OBJECTIVES. Although many children with obstructive sleep apnea syndrome have complete resolution of obstructive sleep apnea syndrome after adenotonsillectomy, some patients have persistent obstructive sleep apnea syndrome requiring positive airway pressure treatment. Little is known about positive airway pressure adherence among school-aged children and adolescents.
PATIENTS AND METHODS. We retrospectively reviewed records from January 2000 through December 2004 to assess positive airway pressure adherence following a comprehensive patient- and parent-focused positive airway pressure education program for children 7 to 19 years of age with persistent obstructive sleep apnea syndrome subsequent to indicated adenotonsillectomy. A polysomnogram was obtained before and after initiation of positive airway pressure therapy. Adherence was defined as >4 hours per night and ≥5 nights per week of positive airway pressure use. Clock-counter meters determined hours per night and nights per week of positive airway pressure use; parents estimated hours per night of positive airway pressure use. Nonparametric tests assessed associations between adherence and various clinical parameters and symptoms.
RESULTS. Forty-six patients (56% male; 39% black, 61% white; mean age: 13.6 years; mean BMI: 39.8 kg/m2) were included. Two refused positive airway pressure. Meter readings were available for 27 patients (59%); positive airway pressure was used, on average, 7.0 hours per night, 73% of the week, and for a mean of 18.1 months. Nineteen (70%) were adherent regardless of age. There was good agreement between parental report and meter readings. Patients with greater improvement in apnea-hypopnea index were more likely to be adherent. Clinical parameters and symptoms improved after positive airway pressure therapy regardless of age or adherence.
CONCLUSIONS. In this retrospective study, positive airway pressure adherence and symptom improvement among school-aged children and adolescents was achieved with comprehensive patient and parent education and follow-up.
Untreated obstructive sleep apnea syndrome (OSAS) in children can result in serious health complications.1 It imposes a substantial health care burden,2 along with a negative effect on health-related quality of life.3 Although adenotonsillectomy remains the treatment of choice for most cases of OSAS in children,4 a cumulative cure rate of only 80% is achieved,5 with a percentage of patients with persistent OSAS requiring positive airway pressure (PAP), using single-level or bilevel airway pressure.
PAP therapy is effective for the long-term treatment of OSAS both in adults6,7 and children.8–12 Nonadherence to PAP use, however, limits its effectiveness. In adults, PAP adherence has been reported to vary between 65% and 80%, with 8% to 15% of patients refusing this treatment after a single night's use in the laboratory.13,14 The definition of adherence varies across studies, however. Although many studies have focused on nightly duration of PAP use,15–17 night-to-night consistency and frequency of use may be as important.18 One study, for example, reported that only 46% of adult patients used PAP for ≥4 hours on 70% of the days.19
Studies in children are even more varied, with adherence estimated to be between 50% and 100%.9,10 These estimates, however, have been based on parental reports or questionnaires sent to pediatric practitioners. Two recent studies measured objective PAP adherence in children using clock-counter/meter readings.20,21 Nevertheless, all of these studies included patients from a wide age range of 6 months to 18 years, with adolescents reported to be the least adherent.9,10,20,21 One might argue that adherence in younger children is primarily a function of parent cooperation. Adherence by school-aged children and adolescents, however, may need to be examined against a background of changing physical, psychological, and social development unique to this group.22–24
Studies in adults suggest that high PAP adherence rates can be achieved in an environment that fosters patient education, comprehensive follow-up, and integrated care.17,25 Variables found to be significantly correlated with PAP usage in adults include higher apnea-hypopnea index (AHI), higher Epworth sleepiness scores, higher baseline BMI, male gender, older age, and patient support and education.15,17,25–28 Only 1 similar study has been conducted in children,20 and no similar studies have been conducted on school-aged children and adolescents, the group commonly considered to be least adherent. In a retrospective study, we describe nasal PAP effectiveness and adherence among school-aged children and adolescents who had been followed in our clinic through a comprehensive program dedicated to PAP education and follow-up.
