pediatrics
February 2015, VOLUME135 /ISSUE 2

# Quality of Life and Obstructive Sleep Apnea Symptoms After Pediatric Adenotonsillectomy

1. Susan L. Garetz, MDa,
2. Ron B. Mitchell, MDb,
3. Portia D. Parker, MSc,
4. Reneé H. Moore, PhDd,
5. Carol L. Rosen, MDe,
6. Bruno Giordani, PhDf,
7. Hiren Muzumdar, MDg,
8. Shalini Paruthi, MDh,
9. Lisa Elden, MDi,
10. Paul Willging, MDj,
11. Dean W. Beebe, PhDk,
12. Carole L. Marcus, MBBChl,
13. Ronald D. Chervin, MD, MSm, and
14. Susan Redline, MD, MPHn
2. fDepartments of Psychiatry and Psychology and Sleep Disorders Center, and
3. mDepartment of Neurology and Sleep Disorders Center, University of Michigan Health Center, Ann Arbor, Michigan;
4. bDepartments of Otolaryngology and Pediatrics, Utah Southwestern and Children’s Medical Center, Dallas, Texas;
5. cSAS Institute Inc, Cary, North Carolina;
6. dDepartment of Statistics, North Carolina State University, Raleigh, North Carolina;
7. eDepartment of Pediatrics, Rainbow Babies & Children’s Hospital, University Hospitals Case Medical Center, Case Western Reserve University School of Medicine, Cleveland, Ohio;
8. gDivision of Pulmonary Medicine, Allergy, & Immunology, Children’s Hospital of Pittsburgh of UPMC, University of Pittsburgh, Pittsburgh, Pennsylvania;
9. hDepartment of Pediatrics, Cardinal Glennon Children’s Medical Center, Saint Louis University, St Louis, Missouri;
10. Departments of iOtolaryngology and
11. lPediatrics, Sleep Center, Children’s Hospital of Philadelphia, University of Pennsylvania, Philadelphia, Pennsylvania;
12. jDepartment of Otolaryngology–Head and Neck Surgery, University of Cincinnati College of Medicine, and
13. kDepartment of Pediatrics, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio; and
14. nDepartments of Medicine and Neurology, Brigham and Women’s Hospital, and Department of Medicine, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, Massachusetts

## Abstract

BACKGROUND AND OBJECTIVES: Data from a randomized, controlled study of adenotonsillectomy for obstructive sleep apnea syndrome (OSAS) were used to test the hypothesis that children undergoing surgery had greater quality of life (QoL) and symptom improvement than control subjects. The objectives were to compare changes in validated QoL and symptom measurements among children randomized to undergo adenotonsillectomy or watchful waiting; to determine whether race, weight, or baseline OSAS severity influenced changes in QoL and symptoms; and to evaluate associations between changes in QoL or symptoms and OSAS severity.

METHODS: Children aged 5 to 9.9 years with OSAS (N = 453) were randomly assigned to undergo adenotonsillectomy or watchful waiting with supportive care. Polysomnography, the Pediatric Quality of Life inventory, the Sleep-Related Breathing Scale of the Pediatric Sleep Questionnaire, the 18-item Obstructive Sleep Apnea QoL instrument, and the modified Epworth Sleepiness Scale were completed at baseline and 7 months. Changes in the QoL and symptom surveys were compared between arms. Effect modification according to race and obesity and associations between changes in polysomnographic measures and QoL or symptoms were examined.

RESULTS: Greater improvements in most QoL and symptom severity measurements were observed in children randomized to undergo adenotonsillectomy, including the parent-completed Pediatric Quality of Life inventory (effect size [ES]: 0.37), the 18-item Obstructive Sleep Apnea QoL instrument (ES: –0.93), the modified Epworth Sleepiness Scale score (ES: –0.42), and the Sleep-Related Breathing Scale of the Pediatric Sleep Questionnaire (ES: –1.35). Effect modification was not observed by obesity or baseline severity but was noted for race in some symptom measures. Improvements in OSAS severity explained only a small portion of the observed changes.

