Original Research—Pediatric Otolaryngology
Adenotonsillectomy vs Observation for Management of Mild Obstructive Sleep Apnea in Children
Otolaryngology– Head and Neck Surgery 2014, Vol 150(1) 126–132 Ó American Academy of Otolaryngology—Head and Neck Surgery Foundation 2013 Reprints and permission: sagepub.com/journalsPermissions.nav DOI: 10.1177/0194599813509780 http://otojournal.org
Peter G. Volsky, MD1, Meghan A. Woughter2, Hind A. Beydoun, PhD3, Craig S. Derkay, MD1,4, and Cristina M. Baldassari, MD1,4
No sponsorships or competing interests have been disclosed for this article.
Abstract Objective. To determine the impact of adenotonsillectomy vs observation on quality of life (QOL) in children with mild obstructive sleep apnea (OSA). Study Design. Prospective, nonrandomized trial. Setting. Tertiary children’s hospital. Subjects and Methods. Sixty-four children (ages 3-16 years) with mild OSA (apnea hypopnea index between 1 and 5 on polysomnogram) completed the study. Caregivers chose between management options of adenotonsillectomy and observation and completed validated QOL instruments (OSA18 and Children’s Health Questionnaire) at baseline, early, and late follow-ups. The primary outcome measure was QOL. Results. Thirty patients chose adenotonsillectomy, while 34 were observed. Total OSA-18 scores at baseline were significantly poorer (P = .01) in the surgery group (72.3) compared with the observation group (58.5). Four months following surgery, OSA-18 scores improved by 39.1 points over baseline (P = .0001), while there was no change for the observation group (P = .69). After 8 months, OSA-18 scores remained improved in the surgery group, and observation group scores improved by 13.4 points over baseline (P = .005). While OSA-18 scores at the late follow-up visit were poorer in the observation group, the difference was not statistically significant (P = .05). Six observation patients opted for adenotonsillectomy during the study. Conclusion. Quality of life significantly improves in children with mild OSA after adenotonsillectomy. In children with mild OSA who are observed, QOL improvements at early follow-up are less pronounced, but significant improvements in QOL are evident after 8 months. QOL instruments may be useful tools to help providers determine which children with mild OSA may benefit from early intervention. Keywords sleep apnea, obstructive, tonsillectomy, adenoidectomy, child, quality of life
Received June 7, 2013; revised September 27, 2013; accepted October 1, 2013.
O
bstructive sleep apnea (OSA) is a sleep-related breathing disorder that is characterized by intermittent episodes of upper airway collapse during sleep. It is a public health concern since it affects 2% to 3% of preschool-age children in the United States.1 Pediatric OSA has been linked to metabolic changes, growth inhibition, and cardiovascular sequelae. Furthermore, there is a growing body of literature demonstrating the negative impact of OSA on quality of life (QOL) and cognitive function in children. Pediatric OSA has been associated with behavior problems, poor attention, memory and cognitive deficits, and poor school performance.2 The gold standard for diagnosis of pediatric OSA is a full-night polysomnogram (PSG). The severity of OSA is categorized according to the obstructive apnea hypopnea index (AHI) on the PSG. According to the most commonly used system, an AHI between 1 and 5 indicates mild OSA.3 Although PSG reliably measures the presence of OSA and provides an objective scale for OSA severity, it fails to quantify the impact of OSA on a child’s general well-being, including emotional and behavioral health. Interestingly, QOL does not correlate with AHI in pediatric patients.4
1
Department of Otolaryngology–Head and Neck Surgery, Eastern Virginia Medical School, Norfolk, Virginia, USA 2 Eastern Virginia Medical School, Norfolk, Virginia, USA 3 Graduate Program in Public Health, Eastern Virginia Medical School, Norfolk, Virginia, USA 4 Department of Pediatric Otolaryngology, Children’s Hospital of the King’s Daughters, Norfolk, Virginia, USA This research was presented at the American Society of Pediatric Otolaryngology (ASPO) Annual Meeting; April 26-28, 2013; Alexandria, Virginia. Corresponding Author: Cristina M. Baldassari, MD, Department of Otolaryngology–Head and Neck Surgery, Eastern Virginia Medical School, 601 Children’s Lane, 2nd Floor, Norfolk, VA 23507, USA. Email: [email protected]
Downloaded from oto.sagepub.com at NORTH CAROLINA STATE UNIV on March 18, 2015
Volsky et al
127
Quality of life is increasingly recognized as an important health outcome measure in clinical medicine. A recent meta-analysis5 demonstrated poor QOL scores in children with OSA and significant improvements in QOL following adenotonsillectomy, irrespective of AHI. The authors concluded that the impact of OSA affects children’s behavior, daytime functioning, and family life.5 Thus, many children with mild OSA potentially warrant treatment. In children, adenotonsillar hyperplasia is the most common cause of upper airway obstruction. Thus, the primary treatment for pediatric OSA is adenotonsillectomy.6 Numerous studies have demonstrated improvements in neuropsychological outcomes such as behavior, school performance, and attention following surgery for pediatric OSA.1 There is also a growing body of literature highlighting improvements in QOL following adenotonsillectomy.5 Management of children with mild OSA, however, remains controversial, since the natural history of mild OSA is not well understood.7 Thus, it is challenging to counsel caregivers about how best to treat mild OSA. Our primary objective was to determine the impact of adenotonsillectomy vs observation on QOL in children with mild OSA.
Methods A prospective trial was designed to assess the impact of adenotonsillectomy vs observation on QOL in children with mild OSA. The protocol was approved by the Eastern Virginia Medical School Institutional Review Board and registered with the National Institutes of Health website, clinicaltrials.gov. The study was carried out from March 2011 to January 2013. Subjects were recruited from the outpatient pediatric otolaryngology clinic at The Children’s Hospital of the King’s Daughters, a tertiary care children’s hospital in Norfolk, Virginia. Inclusion criteria were (1) children 3 to 16 years old; (2) new diagnosis of mild OSA, defined by AHI from 1 to 5 on full-night PSG; and (3) tonsillar hypertrophy on physical examination. The Brodsky grading scale,8 which is based on the ratio of tonsil size to the tonsillar pillar, was used to assess for tonsillar hypertrophy. Children with grade 2, 3, and 4 tonsils were considered to have tonsillar hypertrophy. Subjects with craniofacial abnormalities, cerebral palsy, Trisomy 21, or prior adenotonsillar surgery were excluded. Children with parents who could not speak English were also excluded. Two validated instruments, the Obstructive Sleep Apnea 18 (OSA-18) and the Child Health Questionnaire Parent Form 28 (CHQ PF-28), were used to assess QOL. The OSA-18, developed by Franco et al,9 is an 18-item questionnaire specific to OSA that measures the relative severity of sleep-related problems with a Likert scale ranging from 1 (none of the time) to 7 (all of the time). The OSA-18 instrument yields a total score and 5 domain scores, including scores for sleep disturbance, physical suffering, emotional distress, daytime problems, and caregiver concerns. Total OSA-18 QOL scores between 60 and 80 indicate a moderate QOL impairment.9 Higher OSA-18 scores are associated
