauses, gasping, neck extension, daytime sleepiness, adenoid facies, and sleep tape) as predictors was conducted. Only sleep tape was a significant predictor (P = .009); the area under the ROC curve was .84. A simpler model containing only the sleep tape predictor was not significantly inferior to the full model (likelihood ratio test P = .090); area under the ROC curve for this reduced model was .75. In other words, although the sleep tape was a moderately useful predictor, the addition of the other items did not significantly increase the predictive utility.
Of the 37 children who underwent T&A, the only major complication was a right thigh burn to 1 child in the initial PSG+ group secondary to improper application of the grounding of the electrocautery unit. The wound healed with dressing changes, but the child was hospitalized for 3 days. Children were observed in the hospital for a mean of 21 (12) hours after surgery. Children in the PSG– T&A group who were at least 3 years of age were usually discharged from the hospital 6 hours after surgery, whereas the children in the PSG+ group were admitted for pulse oximetry observation for 1 night unless the PSG was considered mild enough to warrant a 6-hour stay. Children were observed either on the floor or in a monitored unit at the discretion of the attending surgeon. There were no postoperative respiratory complications, episodes of postoperative hemorrhage, or readmissions to the hospital.
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Our results demonstrate that the signs and symptoms of OSA improve after T&A regardless of whether OSA is documented by PSG. At follow-up, the children in the PSG– T&A group had significant improvement in the total clinical assessment score as well as the subset scores as compared with children in the PSG– nonsurgery group. The subset scores were designed to separate the items that would automatically be decreased after T&A (tonsil size and adenoid size on the lateral neck radiograph, score B) from the items that would solely reflect clinical improvement (score A). Eighty-two percent of the PSG– T&A group were asymptomatic at follow-up as compared with 22% of the PSG– nonsurgery patients, resulting in a statistically significant difference between the groups. Our findings were close to the assumptions in our sample size estimation, which predicted that 90% of the PSG– T&A children and 20% of the PSG– nonsurgery children would be asymptomatic at follow-up. There was also a significant difference between the groups in children who were moderately symptomatic (clinical assessment score 20 but <40) at final evaluation. One (11%) child in the PSG– nonsurgery group developed OSA at follow-up, a finding that is comparable to the 10% reported rate of the progression of primary snoring to OSA.
A weakness of our findings is that despite randomization at study outset, the PSG– nonsurgery group had a significantly lower initial total clinical assessment score than the PSG– T&A group. The difference resulted from findings on the physical examination (score B) as opposed to differences in symptoms (score A) and occurred because of the small sample size of our patient population. To correct for the differences in initial clinical assessment scores, we compared changes in clinical assessment scores between the 2 groups at the final evaluation.
The clinical assessment score may prove to be a useful tool for the office diagnosis of pediatric OSA. Additional study to validate the score against PSG and other external measures such as quality-of-life instruments is needed. Its responsiveness to longitudinal change must also be evaluated further. Our study demonstrated a significant difference in change scores between the PSG– T&A and PSG– nonsurgery groups using a score of 40 as the requirement for entry, but there also may be a role for treatment of children with scores 20 but <40.
Although we demonstrated significant improvement in clinical scores, are these changes clinically relevant? Snoring was only 1 of >20 nighttime and daytime symptoms that brought these children to medical attention. Entry into the study required a clinical assessment that revealed more than just snoring. Symptoms of upper airway obstruction have been shown to have a significant impact on children’s quality of life. De Serres et al19 administered the Obstructive Sleep Disorders-6 Survey, a validated health-related quality-of-life instrument, to the caregivers of 101 children from 7 tertiary care pediatric otolaryngology practices across the United States, before and after T&A performed for treatment of sleep-disordered breathing. Children’s sleep-disordered breathing was diagnosed on the basis of clinical assessment as only 8% had preoperative sleep studies. Domains of the survey most affected at initial evaluation were physical suffering, sleep disturbance, and caregiver concern. Postoperatively, 90% of children had improvement in quality of life, which was considered large in 75% and moderate in 6%.
Goldstein et al20 administered another validated quality-of-life survey, the OSA-18, along with a standardized measure of children’s behavior, the Child Behavior Checklist, to the caregivers of 64 children before T&A and 3 months postoperatively. Surgery was performed for the treatment of upper airway obstruction in 84% of children, 92% of whom received a diagnosis on a clinical basis as only 8% had preoperative PSG. Postoperatively, a large change in quality of life was found for the OSA-18 domains of sleep disturbance, caregiver concerns, and physical symptoms, and a moderate change was found for emotional symptoms and daytime function. As in the study by de Serres et al,19 the change in the impact of sleep-disordered breathing on quality of life was highly significant after T&A. The study also found significant improvement in behavioral and emotional difficulties after T&A as measured by the Child Behavior Checklist Total Problem Score and the individual scales of Withdrawn, Somatic Complaints, Anxious/Depressed, Thought Problems, Attention Problems, Delinquent Behavior, Aggressive Behavior, Sleep Problems, and Destructive Behavior.