CONCLUSIONS: Adenotonsillectomy compared with watchful waiting resulted in significantly more improvements in parent-rated generic and OSAS-specific QoL measures and OSAS symptoms.

• obstructive sleep apnea
• pediatric
• quality of life

#### What’s Known on This Subject:

Pediatric obstructive sleep apnea syndrome (OSAS) has been associated with decreased health-related quality of life (QoL). Observational studies suggest that adenotonsillectomy for pediatric OSAS improves QoL, but these studies did not use a randomized study design or a control group of children with OSAS managed nonsurgically.

A prospective, randomized controlled study of adenotonsillectomy for pediatric OSAS showed significantly greater QoL and symptom improvements in children undergoing adenotonsillectomy than in the nonsurgical control arm. The extent of improvement was not appreciably influenced by baseline OSAS severity or obesity.

Obstructive sleep apnea syndrome (OSAS) affects ∼1% to 3% of children1,2 and has been associated with an increased risk for developing cardiovascular and other systemic morbidities.3,4 Even milder forms of sleep-disordered breathing have been associated with behavioral disturbances in children.59 Pediatric OSAS has also been associated with decreased health-related quality of life (QoL). Studies (including a recent meta-analysis) demonstrated that children with OSAS had generic health-related QoL scores lower than healthy children and similar to children with juvenile rheumatoid arthritis.1013 Validated QoL instruments have shown moderate to large impairment of disease-specific QoL in more than one-half of surveyed children.14

The first-line surgical treatment of pediatric OSAS is adenotonsillectomy (AT). Rising health care costs and emphasis on evidence-based medicine have resulted in scrutiny of common surgical procedures, including measurement of outcomes meaningful to patients. Observational studies have suggested that in pediatric OSAS, AT improves both short-term and longer term QoL.1518 However, none of these studies used a randomized study design or a control group of children with OSAS who were not treated surgically.

The recently completed CHAT (Childhood Adenotonsillectomy Trial) was the first multisite, prospective, randomized controlled study on the effects of AT for the treatment of pediatric OSAS. Although the primary cognitive test outcome did not differ significantly between the surgical and observational treatment arms, children randomized to early AT (eAT) rather than watchful waiting had improved behavior and QoL as well as higher rates of OSAS resolution on polysomnography (PSG).19 For the present report, the impact of AT on QoL was quantified (including specific domains of function) and on OSAS symptoms. The extent to which race, baseline OSAS severity, or obesity affected treatment responses was also explored. Finally, we characterized relationships between changes in PSG indices of OSAS severity and changes in symptom and QoL measures.

## Methods

As part of the CHAT study, 453 children 5 to 9.9 years of age with OSAS were randomly assigned to either AT within 1 month (eAT) or watchful waiting with supportive care (WWSC). A detailed description of the methods of the CHAT study has been published.20 In brief, children with OSAS were recruited from pediatric sleep centers, otolaryngology and pediatric clinics, and the general community from 6 clinical sites from 2007 to 2011. Institutional review board approval was obtained from participating clinical centers, children provided assent if old enough, and caregivers provided written informed consent.

All children underwent standardized PSG; studies were scored at a central reading center to ensure uniformity.19 PSG inclusion criteria included OSAS, defined as an obstructive apnea index (OAI) ≥1 or obstructive apnea hypopnea index (AHI) ≥2. OAI is the number of obstructive apneic events per hours of total sleep time and AHI is the number of mixed or obstructive apneic events and hypopneic events associated with a ≥50% reduction in airflow and either ≥3% oxygen desaturation or electroencephalographic arousal per hours of total sleep time. Children with severe OSAS as defined by an OAI >20, an AHI >30, or oxygen saturation ≤90% for >2% of total sleep time were excluded. All children were deemed appropriate surgical candidates by otolaryngologists.

In addition to PSG data, physical examination and validated survey information were collected at baseline and 7 months later. Demographic information included age, gender, race, height, weight, ethnicity, maternal education, and family income. BMI and z scores were calculated by using standardized formulas.21

Generic and disease-specific health-related QoL and severity of OSAS symptoms were assessed with survey instruments validated for these ages. Generic health-related QoL was measured by using the parent and child versions of the Pediatric Quality of Life (PedsQL) inventory that assess physical, emotional, social, and school functioning.22 The parallel child and parent-proxy forms differ only in use of age-appropriate language. For children ages 5 to 7 years, the survey was administered by an interviewer. Scoring is performed by linear transformation of the 23 item scores to a scale of 0 to 100. Higher values indicate better QoL.