with poorer QOL. The CHQ PF-28 (HealthActCHQ, Inc, Cambridge, Massachusetts) is a measure of a child’s overall or global QOL for which normative data are available for healthy children.10 The CHQ PF-28 is composed of 28 questions. The CHQ PF-28 scores are reported as 2 subsets, Physical Functioning and Psychosocial Functioning, with higher scores indicating better QOL. Children with newly diagnosed mild OSA on PSG were recruited. The authors initially had difficulty recruiting patients for randomization, so a nonrandomized arm was added. Consequently, parents could opt out of randomization by choosing to enroll in either the adenotonsillectomy (surgery) group or the observation group. Medical management with intranasal steroids and montelukast was not offered to patients enrolled in this study. Parents, with input from the child, completed both the OSA-18 and CHQ PF-28 at baseline and at 2 follow-up visits. The target intervals for follow-up assessments were 3 and 6 months following enrollment in the observation group and 3 and 6 months following adenotonsillectomy in the surgery group. At least 1 attempt was made by phone to contact patients who failed to follow up. Subjects enrolled in the observation group with symptomatic exacerbation were able to drop out of this group and undergo adenotonsillectomy after 3 months. Such children were labeled the ‘‘failed observation’’ group. We performed an intention-to-treat analysis in which these data were analyzed as part of the observation group. All PSGs were performed in dedicated pediatric sleep laboratories in accordance with the American Academy of Sleep Medicine (AASM) guidelines11 and were scored by pediatric sleep medicine specialists. Standard procedure included a 6-lead electroencephalogram (EEG), bilateral electrooculogram (EOG) leads, and 1 submental and 2 tibial electromyograms (EMGs). Respiratory measurements were recorded using inductance pneumography, airflow using a nasal pressure transducer, oxygen saturation by pulse oximetry, and carbon dioxide using a carbon dioxide sensor. Information obtained from each PSG included sleep efficiency, number and classification of arousals and respiratory events, and oxygen and carbon dioxide levels. Demographic data, including age, sex, ethnicity, and body mass index (BMI) percentiles (calculated according to the US Centers for Disease Control and Prevention guidelines),12 were also recorded. A power analysis was conducted. A group size of 37 was deemed sufficient to detect a 20-point difference in OSA-18 QOL scores using a 2-group t test with 80% power and a 0.05 two-tailed significance level. It was anticipated that subjects would enroll in the surgery and observation groups on a 1:1 ratio, so a total of 74 subjects were needed. Due to anticipated subject withdrawal, the authors planned to enroll at least 150 patients. Descriptive statistics, including means and standard deviation for continuous variables or frequencies and percentages for categorical variables, were used to evaluate baseline characteristics of observation and surgery groups. The x2 test was used to compare the distribution of categorical variables across groups, while independent
Downloaded from oto.sagepub.com at NORTH CAROLINA STATE UNIV on March 18, 2015
128
Otolaryngology–Head and Neck Surgery 150(1)
Table 1. Subject demographic and polysomnogram data. Subjects Completing the Study
Sex Male Female Tonsil grade 2 3 4 Age, mo BMI, kg/m2 BMI percentile Total AHI O2 nadir Arousal index Peak end-tidal CO2 Baseline OSA-18 score Baseline CHQ PF-28 Physical Functioning Psychosocial Functioning
Subjects Withdrawn
Surgery Group (n = 30)
Observation Group (n = 34)
16 14
20 14
10 15 4 80.8 6 33.7 18.9 6 6.1 68.9 6 28.6 3.3 6 1.0 91.2 6 2.4 11.8 6 3.8 50.1 6 5.9 72.3 6 20.0
10 19 5 80.0 6 41.9 20.2 6 6.2 74.0 6 25.5 2.9 6 1.1 90.8 6 4.1 10.2 6 3.4 48.9 6 4.2 58.5 6 21.5
44.9 6 15.4 39.8 6 13.9
50.6 6 7.8 46.3 6 11.8
X-Surgery Group (n = 29)
X-Observation Group (n = 20)
9 20
9 11
.93 .41 .45 .22 .65 .091 .39 .01
6 15 7 69.7 6 25.8 18.4 6 4.1 67.1 6 34.9 3.1 6 1.2 90.4 6 2.5 11.4 6 3.7 48.2 6 5.2 72.6 6 16.3
5 7 5 97.4 6 43.3 18.6 6 6.2 61.6 6 33.7 2.4 6 1.0 91.9 6 1.6 9.6 6 2.5 45.3 6 5.7 54.6 6 20.7
.07 .05
47.0 6 11.7 36.5 6 14.4
53.4 6 8.5 47.5 6 12.2
P Value .66
.72
P Value, Surgery vs X-Surgery
P Value, Observation vs X-Observation
.22
.78
.43
.30
.16 .72 .82 .46 .20 .67 .27 .93
.15 .39 .14 .07 .16 .46 .03 .52
.56 .38
.23 .74
Abbreviations: AHI, apnea hypopnea index; BMI, body mass index; CHQ PF-28, Child Health Questionnaire Parent Form 28; CO2, carbon dioxide.
samples t tests were used for continuous variables. The OSA-18 and CHQ PF-28 scores were compared across groups using independent samples t test at baseline, early follow-up, and late follow-up. Wilcoxon’s rank sum test was used to assess for differences in OSA-18 and CHQ PF28 scores within the surgery and observation groups between baseline, early follow-up, and late follow-up. Finally, a paired t test was used to examine OSA-18 and CHQ PF-28 score changes within each group. Values of P \ .05 were considered significant.
Results Subjects One hundred eleven subjects agreed to participate and were enrolled in either the surgery or observation group based on parent and child preference. An additional 2 patients agreed to randomization to the surgery or observation group. Twenty-nine subjects were withdrawn due to failure to complete follow-up QOL surveys, and 3 subjects elected for voluntarily withdrawal. An additional 17 patients were deactivated at the study termination before they returned for follow-up. Thus, 64 children were included in the final analysis. Thirty subjects comprised the surgery group, while 34 subjects comprised the observation group. There were no differences between the 2 groups in terms of demographic and polysomnographic data (Table 1). The duration of time between enrollment and early follow-up was an average of 113 6 29 days (approximately
4 months). The late follow-up survey was completed 232 6 51 days (approximately 8 months) after enrollment in the observation group or following adenotonsillectomy in the surgery group. Twenty-five percent (n = 16) of subjects completed both early and late QOL surveys.
OSA-18 QOL Scores Total OSA-18 scores at baseline were significantly poorer (P = .01) in the surgery group (72.3 6 20.0) compared with the observation group (58.5 6 21.5). Between the 2 groups, there were also differences in subscores of Sleep Disturbance, Daytime Function, and Caregiver Concerns (Table 2). The mean OSA-18 total score for the surgery group significantly improved (P = .0001) from baseline to the early follow-up by 39.1 6 18.8 points. Significant gains in QOL were also evident for the 5 OSA-18 domain scores following adenotonsillectomy (Table 2). At the early follow-up visit, the total OSA-18 score for the observation group did not significantly change (P = .69) from the initial baseline score (Table 2). When the 2 groups were compared at the early follow-up interval, the surgery group had significantly (P = .0001) better OSA-18 QOL scores than the observation group. Improvements noted in OSA-18 QOL scores following adenotonsillectomy were maintained at the late follow-up (Table 2). Interestingly, in the observation group, there was also a significant improvement (P = .005) in the total
Downloaded from oto.sagepub.com at NORTH CAROLINA STATE UNIV on March 18, 2015
Volsky et al
129
Table 2. OSA-18 QOL scores for surgery and observation groups.a Surgery Group (S)
Observation Group (O)
Pb
Pc
Pc
Baseline (T0)
(S0, n = 30)
(O0, n = 34)
S0 vs O0
Sleep Disturbances Physical Symptoms Emotional Symptoms Daytime Function Caregiver Concerns Total
17.3 6 4.9 13.8 6 6.0 13.4 6 5.9 13.0 6 4.7 14.7 6 6.8 72.3 6 20.0
14.3 6 4.8 11.9 6 4.8 11.0 6 5.6 9.8 6 4.6 10.1 6 5.8 58.5 6 21.5
.019d .17 .09 .008d .006d .01d
Early Follow-up (T1)
(S1, n = 24)
(O1, n = 25)
S1 vs O1
S1 vs S0
O1 vs O0
Sleep Disturbances Physical Symptoms Emotional Symptoms Daytime Function Caregiver Concerns Total
6.3 6 2.8 6.8 6 3.2 8.4 6 4.5 5.8 6 2.9 6.6 6 3.7 33.9 6 14.6
14.6 6 6.4 12.0 6 5.6 10.9 6 5.8 9.4 6 5.0 11.2 6 5.9 58.2 6 24.5
\.0001d .0003d .09d .004d .002d .0001d
\.0001d \.0001d \.0001d \.0001d \.0001d \.0001d
Late Follow-up (T2)
(S2, n = 11)
(O2, n = 18)
S2 vs O2
S2 vs S0
Sleep Disturbances Physical Symptoms Emotional Symptoms Daytime Function Caregiver Concerns Total
6.7 6 3.2 6.5 6 2.3 8.3 6 4.3 6.3 6 3.2 5.8 6 2.7 33.6 6 8.6
12.0 6 5.6 9.7 6 6.0 8.6 6 6.8 7.1 6 4.1 8.2 6 4.8 45.1 6 21.9
.003d .04d .85 .58 .08 .05
.0002d .01d .01d .003d .0007d .0002d
.60 .40 .81 .86 .19 .69 O2 vs O0 .004d .52 .03d .01d .31 .005d
Abbreviation: QOL, quality of life. a Higher scores indicate poorer QOL. Early follow-up (T1): 113 6 29 days. Late follow-up (T2): 232 6 51 days. b Independent samples t test. c Paired t test. d P \.05 is statistically significant.