As in the other studies of the accuracy of clinical diagnosis in predicting a positive PSG, esophageal pressure monitoring for detecting UARS was not performed. Children with UARS demonstrate abnormally increased upper airway resistance, resulting in increased respiratory effort and sleep fragmentation without apneic or hypopneic episodes or desaturation. The clinical effects of UARS are identical to OSA and treatment is the same.30 The children with negative PSG in our study may have been experiencing UARS, with the resulting resolution of signs and symptoms in those who underwent T&A. Because the measurement of UARS is not routine in most centers, children with sleep-disordered breathing are often missed by standard PSG.13 Clinical assessment remains a valuable tool in assessing upper airway obstruction and sleep-disordered breathing and determining the need for T&A.
Our accuracy of predicting a positive PSG in children with a clinical assessment suggestive of OSA (48%) when UARS is not evaluated agrees with the findings of previous studies of pediatric OSA.5–11 Our definition of a positive PSG was an RDI 5. Many sleep laboratories consider an AI 1 to be abnormal, based on studies of normal children.31 If we considered children with an AI 1 to have abnormal PSG, then 6 of the 15 PSG– T&A children and 5 of the 14 PSG– nonsurgery children would have had initial positive sleep studies, increasing our clinical accuracy to 68%. However, 13 (41%) of the 32 children who underwent T&A would have had positive PSG at final evaluation, which is much higher than the published reports. Because most of the children were asymptomatic at final evaluation, an AI 1 seems too strict a definition for a positive PSG. Using the RDI 5 as the definition of a positive PSG, 10% of the children with initial PSG positive for OSA had final PSG still positive for apnea. This finding agrees with previous studies that report that T&A is curative in 85% to 95% of otherwise healthy children.4,5
Using a logistic regression model, only sleep tape was a moderately useful predictor of a positive PSG. The addition of 6 items from the clinical assessment did not increase the predictive utility of the model. Previous studies have also demonstrated that scores that were based on a subset of items from the clinical assessment were not sufficiently predictive to determine which children would have a positive PSG.6,7,11,32
The sleep tape had a sensitivity of 88%, a specificity of 52%, and a positive predictive value of 62% in predicting a positive PSG. These results are similar to those of Lamm et al,33 who found that an audiotape had a median sensitivity of 71% and median specificity of 80% in predicting a positive PSG in 29 children who were referred to a sleep laboratory for evaluation of sleep-disordered breathing. Although not specific enough to distinguish children with positive and negative studies, the sleep tape is an inexpensive, convenient method to confirm the parents’ description of the child’s nighttime breathing difficulties. Home videotape recording of the child’s nighttime breathing has also been shown to be a reliable screening method for OSA with a sensitivity of 94% and a specificity of 68% in predicting a positive study.34
In early studies of pediatric OSA, 27% to 69% of children presented with failure to thrive.3 Now that affected children are identified earlier, failure to thrive is less common. Five percent of the children in our study presented with failure to thrive, which agrees with more recent reports of a 4% to 13% incidence.6,7 Thirty-four percent of the children in our study presented with obesity (BMI 95th percentile), which is similar to recent reports of an incidence of 26% to 40%.6,7 As demonstrated in previous studies, substantial weight gain is often found after T&A even in children who are obese preoperatively.35
In adults, hypertension is a common complication of OSA. Early reports found systemic hypertension in 10% to 25% of children with OSA, although these children were severely affected.36,37 In a more recent report, Kunzman et al38 found no increased incidence of systemic hypertension in 22 children with OSA as compared with control subjects. Goldstein et al7 found that 18% of children who presented for evaluation of OSA were hypertensive. Marcus et al39 found that 41 children with OSA had a significantly higher diastolic blood pressure than 26 children with primary snoring, although there was no significant difference in systolic blood pressure between the 2 groups. BMI was a significant predictor of elevated blood pressure. In our study, 20% of the children presented with systemic hypertension, 38% of whom were obese. It resolved in all children by final evaluation, although 3 new children were hypertensive, 2 of whom were obese. As indicated in previous reports, systemic hypertension is a complication of sleep-disordered breathing, especially in obese children. Early studies reported that between 10% and 56% of children with OSA had pulmonary hypertension or cor pulmonale.8,14,36 Now that children receive a diagnosis at a much earlier stage, the incidence is much lower. In our study of otherwise healthy children, only 1 (2%) child presented with mild pulmonary hypertension diagnosed by echocardiography.