Disease-specific health-related QoL was assessed by caregivers by using the 18-item Obstructive Sleep Apnea (OSA-18) tool. This instrument focuses on perceived impact of OSAS on 5 domains: sleep disturbance, physical suffering, emotional distress, daytime problems, and caregiver concerns.23 Items are scored on a 7-point scale and totaled, providing a severity score of 18 to 126, with lower scores representing higher QoL. Mean scores for healthy children with no OSAS symptoms are in the range of 31.2 ± 10.4.24 Scores >60 suggest a moderate impact.

To assess OSAS symptom severity, caregivers completed the Sleep-Related Breathing Disorder (SRBD) scale of the Pediatric Sleep Questionnaire (PSQ) and the Epworth Sleepiness Scale modified for children (mESS). The PSQ SRBD scale contains 22 yes/no questions and provides both a total score, as the proportion of all symptoms endorsed by the caregiver, and subscale scores for snoring, daytime sleepiness, and behavior. The mean of yes (1) and no (0) responses generates a score between 0 and 1, with higher scores indicating greater symptom severity. Values ≥0.33 have been proposed as identifying higher risk for pediatric OSAS.25 The sleepiness subscale has been validated against objective sleepiness in children.26 On the mESS, caregivers rate the likelihood of their child falling asleep from 0 (never) to 3 (almost always) in 8 situations. Scores range from 0 to 24, with higher scores indicating more sleepiness.27

PSG parameters used to assess OSAS severity were the AHI and oxygen desaturation index (ODI [ie, number of episodes of oxygen desaturation ≥3% per hour of sleep]). The AHI reflects both sleep fragmentation and hypoxemia, whereas the ODI more specifically assesses intermittent hypoxemia.

Baseline demographic variables are summarized according to treatment arm (ie, eAT, WWSC) as mean ± SD values for continuous variables or frequency (%) for categorical variables. Baseline comparisons of QoL and symptom measurements according to study arm were examined by using 2-sample independent t tests (unadjusted P value) or analysis of covariance (ANCOVA). These and all other ANCOVA models were adjusted for site, race (African American versus non–African American), age (5–7 vs 8–10 years), and overweight status (≥85th vs <85th BMI percentile) as the primary analysis and site. Race (African American versus non–African American), gender, age (continuous), obesity (≥95th vs <95th BMI percentile), maternal education (less than high school, high school or higher, or missing), income (<$30 000, ≥$30 000, or missing), and baseline log AHI were included in the secondary analysis.19 To assess whether the WWSC and eAT arms experienced a differential change in QoL and symptom measurements, unadjusted analysis of variance and adjusted ANCOVA models were fit with the QoL and symptom outcomes expressed as change from baseline to follow-up. Additional ANCOVA models included interaction terms to assess effect modification for treatment response according to baseline OSAS severity, race, and weight. Furthermore, linear regression models were used to assess associations between change in QoL or symptoms and change in PSG measures (log transformed to approximate normal distribution). In this last regression model, data from the 2 treatment arms were combined. This technique was used because OSAS resolution, defined as AHI <2 and OAI <1 at follow-up, was observed in a large proportion of subjects in both treatment arms (46% of WWSC subjects and 79% of eAT subjects). Sensitivity analyses were conducted, however, stratified by treatment arm. A total of 24 children (16 in the WWSC arm and 8 in the eAT arm) were not treated per protocol. Exploratory analyses performed for the original CHAT publication did not yield appreciable changes in results when those subjects were excluded from the analyses. Cohen’s d effect size was calculated as (mean change difference)/(pooled SD). Statistical analyses were performed by using SAS version 9.3 (SAS Institute, Inc, Cary, NC) with α cutoff of ≤0.01.