CHQ PF-28 QOL Scores At baseline, the Physical and Psychosocial Functioning total scores were similar between the surgery and observation groups (P = .05; Table 3). Among the CHQ PF-28 subscores, Physical Functioning (P = .011) and Role/Social Limitations—Physical (P = .025) were poorer at baseline in the surgery group. After adenotonsillectomy, improvement occurred in the Physical and Psychosocial Functioning scores of the surgery group (Table 3). Also, 11 subscales improved from baseline to the early follow-up in the surgery group, including Global Health (P \ .0001), Physical Functioning (P = .005), Role/Social Limitations—Emotional/Behavioral
120
OSA-18 Total Score
OSA-18 QOL score, with a 13-point change between the baseline and late follow-up visits (Figure 1). Three domain scores—Sleep Disturbances, Emotional Symptoms, and Daytime Function—also improved in the observation group during that time. Total OSA-18 scores at the late follow-up visit were poorer in the observation group compared with the surgery group. However, this difference was not statistically significant (P = .05).
100 80
72.3 (64.9, 79.7) 58.2 (48.6, 67.8)
60
45.1 (35.2, 54.9)
58.5 (51.3, 65.7)
40 33.9 (28.0, 39.7)
20 0
Baseline
Early follow-up
33.6 (28.7, 38.5)
Late follow-up
Time Interval Surgery Group
Observation Group
Figure 1. Total OSA-18 quality-of-life (QOL) scores. Total OSA18 QOL scores with 95% confidence intervals are depicted at baseline, early follow-up (113 6 29 days), and late follow-up (232 6 51 days) for both the surgery and observation groups. Higher OSA-18 Total Scores indicate poorer QOL.
Downloaded from oto.sagepub.com at NORTH CAROLINA STATE UNIV on March 18, 2015
130
Otolaryngology–Head and Neck Surgery 150(1)
Table 3. CHQ PF-28 QOL scores for surgery and observation groups.a Surgery Group (S)
Observation Group (O)
Pb
Baseline (T0)
(S0, n = 30)
(O0, n = 34)
S0 vs O0
Total—Physical Total—Psychosocial
44.9 6 15.4 39.8 6 13.9
50.6 6 7.8 46.3 6 11.8
.07 .05
Early Follow-up (T1)
(S1, n = 24)
(O1, n = 25)
S1 vs O1
Pc
Pc
S1 vs S0
O1 vs O0
d
Total—Physical Total—Psychosocial
52.3 6 9.4 50.1 6 12.7
47.2 6 12.8 47.2 6 10.1
.13 .40
.02 .003d
.66 .24
Late Follow-up (T2)
(S2, n = 11)
(O2, n = 18)
S2 vs O2
S2 vs S0
O2 vs O0
Total—Physical Total—Psychosocial
50.3 6 6.2 50.6 6 8.7
49.9 6 12.8 52.0 6 10.9
.93 .74
.02d .28
.89 .02d
Abbreviations: CHQ PF-28, Child Health Questionnaire Parent Form 28; QOL, quality of life. a Lower scores indicate poorer QOL. Early follow-up (T1): 113 6 29 days. Late follow-up (T2): 232 6 51 days. b Independent samples t test. c Paired t test. d P \.05 is statistically significant.
(P = .004), Behavior (P = .001), Global Behavior Item (P = .007), Mental Health (P = .009), Self-Esteem (P = .01), General Health Perceptions (P = .01), Change in Health (P \ .0001), Parental Impact—Emotional (P = .0002), and Family Activities (P = .002). In the observation group, there was no change in the total CHQ PF-28 scores between baseline and early follow-up, while the subscale of Family Cohesion (P = .02) actually deteriorated. When comparing the 2 groups at early follow-up, total scores for Physical and Psychosocial Functioning were similar (Table 3). However, the surgery group did have better QOL scores in the 4 subscales of Global Health (P = .033), Physical Functioning (P = .037), Behavior (P = .02), and Change in Health (P \ .0001). From baseline to the late follow-up, the surgery group maintained improvements in CHQ PF-28 total Physical and Psychosocial Functioning scores (Table 3). In the observation group, the Psychosocial Functioning score was improved (P = .02) from baseline, but there was not a significant change noted in the Physical Functioning score (Table 3). At the time of late follow-up, there was no difference in Physical and Psychosocial Functioning between groups (Table 3).
Failed Observation Group Six subjects initially enrolled in the observation group experienced an exacerbation of OSA symptoms during the study, and their caregivers elected to proceed with adenotonsillectomy. These children were labeled the ‘‘failed observation’’ group. In 5 children, exacerbation of OSA symptoms occurred in the first 6 months of enrollment, while 1 subject deteriorated after 6 months. The baseline demographic and polysomnographic variables for the failed observation group and the observation group were similar (P . .05). Interestingly, the mean baseline OSA-18 total score for the
failed observation group was 73.0 6 24.9, which was poorer than the mean total score for the observation group, 55.4 6 19.8. This difference, however, was not significant (P = .07).