PSG has been recommended before T&A to identify children who are at risk for postoperative respiratory complications. Although the risk of postoperative respiratory complications in the general pediatric population ranges from 0% to 1.3%, rates of 16% to 27% have been reported in children with OSA.40–43 Risk factors are age under 3, pulmonary hypertension or other cardiac abnormalities, craniofacial syndromes, failure to thrive, hypotonia, acute airway obstruction, morbid obesity, and severe sleep study indices. Anesthetic technique including the use of opioids has not been shown to influence the rate of postoperative respiratory complications.43 The majority of children described in these studies would have been identified by preoperative clinical assessment without the need for sleep studies. In addition, most postoperative respiratory complications occur within 2 hours after surgery.44,45 As long as appropriate facilities and staff are available for treatment of these children, including the option for overnight observation, a preoperative sleep study is not necessarily needed. PSG has been most useful to confirm the diagnosis of OSA and document its severity in the following situations: children who are younger than 2 years; high-risk patients for which surgery is contraindicated; children with craniofacial anomalies, morbid obesity, or cerebral palsy; when there is a discrepancy between the history and physical examination; and children who remain symptomatic after T&A. PSG is also a prerequisite to treatment with nasal continuous or bilevel positive airway pressure in high-risk children or surgical failures.13,46
|
CONCLUSIONS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
Presented at the Annual Meeting of the American Society of Pediatric Otolaryngology; May 4, 2003; Nashville, TN.
We thank Alex Sternberg, ScD, for performing the polysomnograms; Jennifer J. Black for assistance with analysis of the voice tapes; and Kazi Azam, RDCS, FASE, and Denise Lei, RDCS, for performing the echocardiograms. Ari J. Goldsmith, MD; Jessica W. Lim, MD; and Krishnamurthi Sundaram, MD, along with the SUNY Downstate otolaryngology residents, performed the T&A procedures. We thank Ellen M. Mandel, MD, and Jonathan D. Finder, MD, for assistance with study design and Richard M. Rosenfeld, MD, MPH, for critical review of the manuscript. We thank Tiffany Morgan for administrative support.
|
FOOTNOTES |
|---|
Reprint requests to (N.A.G.) Department of Otolaryngology, SUNY Downstate Medical Center, 450 Clarkson Ave, Box 126, Brooklyn, NY 11203. E-mail: ngoldstein{at}0040downstate.edu
Dr Pugazhendhi’s current affiliation is the Department of Pediatrics, Medical College of Virginia, Richmond, Virginia.
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  K. Murto, N. Eipe, and E. McLeod 473881 - PEDIATRIC PERIOPERATIVE SCREENING FOR SLEEP APNEA PRIOR TO ADENOTONSILLECTOMY Can J Anesth, June 1, 2008; 55(suppl_1): 473881 - 473881. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Buskens, B. van Staaij, J. van den Akker, A. W. Hoes, and A. G. M. Schilder Adenotonsillectomy or Watchful Waiting in Patients With Mild to Moderate Symptoms of Throat Infections or Adenotonsillar Hypertrophy: A Randomized Comparison of Costs and Effects Arch Otolaryngol Head Neck Surg, November 1, 2007; 133(11): 1083 - 1088. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. J. Paavonen, S. Strang-Karlsson, K. Raikkonen, K. Heinonen, A.-K. Pesonen, P. Hovi, S. Andersson, A.-L. Jarvenpaa, J. G. Eriksson, and E. Kajantie Very Low Birth Weight Increases Risk for Sleep-Disordered Breathing in Young Adulthood: The Helsinki Study of Very Low Birth Weight Adults Pediatrics, October 1, 2007; 120(4): 778 - 784. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Guilleminault, J. H. Lee, and A. Chan Pediatric Obstructive Sleep Apnea Syndrome Arch Pediatr Adolesc Med, August 1, 2005; 159(8): 775 - 785. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Kotagal Childhood obstructive sleep apnoea BMJ, April 30, 2005; 330(7498): 978 - 979. [Full Text] [PDF]
|
|
|
|
 |
|
 Journal Watch Arch. Dis. Child., December 1, 2004; 89(12): 1175 - 1176. [Full Text] [PDF]
|
|
|
|
 |
|
 Snoring, Sleep Apnea, and Surgery in Children Journal Watch Pediatrics and Adolescent Medicine, August 9, 2004; 2004(809): 9 - 9. [Full Text]
|
|
|
|
 |
|
 Snoring, Sleep Apnea, and Surgery in Children Journal Watch (General), July 16, 2004; 2004(716): 4 - 4. [Full Text]
|
|
| ||||||||||||||