## Results

No significant baseline differences in demographic characteristics, QoL, or symptom survey total scores were seen between treatment arms. There was a significant difference in the emotional function domain of the parent PedsQL with a higher score seen in the eAT arm (Tables 1 and 2).

TABLE 1

Demographic Characteristics of the Study Population at Baseline

TABLE 2

QoL and Symptom Measures According to Treatment Arm at Baseline

### Generic Health-Related QoL (PedsQL)

The PedsQL parent-reported total score improved significantly more in the eAT group compared with the WWSC group (5.9 ± 13.6 points vs 0.95 ± 13.3 points [P < .01], yielding an effect size of 0.37) (Table 3). Significance did not alter with adjustment for site, continuous age, race, socioeconomic status, or obesity status. This difference was driven by highly significant differences for the school (academic performance) and physical function domains. The change scores for the emotional function domain also differed between arms after adjusting for baseline differences (P < .01). Changes in the social function (peer interaction) domain were not significantly different between the arms. No significant differences in change scores between treatment groups in the PedsQL total score or the 4 domains were noted in the surveys answered directly by the children.

TABLE 3

Change Scores in QoL and Symptom Measures by Treatment Arm

### Disease-Specific Health-Related QoL (OSA-18)

Total OSA-18 scores in the eAT group improved more than in the WWSC group by –21.4 ± 16.5 vs –4.5 ± 19.3, producing a large effect size of –0.93 (P < .01) (Table 3). Significant differences in the change scores between the 2 arms were seen in all of the individual domains of the OSA-18, including sleep disturbance, daytime problems, physical suffering, caregiver concerns, and emotional distress (all P ≤ .01).

### Symptom Questionnaires (PSQ SRBD and mESS Scores)

For the Sleep-Related Breathing Scale of the Pediatric Sleep Questionnaire (PSQ-SRBD), a –0.28 ± 0.2 point change in the eAT group and –0.03 ± 0.2 change in the WWSC group produced a large effect size of –1.35 (P < .01) for the differences between arms (Table 3). Moreover, significant differences in the change scores between treatment arms were seen for the behavior, sleepiness, and snoring subscales (all P < .01). Improved sleepiness was corroborated by significant improvement in the mESS score in the eAT group of –2.01 ± 4.7 compared with 0.28 ± 4.1 in the WWSC arm, with a moderate effect size of –0.42 (P < .01).

Change score differences between the treatment arms for the QoL and symptom survey total scores are summarized in Fig 1.

FIGURE 1

Summary of differences in QoL and OSAS symptom score changes in the eAT and WWSC arms. Absolute values were used when change scores were negative to facilitate comparisons of effect magnitude. *P < .01 for difference between arms, adjusted for site, race, age, obese (<95 or >95 BMI percentile), gender, maternal education (less than high school, high school or higher, or missing), income (>$30 000, ≤$30 000, or missing), log baseline AHI, and baseline outcome variable.

### Assessment of Effect Modification by Race and Baseline Weight and OSAS Severity

Weight did not influence the associations between treatment arm and QoL or symptoms (Table 4, all P > .05). Interaction terms for race were not significant for models for the majority of QoL and symptom outcomes. In contrast, effect modification by race was observed for the association between intervention group and both the PSQ-SRBD total score and behavior subscale, even after adjustment for measures of socioeconomic status (Table 5). Specifically, smaller relative improvements associated with AT were reported by caregivers of African-American children compared with non–African-American children for those 2 symptom measures (P = .01 and <.01, respectively, for the relevant interaction terms). These differences persisted in analyses restricted to the 76 African-American children and 81 non–African-American children in the eAT arm whose OSAS resolved by PSG (P values for the fully adjusted models all <.01, data not shown).

TABLE 4

Effect Modification on Change: Weight Category (Normal Versus Overweight Versus Obese)

TABLE 5

Effect Modification on Change: Race (African American Versus Non–African American)

Baseline OSAS severity (AHI or ODI quartiles) also did not influence the association between treatment arm and QoL or symptoms (all P > .01, data not shown).