Discussion A recently published meta-analysis highlighted the impact of OSA on QOL in children.5 The authors showed that pediatric patients with OSA have a poorer health status than healthy children. While full-night PSG remains the gold standard for determining the severity of OSA in children, numerous studies have shown a poor correlation between AHI and QOL scores.5 Furthermore, behavioral problems associated with OSA were not related to the severity of OSA as defined by the AHI on PSG.13 Thus, QOL assessments are being increasingly recognized as an important part of the evaluation of patients with OSA. In addition to being used to study the disease burden of OSA, QOL assessments are also being used to study outcomes of treatment, including adenotonsillectomy. The appropriate management for mild OSA in children continues to be debated. Some providers advocate for watchful waiting. This conservative approach is affirmed by a study by Calhoun et al14 that demonstrated that children with mild OSA have neurocognitive functioning equivalent to control patients. Proponents for more aggressive treatment cite recent studies that demonstrate neurocognitive deficits in children with only mild airway obstruction, including primary snoring.15,16 Mitchell and Kelly4 examined QOL scores before and after adenotonsillectomy for children with mild sleepdisordered breathing (defined as an AHI \5) and children with OSA (defined as an AHI .5). The authors reported poor QOL in both groups; there were no significant differences in preoperative QOL scores between the mild sleep-
Downloaded from oto.sagepub.com at NORTH CAROLINA STATE UNIV on March 18, 2015
Volsky et al
131
disordered breathing and OSA groups. Our results also showed a significant impact of sleep disturbance on QOL despite the subjects having only mild OSA by PSG criteria. Why do children with mild OSA have such poor quality of life? While AHI is the main parameter on PSG used to determine OSA severity in children, perhaps other parameters, such as the arousal index, better correlate with QOL scores and reflect the disease burden of OSA. Quality-oflife studies demonstrate that the impact of OSA extends beyond sleep to manifest itself in multiple aspects of a child’s life, including daytime functioning, behavior, and family interactions. A complex relationship exists between pediatric OSA, behavior, and QOL. Regardless, children with mild OSA often report poor QOL and can be very symptomatic. Such children deserve consideration for treatment. To the authors’ knowledge, no prior studies have assessed the impact of observation vs adenotonsillectomy on QOL in children with mild OSA. A change in the OSA18 total score from 80 to 60 indicates a change from severe to moderate impact on QOL.9 In our study, baseline OSA18 scores for the surgery group indicated a moderate impact of sleep disturbance on QOL, while those in the observation group had OSA-18 scores in the mild range. Following adenotonsillectomy, there were significant improvements in QOL scores in the surgery group, while the observation group scores remained stable at the early follow-up visit. To this end, the OSA-18 may be a useful tool to help providers decide which children with mild OSA may benefit from early intervention. According to our findings, children with mild OSA who have OSA-18 scores in the moderate range (60-80) experience significant improvement in QOL following surgery and are good candidates for adenotonsillectomy. Our findings also demonstrate that some subjects with mild OSA will experience improvements in QOL scores without any intervention. Thus, observation with close follow-up is an option for children with mild OSA with total OSA-18 scores less than 60. The natural history of mild sleep apnea is still being investigated. Ng et al8 studied the natural history of mild OSA in 45 children. All subjects had a repeat PSG 2 years after the initial diagnosis of mild OSA. Disease progression, as evidenced by an increasing AHI, was noted in 29% (n = 13) of children. Progression of OSA in the study by Ng et al8 was determined by PSG alone; QOL was not assessed in this study. Eighteen percent of children (n = 6) in our study developed increasing obstructive symptoms and failed observation. This failed observation group had a baseline OSA-18 total score similar to that of the surgery group, suggesting that these children had poorer QOL. This difference did not reach statistical significance, but this may have been due to small sample size. Interestingly, many children in the observation group demonstrated improvements in QOL at the late follow-up interval. The reason for such improvements is unclear. Recent research has demonstrated an increased
inflammatory response, including alterations in T-cell lymphocyte populations, in children with OSA.17 Perhaps over time, the body is able to mount an immune response to the initial inflammatory insult, which translates into clinical improvement in QOL. Other potential explanations for QOL improvements include growth of the airway and diminished lymphoid tissue. Further research is necessary to determine which children with mild OSA will improve with observation and which children will experience progression of disease. To our knowledge, this is the first study to compare QOL outcomes for adenotonsillectomy vs observation in children with mild OSA. Strengths of the study include utilization of full-night PSG and the use of validated instruments to assess QOL. Limitations include possible selection bias due to lack of randomization of subjects and the possible placebo effect of surgery on QOL scores. The authors had significant difficulty recruiting patients for randomization because most caregivers had a strong preference for either observation or surgery when the options were presented. In addition, a substantial number of subjects were withdrawn for failure to follow up. This may be due in part to our location in an area with a large military population where transfers are common. In addition, medical management such as montelukast was not offered to patients enrolled in this study. Such medications should be considered in future treatment protocols for mild OSA.
Conclusion Mild OSA can have a substantial impact on QOL in pediatric patients. Quality of life significantly improves following adenotonsillectomy. There is also a subset of patients with mild OSA in whom QOL scores improve after an observation period. Thus, observation with close follow-up is another treatment option. Acknowledgment Mrs Kelly Sylvester assisted with collection of surveys.
Author Contributions Peter G. Volsky, study design, recruitment of subjects, data management, interpretation of data, drafting, critical revision, and final approval of manuscript; Meghan A. Woughter, collection and interpretation of data, drafting and final approval of manuscript; Hind A. Beydoun, study design, data management and interpretation, statistical analysis, critical revision and final approval of manuscript; Craig S. Derkay, recruitment of subjects/data collection, interpretation of results, critical revision and final approval of manuscript; Cristina M. Baldassari, study design, recruitment of subjects, interpretation of data, critical revision and final approval of manuscript.
Disclosures Competing interests: None. Sponsorships: None. Funding source: None.
Downloaded from oto.sagepub.com at NORTH CAROLINA STATE UNIV on March 18, 2015
132
Otolaryngology–Head and Neck Surgery 150(1)
References 1. Katz ES, D’Ambrosio CM.Pediatric obstructive sleep apnea syndrome. Clin Chest Med. 2010;31:221-234. 2. Ray R, Bower C.Pediatric obstructive sleep apnea: the year in review. Curr Opin Otolaryngol Head Neck Surg. 2005;13: 360-365. 3. Wagner MH, Torrez DM.Interpretation of the polysomnogram in children. Otolaryngol Clin North Am. 2007;40:745-759. 4. Mitchell RB, Kelly J.Quality of life after adenotonsillectomy for SDB in children. Otolaryngol Head Neck Surg. 2005;133: 569-572. 5. Baldassari CM, Mitchell RB, Schubert C, Rudnick EF. Pediatric obstructive sleep apnea and quality of life: a metaanalysis. Otolaryngol Head Neck Surg. 2008;138:265-273. 6. Darrow DH.Surgery for pediatric sleep apnea. Otolaryngol Clin North Am. 2007;40:855-875. 7. Li AM, Au CT, Ng SK, et al. Natural history and predictors for progression of mild childhood obstructive sleep apnoea. Thorax. 2010;65:27-31. 8. Ng SK, Lee DL, Li AM, et al. Reproducibility of clinical grading of tonsillar size. Arch Otolaryngol Head Neck Surg. 2010; 136:159-162. 9. Franco RA, Rosenfeld RM, Rao M.Quality of life for children with obstructive sleep apnea. Otolaryngol Head Neck Surg. 2000;123:9-16.
10. HealthActCHQ. The CHQ Scoring and Interpretation Manual. Cambridge, MA: HealthActCHQ; 2008. 11. Iber CA, Chesso A, Quan SF. The AASM Manual for the Scoring of Sleep and Associated Events: Rules, Terminology and Technical Specifications. Darien, IL: American Academy of Sleep Medicine; 2007. 12. US Centers for Disease Control and Prevention. Children’s BMI tool for schools. www.cdc.gov/healthyweight/assessing/ bmi/childrens_BMI/tool_for_schools.html. Accessed January 10, 2013. 13. Mitchell RB, Kelly J.Behavioral changes in children with mild sleep-disordered breathing or obstructive sleep apnea after adenotonsillectomy. Laryngoscope. 2007;117:1685-1688. 14. Calhoun SL, Mayes SD, Vgontzas AN, Tsaoussoglou M, Shifflett LJ, Bixler EO.No relationship between neurocognitive functioning and mild sleep disordered breathing in a community sample of children. J Clin Sleep Med. 2009;5:228-234. 15. Kennedy JD, Blunden S, Hirte C, et al. Reduced neurocognition in children who snore. Pediatr Pulmonol. 2004;37:330-337. 16. O’Brien LM, Mervis CB, Holbrook CR, et al. Neurobehavioral implications of habitual snoring in children. Pediatrics. 2004; 114:44-49. 17. Tan H, Gozal D, Wany Y, et al. Alterations in circulating Tcell lymphocyte populations in children with obstructive sleep apnea. Sleep. 2013;36:913-922.