### Association of QoL and OSAS Symptoms With PSG Measures

In general, improvements in OSAS severity measured by using PSG explained only a small portion of the variance in the QoL and symptom change scores. Change in AHI correlated, albeit weakly, with change in mESS (partial r2 = 0.03, P < .01), OSA-18 (partial r2 = 0.07, P < .01), PSQ SRBD scale (partial r2 = 0.14, P < .01), PSQ snoring subscale (partial r2 = 0.17, P < .01), and PSQ sleepiness subscale (partial r2 = 0.03, P < .01) (Table 6). Small but significant associations were also seen between change in ODI and OSA-18 total score (partial r2 = 0.05, P < .01), PSQ SRBD scale (partial r2 = 0.09, P < .01), PSQ snoring subscale (partial r2 = 0.09, P < .01), and PSQ sleepiness subscale (partial r2 = 0.02, P < .01) (Table 7). In contrast, changes in AHI and ODI were not significantly associated with changes in generic health-related QoL.

TABLE 6

Association Between QoL and Symptom Change Scores and PSG Change Scores (Log AHI)

TABLE 7

Association Between QoL and Symptom Change Scores and PSG Change Scores (log ODI)

## Discussion

This large, randomized controlled trial of children with OSAS found that key symptoms and QoL improved substantially more after eAT than WWSC. Benefits from eAT were evident in generic and disease-specific health-related QoL (as measured by using the PedsQL and OSA-18) and in OSAS symptoms (as reflected by using the PSQ SRBD scale and the mESS). Moderate to large improvements were observed for most QoL and symptom measurements, including the parent-completed PedsQL (total score, school, emotional, and physical function domains), OSA-18 (total and all 5 domains), mESS, and the PSQ SRBD (total score and snoring, sleepiness, and inattentive/behavioral subscales). Improvement in OSAS severity measured by using PSG variables explained only a small proportion of the improvements seen in OSAS symptoms and QoL. These observations have important clinical implications for the many children with OSAS who are evaluated for AT. The findings are of particular relevance, given the growing interest from patients, payers, and providers that QoL and symptom outcomes be highlighted in the management of chronic health conditions.

Clinically, many factors are considered when making a decision to perform surgery or to judge the success of surgical interventions. Previous studies measuring success rates for AT in children with OSAS have often focused somewhat narrowly on normalization of PSG parameters, with reports of surgical success rates ranging from 27% to 83%; lower cure rates are typically reported in obese children.19,2831 This emphasis on PSG measures of disease resolution may be partially due to assumptions that PSG severity parallels severity of the symptoms seen with OSAS. This concept is not well supported by the current literature for either neurobehavioral morbidity or QoL.19,26 Previous studies of children with OSAS have shown no association between baseline OSA-18 scores and severity of OSAS on PSG.32 Moreover, studies have failed to demonstrate clear correlation between extent of PSG improvements after AT and improvement in QoL.33,34 In CHAT, we observed correlations between changes in AHI or ODI and changes in QoL and symptom severity measures. However, PSG improvements explained only a small portion of the variance for the change scores (partial r2 ranging from <0.01 to 0.17). Thus, both previous literature and current data indicate that using PSG results as the sole metric for effectiveness of AT in pediatric OSAS may neglect other benefits that are important to children and their families.

The large proportion of our subjects who were overweight or obese allowed for subgroup analysis of QoL and symptoms. Increased likelihood of persistent OSAS after AT in obese children has been well documented, including a meta-analysis of 23 studies.28,30 Obesity has also been associated with decreased QoL in children.35 Improvement in QoL after AT for OSAS in the obese population has, however, been reported. A study of children with OSAS and BMI >95% showed improvement in OSA-18 general and domain scores despite lack of resolution of OSAS in the majority of subjects.31 In the present analysis, although only obese children considered to be candidates for AT were included, obesity did not influence the relative changes in QoL or OSAS symptom severity with each intervention. These findings are supported by a study of QoL in children with severe obesity which showed that of 7 obesity-related comorbidities, only OSAS was associated with significant decreases in QoL.35 The improved QoL and symptom outcomes seen in obese children support a clinically beneficial effect of surgery relative to watchful waiting for children in this group for whom treatment controversies exist.