Downloaded from oto.sagepub.com at NORTH CAROLINA STATE UNIV on March 18, 2015
Adenotonsillectomy vs Observation for Management of Mild Obstructive Sleep Apnea in Children
Otolaryngology– Head and Neck Surgery 2014, Vol 150(1) 126–132 Ó American Academy of Otolaryngology—Head and Neck Surgery Foundation 2013 Reprints and permission: sagepub.com/journalsPermissions.nav DOI: 10.1177/0194599813509780 http://otojournal.org
Peter G. Volsky, MD1, Meghan A. Woughter2, Hind A. Beydoun, PhD3, Craig S. Derkay, MD1,4, and Cristina M. Baldassari, MD1,4
No sponsorships or competing interests have been disclosed for this article.
Abstract Objective. To determine the impact of adenotonsillectomy vs observation on quality of life (QOL) in children with mild obstructive sleep apnea (OSA). Study Design. Prospective, nonrandomized trial. Setting. Tertiary children’s hospital. Subjects and Methods. Sixty-four children (ages 3-16 years) with mild OSA (apnea hypopnea index between 1 and 5 on polysomnogram) completed the study. Caregivers chose between management options of adenotonsillectomy and observation and completed validated QOL instruments (OSA18 and Children’s Health Questionnaire) at baseline, early, and late follow-ups. The primary outcome measure was QOL. Results. Thirty patients chose adenotonsillectomy, while 34 were observed. Total OSA-18 scores at baseline were significantly poorer (P = .01) in the surgery group (72.3) compared with the observation group (58.5). Four months following surgery, OSA-18 scores improved by 39.1 points over baseline (P = .0001), while there was no change for the observation group (P = .69). After 8 months, OSA-18 scores remained improved in the surgery group, and observation group scores improved by 13.4 points over baseline (P = .005). While OSA-18 scores at the late follow-up visit were poorer in the observation group, the difference was not statistically significant (P = .05). Six observation patients opted for adenotonsillectomy during the study. Conclusion. Quality of life significantly improves in children with mild OSA after adenotonsillectomy. In children with mild OSA who are observed, QOL improvements at early follow-up are less pronounced, but significant improvements in QOL are evident after 8 months. QOL instruments may be useful tools to help providers determine which children with mild OSA may benefit from early intervention. Keywords sleep apnea, obstructive, tonsillectomy, adenoidectomy, child, quality of life
Received June 7, 2013; revised September 27, 2013; accepted October 1, 2013.
O
bstructive sleep apnea (OSA) is a sleep-related breathing disorder that is characterized by intermittent episodes of upper airway collapse during sleep. It is a public health concern since it affects 2% to 3% of preschool-age children in the United States.1 Pediatric OSA has been linked to metabolic changes, growth inhibition, and cardiovascular sequelae. Furthermore, there is a growing body of literature demonstrating the negative impact of OSA on quality of life (QOL) and cognitive function in children. Pediatric OSA has been associated with behavior problems, poor attention, memory and cognitive deficits, and poor school performance.2 The gold standard for diagnosis of pediatric OSA is a full-night polysomnogram (PSG). The severity of OSA is categorized according to the obstructive apnea hypopnea index (AHI) on the PSG. According to the most commonly used system, an AHI between 1 and 5 indicates mild OSA.3 Although PSG reliably measures the presence of OSA and provides an objective scale for OSA severity, it fails to quantify the impact of OSA on a child’s general well-being, including emotional and behavioral health. Interestingly, QOL does not correlate with AHI in pediatric patients.4
1
Department of Otolaryngology–Head and Neck Surgery, Eastern Virginia Medical School, Norfolk, Virginia, USA 2 Eastern Virginia Medical School, Norfolk, Virginia, USA 3 Graduate Program in Public Health, Eastern Virginia Medical School, Norfolk, Virginia, USA 4 Department of Pediatric Otolaryngology, Children’s Hospital of the King’s Daughters, Norfolk, Virginia, USA This research was presented at the American Society of Pediatric Otolaryngology (ASPO) Annual Meeting; April 26-28, 2013; Alexandria, Virginia. Corresponding Author: Cristina M. Baldassari, MD, Department of Otolaryngology–Head and Neck Surgery, Eastern Virginia Medical School, 601 Children’s Lane, 2nd Floor, Norfolk, VA 23507, USA. Email: [email protected]
Downloaded from oto.sagepub.com at NORTH CAROLINA STATE UNIV on March 18, 2015
Volsky et al
127
Quality of life is increasingly recognized as an important health outcome measure in clinical medicine. A recent meta-analysis5 demonstrated poor QOL scores in children with OSA and significant improvements in QOL following adenotonsillectomy, irrespective of AHI. The authors concluded that the impact of OSA affects children’s behavior, daytime functioning, and family life.5 Thus, many children with mild OSA potentially warrant treatment. In children, adenotonsillar hyperplasia is the most common cause of upper airway obstruction. Thus, the primary treatment for pediatric OSA is adenotonsillectomy.6 Numerous studies have demonstrated improvements in neuropsychological outcomes such as behavior, school performance, and attention following surgery for pediatric OSA.1 There is also a growing body of literature highlighting improvements in QOL following adenotonsillectomy.5 Management of children with mild OSA, however, remains controversial, since the natural history of mild OSA is not well understood.7 Thus, it is challenging to counsel caregivers about how best to treat mild OSA. Our primary objective was to determine the impact of adenotonsillectomy vs observation on QOL in children with mild OSA.
Methods A prospective trial was designed to assess the impact of adenotonsillectomy vs observation on QOL in children with mild OSA. The protocol was approved by the Eastern Virginia Medical School Institutional Review Board and registered with the National Institutes of Health website, clinicaltrials.gov. The study was carried out from March 2011 to January 2013. Subjects were recruited from the outpatient pediatric otolaryngology clinic at The Children’s Hospital of the King’s Daughters, a tertiary care children’s hospital in Norfolk, Virginia. Inclusion criteria were (1) children 3 to 16 years old; (2) new diagnosis of mild OSA, defined by AHI from 1 to 5 on full-night PSG; and (3) tonsillar hypertrophy on physical examination. The Brodsky grading scale,8 which is based on the ratio of tonsil size to the tonsillar pillar, was used to assess for tonsillar hypertrophy. Children with grade 2, 3, and 4 tonsils were considered to have tonsillar hypertrophy. Subjects with craniofacial abnormalities, cerebral palsy, Trisomy 21, or prior adenotonsillar surgery were excluded. Children with parents who could not speak English were also excluded. Two validated instruments, the Obstructive Sleep Apnea 18 (OSA-18) and the Child Health Questionnaire Parent Form 28 (CHQ PF-28), were used to assess QOL. The OSA-18, developed by Franco et al,9 is an 18-item questionnaire specific to OSA that measures the relative severity of sleep-related problems with a Likert scale ranging from 1 (none of the time) to 7 (all of the time). The OSA-18 instrument yields a total score and 5 domain scores, including scores for sleep disturbance, physical suffering, emotional distress, daytime problems, and caregiver concerns. Total OSA-18 QOL scores between 60 and 80 indicate a moderate QOL impairment.9 Higher OSA-18 scores are associated