OSAS has also been shown to be more common in African-American children.36 More than one-half (55%) of the CHAT study participants were African American, which enabled evaluation for effect modification of race on the changes in QoL and symptoms between treatment arms. A significant effect modification of treatment by race was seen when comparing African-American versus non−African-American study participants for the PSQ SRBD total score and behavior subscale. Specifically, caregivers of American-African children in the eAT arm reported less improvement in children’s behavior than did caregivers of non−African-American children. These differences persisted after adjustment for socioeconomic status and in an analysis restricted to children in whom OSAS resolved by PSG. In conjunction with the lack of improvement noted by the child-completed PedsQL survey, however, it must be considered that differing caregiver expectations about the beneficial effects of surgery or what constitutes problematic behavior may have influenced responses.

In the present study, none of the child-reported PedsQL measurements differed significantly between the 2 treatment groups. Previous studies have shown an ability of the child PedsQL to detect significant differences in the summary and domain scores between healthy children and children with a variety of chronic diseases.37 However, OSAS was not specifically evaluated. Conceivably, children have difficulty recognizing their own sleepiness, irritability, or decreased concentration or do not consider those symptoms as problematic as the pain or physical limitations experienced with other chronic diseases. An alternate explanation is that the improvements reported by caregivers represent a desire to justify surgical interventions.

The major strengths of this study of health-related QoL and OSAS symptoms in pediatric patients undergoing AT for OSAS were a large, diverse sample recruited from multiple pediatric centers and use of a randomized design with a control group and highly rigorous and standardized measurement approaches. The study addressed patient-reported outcomes, which are increasingly recognized as important to patients and other stakeholders in health care. However, it should be noted that measures of QoL are inherently subjective, and in the setting of a surgical trial with an inability to blind participants, it is possible that the larger improvements in QoL and symptom measurements seen in the eAT arm could reflect a surgical placebo effect or variability of caregivers in assessing symptoms. However, the significant (albeit small) correlation with PSG improvement provides support for treatment-associated effects. An additional shortcoming was the limited follow-up period of 7 months.

## Conclusions

This large, multisite, prospective, randomized controlled study of AT for PSG-documented pediatric OSAS found that key parent-reported measures of QoL and symptoms, or “patient- centered outcomes,” improved substantially and significantly more in children treated with surgical AT than in children treated with WWSC. Improvements in QoL and OSAS symptoms were associated with improvement in PSG indicators of disease severity; however, only a small proportion of the observed QoL and symptomatic improvement was explained by PSG improvement. This study strongly supports the consideration of metrics beyond those reflected by PSG parameters when evaluating the value of AT in children with symptomatic OSAS.

## Acknowledgments

CHAT gratefully acknowledges the superb support of the CHAT research staff: Jean Arnold, Mary Ellen Carroll, Mary Anne Cornaglia, Beth Ann Compton, Casey Critchlow, Judith Emancipator, Melissa Fernando, Theresa Friederich, Amanda Goodman, Xiaoling Hou, Elise Hodges, Laurie Karamessinis, Kim Lacy, Megan McDougall, Daniel Mobley, Michelle Nicholson, Angela Orlando, Deborah L. Ruzicka, Gauri Sathe, Nancy Scott, Susan Surovec, Omarya Vega, Xingmei Wang, and Catherine Williams. The authors also appreciate the generous participation of the families enrolled in the study. They are grateful for the helpful guidance during the study of the CHAT Data and Safety Monitoring Board: Lynn Taussig, MD (Chair); Thomas Anders, MD; Julie Buring, ScD; Karina Davidson, PhD; Estelle Gauda, MD; Steven Piantadosi, MD, PhD; Bennett Shaywitz, MD; Benjamin Wilfond, MD; Tucker Woodson, MD; and Robert Zeiger, MD.

## Footnotes

• Accepted November 4, 2014.
• Address correspondence to Susan L. Garetz, MD, Department of Otolaryngology-Head and Neck Surgery, University of Michigan Health System, 1540 East Hospital Dr, Ann Arbor, MI 48109-4241. E-mail: garetz{at}umich.edu
• This trial has been registered at www.clinicaltrials.gov (identifier NCT00560859).