with poorer QOL. The CHQ PF-28 (HealthActCHQ, Inc, Cambridge, Massachusetts) is a measure of a child’s overall or global QOL for which normative data are available for healthy children.10 The CHQ PF-28 is composed of 28 questions. The CHQ PF-28 scores are reported as 2 subsets, Physical Functioning and Psychosocial Functioning, with higher scores indicating better QOL. Children with newly diagnosed mild OSA on PSG were recruited. The authors initially had difficulty recruiting patients for randomization, so a nonrandomized arm was added. Consequently, parents could opt out of randomization by choosing to enroll in either the adenotonsillectomy (surgery) group or the observation group. Medical management with intranasal steroids and montelukast was not offered to patients enrolled in this study. Parents, with input from the child, completed both the OSA-18 and CHQ PF-28 at baseline and at 2 follow-up visits. The target intervals for follow-up assessments were 3 and 6 months following enrollment in the observation group and 3 and 6 months following adenotonsillectomy in the surgery group. At least 1 attempt was made by phone to contact patients who failed to follow up. Subjects enrolled in the observation group with symptomatic exacerbation were able to drop out of this group and undergo adenotonsillectomy after 3 months. Such children were labeled the ‘‘failed observation’’ group. We performed an intention-to-treat analysis in which these data were analyzed as part of the observation group. All PSGs were performed in dedicated pediatric sleep laboratories in accordance with the American Academy of Sleep Medicine (AASM) guidelines11 and were scored by pediatric sleep medicine specialists. Standard procedure included a 6-lead electroencephalogram (EEG), bilateral electrooculogram (EOG) leads, and 1 submental and 2 tibial electromyograms (EMGs). Respiratory measurements were recorded using inductance pneumography, airflow using a nasal pressure transducer, oxygen saturation by pulse oximetry, and carbon dioxide using a carbon dioxide sensor. Information obtained from each PSG included sleep efficiency, number and classification of arousals and respiratory events, and oxygen and carbon dioxide levels. Demographic data, including age, sex, ethnicity, and body mass index (BMI) percentiles (calculated according to the US Centers for Disease Control and Prevention guidelines),12 were also recorded. A power analysis was conducted. A group size of 37 was deemed sufficient to detect a 20-point difference in OSA-18 QOL scores using a 2-group t test with 80% power and a 0.05 two-tailed significance level. It was anticipated that subjects would enroll in the surgery and observation groups on a 1:1 ratio, so a total of 74 subjects were needed. Due to anticipated subject withdrawal, the authors planned to enroll at least 150 patients. Descriptive statistics, including means and standard deviation for continuous variables or frequencies and percentages for categorical variables, were used to evaluate baseline characteristics of observation and surgery groups. The x2 test was used to compare the distribution of categorical variables across groups, while independent
Downloaded from oto.sagepub.com at NORTH CAROLINA STATE UNIV on March 18, 2015
128
Otolaryngology–Head and Neck Surgery 150(1)
Table 1. Subject demographic and polysomnogram data. Subjects Completing the Study
Sex Male Female Tonsil grade 2 3 4 Age, mo BMI, kg/m2 BMI percentile Total AHI O2 nadir Arousal index Peak end-tidal CO2 Baseline OSA-18 score Baseline CHQ PF-28 Physical Functioning Psychosocial Functioning
Subjects Withdrawn
Surgery Group (n = 30)
Observation Group (n = 34)
16 14
20 14
10 15 4 80.8 6 33.7 18.9 6 6.1 68.9 6 28.6 3.3 6 1.0 91.2 6 2.4 11.8 6 3.8 50.1 6 5.9 72.3 6 20.0
10 19 5 80.0 6 41.9 20.2 6 6.2 74.0 6 25.5 2.9 6 1.1 90.8 6 4.1 10.2 6 3.4 48.9 6 4.2 58.5 6 21.5
44.9 6 15.4 39.8 6 13.9
50.6 6 7.8 46.3 6 11.8
X-Surgery Group (n = 29)
X-Observation Group (n = 20)
9 20
9 11
.93 .41 .45 .22 .65 .091 .39 .01
6 15 7 69.7 6 25.8 18.4 6 4.1 67.1 6 34.9 3.1 6 1.2 90.4 6 2.5 11.4 6 3.7 48.2 6 5.2 72.6 6 16.3
5 7 5 97.4 6 43.3 18.6 6 6.2 61.6 6 33.7 2.4 6 1.0 91.9 6 1.6 9.6 6 2.5 45.3 6 5.7 54.6 6 20.7
.07 .05
47.0 6 11.7 36.5 6 14.4
53.4 6 8.5 47.5 6 12.2
P Value .66
.72
P Value, Surgery vs X-Surgery
P Value, Observation vs X-Observation
.22
.78
.43
.30
.16 .72 .82 .46 .20 .67 .27 .93
.15 .39 .14 .07 .16 .46 .03 .52
.56 .38
.23 .74
Abbreviations: AHI, apnea hypopnea index; BMI, body mass index; CHQ PF-28, Child Health Questionnaire Parent Form 28; CO2, carbon dioxide.
samples t tests were used for continuous variables. The OSA-18 and CHQ PF-28 scores were compared across groups using independent samples t test at baseline, early follow-up, and late follow-up. Wilcoxon’s rank sum test was used to assess for differences in OSA-18 and CHQ PF28 scores within the surgery and observation groups between baseline, early follow-up, and late follow-up. Finally, a paired t test was used to examine OSA-18 and CHQ PF-28 score changes within each group. Values of P \ .05 were considered significant.
Results Subjects One hundred eleven subjects agreed to participate and were enrolled in either the surgery or observation group based on parent and child preference. An additional 2 patients agreed to randomization to the surgery or observation group. Twenty-nine subjects were withdrawn due to failure to complete follow-up QOL surveys, and 3 subjects elected for voluntarily withdrawal. An additional 17 patients were deactivated at the study termination before they returned for follow-up. Thus, 64 children were included in the final analysis. Thirty subjects comprised the surgery group, while 34 subjects comprised the observation group. There were no differences between the 2 groups in terms of demographic and polysomnographic data (Table 1). The duration of time between enrollment and early follow-up was an average of 113 6 29 days (approximately
4 months). The late follow-up survey was completed 232 6 51 days (approximately 8 months) after enrollment in the observation group or following adenotonsillectomy in the surgery group. Twenty-five percent (n = 16) of subjects completed both early and late QOL surveys.
OSA-18 QOL Scores Total OSA-18 scores at baseline were significantly poorer (P = .01) in the surgery group (72.3 6 20.0) compared with the observation group (58.5 6 21.5). Between the 2 groups, there were also differences in subscores of Sleep Disturbance, Daytime Function, and Caregiver Concerns (Table 2). The mean OSA-18 total score for the surgery group significantly improved (P = .0001) from baseline to the early follow-up by 39.1 6 18.8 points. Significant gains in QOL were also evident for the 5 OSA-18 domain scores following adenotonsillectomy (Table 2). At the early follow-up visit, the total OSA-18 score for the observation group did not significantly change (P = .69) from the initial baseline score (Table 2). When the 2 groups were compared at the early follow-up interval, the surgery group had significantly (P = .0001) better OSA-18 QOL scores than the observation group. Improvements noted in OSA-18 QOL scores following adenotonsillectomy were maintained at the late follow-up (Table 2). Interestingly, in the observation group, there was also a significant improvement (P = .005) in the total
Downloaded from oto.sagepub.com at NORTH CAROLINA STATE UNIV on March 18, 2015
Volsky et al
129
Table 2. OSA-18 QOL scores for surgery and observation groups.a Surgery Group (S)
Observation Group (O)
Pb
Pc
Pc
Baseline (T0)
(S0, n = 30)
(O0, n = 34)
S0 vs O0
Sleep Disturbances Physical Symptoms Emotional Symptoms Daytime Function Caregiver Concerns Total
17.3 6 4.9 13.8 6 6.0 13.4 6 5.9 13.0 6 4.7 14.7 6 6.8 72.3 6 20.0
14.3 6 4.8 11.9 6 4.8 11.0 6 5.6 9.8 6 4.6 10.1 6 5.8 58.5 6 21.5
.019d .17 .09 .008d .006d .01d
Early Follow-up (T1)
(S1, n = 24)
(O1, n = 25)
S1 vs O1
S1 vs S0
O1 vs O0
Sleep Disturbances Physical Symptoms Emotional Symptoms Daytime Function Caregiver Concerns Total
6.3 6 2.8 6.8 6 3.2 8.4 6 4.5 5.8 6 2.9 6.6 6 3.7 33.9 6 14.6
14.6 6 6.4 12.0 6 5.6 10.9 6 5.8 9.4 6 5.0 11.2 6 5.9 58.2 6 24.5
\.0001d .0003d .09d .004d .002d .0001d
\.0001d \.0001d \.0001d \.0001d \.0001d \.0001d
Late Follow-up (T2)
(S2, n = 11)
(O2, n = 18)
S2 vs O2
S2 vs S0
Sleep Disturbances Physical Symptoms Emotional Symptoms Daytime Function Caregiver Concerns Total
6.7 6 3.2 6.5 6 2.3 8.3 6 4.3 6.3 6 3.2 5.8 6 2.7 33.6 6 8.6
12.0 6 5.6 9.7 6 6.0 8.6 6 6.8 7.1 6 4.1 8.2 6 4.8 45.1 6 21.9
.003d .04d .85 .58 .08 .05
.0002d .01d .01d .003d .0007d .0002d
.60 .40 .81 .86 .19 .69 O2 vs O0 .004d .52 .03d .01d .31 .005d
Abbreviation: QOL, quality of life. a Higher scores indicate poorer QOL. Early follow-up (T1): 113 6 29 days. Late follow-up (T2): 232 6 51 days. b Independent samples t test. c Paired t test. d P \.05 is statistically significant.
CHQ PF-28 QOL Scores At baseline, the Physical and Psychosocial Functioning total scores were similar between the surgery and observation groups (P = .05; Table 3). Among the CHQ PF-28 subscores, Physical Functioning (P = .011) and Role/Social Limitations—Physical (P = .025) were poorer at baseline in the surgery group. After adenotonsillectomy, improvement occurred in the Physical and Psychosocial Functioning scores of the surgery group (Table 3). Also, 11 subscales improved from baseline to the early follow-up in the surgery group, including Global Health (P \ .0001), Physical Functioning (P = .005), Role/Social Limitations—Emotional/Behavioral
120
OSA-18 Total Score
OSA-18 QOL score, with a 13-point change between the baseline and late follow-up visits (Figure 1). Three domain scores—Sleep Disturbances, Emotional Symptoms, and Daytime Function—also improved in the observation group during that time. Total OSA-18 scores at the late follow-up visit were poorer in the observation group compared with the surgery group. However, this difference was not statistically significant (P = .05).
100 80
72.3 (64.9, 79.7) 58.2 (48.6, 67.8)
60
45.1 (35.2, 54.9)
58.5 (51.3, 65.7)
40 33.9 (28.0, 39.7)
20 0
Baseline
Early follow-up
33.6 (28.7, 38.5)
Late follow-up
Time Interval Surgery Group
Observation Group
Figure 1. Total OSA-18 quality-of-life (QOL) scores. Total OSA18 QOL scores with 95% confidence intervals are depicted at baseline, early follow-up (113 6 29 days), and late follow-up (232 6 51 days) for both the surgery and observation groups. Higher OSA-18 Total Scores indicate poorer QOL.
Downloaded from oto.sagepub.com at NORTH CAROLINA STATE UNIV on March 18, 2015
130
Otolaryngology–Head and Neck Surgery 150(1)
Table 3. CHQ PF-28 QOL scores for surgery and observation groups.a Surgery Group (S)
Observation Group (O)
Pb
Baseline (T0)
(S0, n = 30)
(O0, n = 34)
S0 vs O0
Total—Physical Total—Psychosocial
44.9 6 15.4 39.8 6 13.9
50.6 6 7.8 46.3 6 11.8
.07 .05
Early Follow-up (T1)
(S1, n = 24)
(O1, n = 25)
S1 vs O1
Pc
Pc
S1 vs S0
O1 vs O0
d
Total—Physical Total—Psychosocial
52.3 6 9.4 50.1 6 12.7
47.2 6 12.8 47.2 6 10.1
.13 .40
.02 .003d
.66 .24
Late Follow-up (T2)
(S2, n = 11)
(O2, n = 18)
S2 vs O2
S2 vs S0
O2 vs O0
Total—Physical Total—Psychosocial
50.3 6 6.2 50.6 6 8.7
49.9 6 12.8 52.0 6 10.9
.93 .74
.02d .28
.89 .02d
Abbreviations: CHQ PF-28, Child Health Questionnaire Parent Form 28; QOL, quality of life. a Lower scores indicate poorer QOL. Early follow-up (T1): 113 6 29 days. Late follow-up (T2): 232 6 51 days. b Independent samples t test. c Paired t test. d P \.05 is statistically significant.
(P = .004), Behavior (P = .001), Global Behavior Item (P = .007), Mental Health (P = .009), Self-Esteem (P = .01), General Health Perceptions (P = .01), Change in Health (P \ .0001), Parental Impact—Emotional (P = .0002), and Family Activities (P = .002). In the observation group, there was no change in the total CHQ PF-28 scores between baseline and early follow-up, while the subscale of Family Cohesion (P = .02) actually deteriorated. When comparing the 2 groups at early follow-up, total scores for Physical and Psychosocial Functioning were similar (Table 3). However, the surgery group did have better QOL scores in the 4 subscales of Global Health (P = .033), Physical Functioning (P = .037), Behavior (P = .02), and Change in Health (P \ .0001). From baseline to the late follow-up, the surgery group maintained improvements in CHQ PF-28 total Physical and Psychosocial Functioning scores (Table 3). In the observation group, the Psychosocial Functioning score was improved (P = .02) from baseline, but there was not a significant change noted in the Physical Functioning score (Table 3). At the time of late follow-up, there was no difference in Physical and Psychosocial Functioning between groups (Table 3).
Failed Observation Group Six subjects initially enrolled in the observation group experienced an exacerbation of OSA symptoms during the study, and their caregivers elected to proceed with adenotonsillectomy. These children were labeled the ‘‘failed observation’’ group. In 5 children, exacerbation of OSA symptoms occurred in the first 6 months of enrollment, while 1 subject deteriorated after 6 months. The baseline demographic and polysomnographic variables for the failed observation group and the observation group were similar (P . .05). Interestingly, the mean baseline OSA-18 total score for the
failed observation group was 73.0 6 24.9, which was poorer than the mean total score for the observation group, 55.4 6 19.8. This difference, however, was not significant (P = .07).
Discussion A recently published meta-analysis highlighted the impact of OSA on QOL in children.5 The authors showed that pediatric patients with OSA have a poorer health status than healthy children. While full-night PSG remains the gold standard for determining the severity of OSA in children, numerous studies have shown a poor correlation between AHI and QOL scores.5 Furthermore, behavioral problems associated with OSA were not related to the severity of OSA as defined by the AHI on PSG.13 Thus, QOL assessments are being increasingly recognized as an important part of the evaluation of patients with OSA. In addition to being used to study the disease burden of OSA, QOL assessments are also being used to study outcomes of treatment, including adenotonsillectomy. The appropriate management for mild OSA in children continues to be debated. Some providers advocate for watchful waiting. This conservative approach is affirmed by a study by Calhoun et al14 that demonstrated that children with mild OSA have neurocognitive functioning equivalent to control patients. Proponents for more aggressive treatment cite recent studies that demonstrate neurocognitive deficits in children with only mild airway obstruction, including primary snoring.15,16 Mitchell and Kelly4 examined QOL scores before and after adenotonsillectomy for children with mild sleepdisordered breathing (defined as an AHI \5) and children with OSA (defined as an AHI .5). The authors reported poor QOL in both groups; there were no significant differences in preoperative QOL scores between the mild sleep-
Downloaded from oto.sagepub.com at NORTH CAROLINA STATE UNIV on March 18, 2015
Volsky et al
131
disordered breathing and OSA groups. Our results also showed a significant impact of sleep disturbance on QOL despite the subjects having only mild OSA by PSG criteria. Why do children with mild OSA have such poor quality of life? While AHI is the main parameter on PSG used to determine OSA severity in children, perhaps other parameters, such as the arousal index, better correlate with QOL scores and reflect the disease burden of OSA. Quality-oflife studies demonstrate that the impact of OSA extends beyond sleep to manifest itself in multiple aspects of a child’s life, including daytime functioning, behavior, and family interactions. A complex relationship exists between pediatric OSA, behavior, and QOL. Regardless, children with mild OSA often report poor QOL and can be very symptomatic. Such children deserve consideration for treatment. To the authors’ knowledge, no prior studies have assessed the impact of observation vs adenotonsillectomy on QOL in children with mild OSA. A change in the OSA18 total score from 80 to 60 indicates a change from severe to moderate impact on QOL.9 In our study, baseline OSA18 scores for the surgery group indicated a moderate impact of sleep disturbance on QOL, while those in the observation group had OSA-18 scores in the mild range. Following adenotonsillectomy, there were significant improvements in QOL scores in the surgery group, while the observation group scores remained stable at the early follow-up visit. To this end, the OSA-18 may be a useful tool to help providers decide which children with mild OSA may benefit from early intervention. According to our findings, children with mild OSA who have OSA-18 scores in the moderate range (60-80) experience significant improvement in QOL following surgery and are good candidates for adenotonsillectomy. Our findings also demonstrate that some subjects with mild OSA will experience improvements in QOL scores without any intervention. Thus, observation with close follow-up is an option for children with mild OSA with total OSA-18 scores less than 60. The natural history of mild sleep apnea is still being investigated. Ng et al8 studied the natural history of mild OSA in 45 children. All subjects had a repeat PSG 2 years after the initial diagnosis of mild OSA. Disease progression, as evidenced by an increasing AHI, was noted in 29% (n = 13) of children. Progression of OSA in the study by Ng et al8 was determined by PSG alone; QOL was not assessed in this study. Eighteen percent of children (n = 6) in our study developed increasing obstructive symptoms and failed observation. This failed observation group had a baseline OSA-18 total score similar to that of the surgery group, suggesting that these children had poorer QOL. This difference did not reach statistical significance, but this may have been due to small sample size. Interestingly, many children in the observation group demonstrated improvements in QOL at the late follow-up interval. The reason for such improvements is unclear. Recent research has demonstrated an increased
inflammatory response, including alterations in T-cell lymphocyte populations, in children with OSA.17 Perhaps over time, the body is able to mount an immune response to the initial inflammatory insult, which translates into clinical improvement in QOL. Other potential explanations for QOL improvements include growth of the airway and diminished lymphoid tissue. Further research is necessary to determine which children with mild OSA will improve with observation and which children will experience progression of disease. To our knowledge, this is the first study to compare QOL outcomes for adenotonsillectomy vs observation in children with mild OSA. Strengths of the study include utilization of full-night PSG and the use of validated instruments to assess QOL. Limitations include possible selection bias due to lack of randomization of subjects and the possible placebo effect of surgery on QOL scores. The authors had significant difficulty recruiting patients for randomization because most caregivers had a strong preference for either observation or surgery when the options were presented. In addition, a substantial number of subjects were withdrawn for failure to follow up. This may be due in part to our location in an area with a large military population where transfers are common. In addition, medical management such as montelukast was not offered to patients enrolled in this study. Such medications should be considered in future treatment protocols for mild OSA.
Conclusion Mild OSA can have a substantial impact on QOL in pediatric patients. Quality of life significantly improves following adenotonsillectomy. There is also a subset of patients with mild OSA in whom QOL scores improve after an observation period. Thus, observation with close follow-up is another treatment option. Acknowledgment Mrs Kelly Sylvester assisted with collection of surveys.
Author Contributions Peter G. Volsky, study design, recruitment of subjects, data management, interpretation of data, drafting, critical revision, and final approval of manuscript; Meghan A. Woughter, collection and interpretation of data, drafting and final approval of manuscript; Hind A. Beydoun, study design, data management and interpretation, statistical analysis, critical revision and final approval of manuscript; Craig S. Derkay, recruitment of subjects/data collection, interpretation of results, critical revision and final approval of manuscript; Cristina M. Baldassari, study design, recruitment of subjects, interpretation of data, critical revision and final approval of manuscript.
Disclosures Competing interests: None. Sponsorships: None. Funding source: None.
Downloaded from oto.sagepub.com at NORTH CAROLINA STATE UNIV on March 18, 2015
132
Otolaryngology–Head and Neck Surgery 150(1)
References 1. Katz ES, D’Ambrosio CM.Pediatric obstructive sleep apnea syndrome. Clin Chest Med. 2010;31:221-234. 2. Ray R, Bower C.Pediatric obstructive sleep apnea: the year in review. Curr Opin Otolaryngol Head Neck Surg. 2005;13: 360-365. 3. Wagner MH, Torrez DM.Interpretation of the polysomnogram in children. Otolaryngol Clin North Am. 2007;40:745-759. 4. Mitchell RB, Kelly J.Quality of life after adenotonsillectomy for SDB in children. Otolaryngol Head Neck Surg. 2005;133: 569-572. 5. Baldassari CM, Mitchell RB, Schubert C, Rudnick EF. Pediatric obstructive sleep apnea and quality of life: a metaanalysis. Otolaryngol Head Neck Surg. 2008;138:265-273. 6. Darrow DH.Surgery for pediatric sleep apnea. Otolaryngol Clin North Am. 2007;40:855-875. 7. Li AM, Au CT, Ng SK, et al. Natural history and predictors for progression of mild childhood obstructive sleep apnoea. Thorax. 2010;65:27-31. 8. Ng SK, Lee DL, Li AM, et al. Reproducibility of clinical grading of tonsillar size. Arch Otolaryngol Head Neck Surg. 2010; 136:159-162. 9. Franco RA, Rosenfeld RM, Rao M.Quality of life for children with obstructive sleep apnea. Otolaryngol Head Neck Surg. 2000;123:9-16.
10. HealthActCHQ. The CHQ Scoring and Interpretation Manual. Cambridge, MA: HealthActCHQ; 2008. 11. Iber CA, Chesso A, Quan SF. The AASM Manual for the Scoring of Sleep and Associated Events: Rules, Terminology and Technical Specifications. Darien, IL: American Academy of Sleep Medicine; 2007. 12. US Centers for Disease Control and Prevention. Children’s BMI tool for schools. www.cdc.gov/healthyweight/assessing/ bmi/childrens_BMI/tool_for_schools.html. Accessed January 10, 2013. 13. Mitchell RB, Kelly J.Behavioral changes in children with mild sleep-disordered breathing or obstructive sleep apnea after adenotonsillectomy. Laryngoscope. 2007;117:1685-1688. 14. Calhoun SL, Mayes SD, Vgontzas AN, Tsaoussoglou M, Shifflett LJ, Bixler EO.No relationship between neurocognitive functioning and mild sleep disordered breathing in a community sample of children. J Clin Sleep Med. 2009;5:228-234. 15. Kennedy JD, Blunden S, Hirte C, et al. Reduced neurocognition in children who snore. Pediatr Pulmonol. 2004;37:330-337. 16. O’Brien LM, Mervis CB, Holbrook CR, et al. Neurobehavioral implications of habitual snoring in children. Pediatrics. 2004; 114:44-49. 17. Tan H, Gozal D, Wany Y, et al. Alterations in circulating Tcell lymphocyte populations in children with obstructive sleep apnea. Sleep. 2013;36:913-922.
Downloaded from oto.sagepub.com at NORTH CAROLINA STATE UNIV on March 18, 2015