The study was approved by the institutional review board at Washington University School of Medicine. Patient records from the St Louis Children's Hospital Sleep Clinic beginning January 2000 through December 2004 were retrospectively reviewed. We included all of the children ≥7 years of age who, despite indicated adenotonsillectomy, had OSAS confirmed by polysomnogram based on published standard criteria.29 The patient was considered to have OSAS if an AHI ≥5 and/or presence of any abnormalities in gas exchange was documented, including a pulse oxygen saturation (Spo2) nadir of ≤85%, Spo2 <90% for ≥10% of total sleep time (TST), peak end-tidal CO2 (ETCO2) ≥55 mmHg, and ETCO2 >50 mmHg for ≥10% of TST. All of the patients underwent a baseline polysomnogram evaluation followed by a polysomnogram PAP titration study. Patients with severe cognitive impairment (who, therefore, were unable to understand or adhere to instructions), who were on PAP therapy for indications other than OSAS, or who had received PAP treatment before January 2000 were excluded.
Patients were assessed in a dedicated sleep clinic by 1 physician, who is a pediatric, board-certified sleep specialist. A set of standard questions was used to determine baseline clinical symptoms observed by parents; this included the presence of snoring, witnessed episodes of apnea and choking respirations during sleep, sleeping patterns, presence of secondary enuresis, and subjective assessment of changes in school performance, behavior, and sleepiness since the onset of nighttime symptoms. Patients and their parents were also given a 30-minute educational session about OSAS as part of their initial clinic visit. Health consequences of OSAS and treatment options, including an overview of PAP therapy, were reviewed before the patients' baseline polysomnogram.
All of the patients underwent baseline polysomnogram evaluation. After baseline data were collected, a brief PAP trial was offered during the last hour of the initial overnight polysomnogram to familiarize patients with PAP therapy. Baseline polysomnogram evaluation results were then discussed with the patient and his or her parents 2 weeks after the initial polysomnogram, and additional PAP education with emphasis on the importance of adherence was provided. A second polysomnogram was obtained to determine optimal PAP pressure settings. Follow-up telephone calls by a dedicated nurse were made shortly after optimal PAP settings were in place, and patients had clinic follow-up visits 2 to 4 weeks later and every 6 months thereafter. Problem areas, if any, were determined at each follow-up visit. The number of hours of PAP use was obtained from the patient and/or parent and, for those patients who used a clock-counter/meter, the number of hours and frequency of use during the week were obtained from the meter readings. Patients with meter readings had PAP devices with a built-in monitoring chip that registered use when the set pressure was maintained and then collected and stored the PAP usage data. Monitored data were downloaded using the Encore Pro Data Management System (Respironics, Murrysville, PA) during the clinic follow-up visit. Detailed meter-reading summaries of the hours of daily use, time of days used, and days used per month were reviewed.
Polysomnography and Institution of PAP Therapy
All of the polysomnograms were performed at the St Louis Children's Hospital sleep laboratory. Patients fell asleep without the aid of sleep medication and were observed for ≥8 hours overnight in the company of ≥1 parent. Parameters measured included chest and abdominal wall movement by piezoelectric respiratory belts (Respironics); heart rate by electrocardiogram; and airflow with sidestream end-tidal capnograph, which provided breath-by-breath assessment of ETCO2 levels (Capnocheck Plus; BCI International, Waukesha, WI) and an oronasal thermistor (Respironics). Arterial oxygen saturation was assessed using pulse oximetry (Spo2) with simultaneous recording of the pulse waveform (BCI International). Bilateral electrooculogram, 8 channels of electroencephalogram, chin electromyogram, and analog output from a body-position sensor were monitored. Tracheal sound was monitored with a microphone sensor (Respironics), and synchronized video recording was performed. All of the measures were digitalized using a commercially available polysomnographic system (Alice III; Respironics), where data were acquired, recorded, and stored. Raw data were manually scored by 30-second epoch, according to published standards.29 Obstructive apnea was defined as the absence of airflow with continued chest and abdominal wall movement for a duration of 2 breath cycles.30 Hypopnea was defined as a ≥50% decrease in nasal airflow with a corresponding hypoxemia and/or arousal.29 The AHI was defined as the number of apneas and hypopneas per hour of TST. The mean Spo2 and Spo2 nadir were determined. Arousals were defined as recommended by the American Sleep Disorders Task Force Report.31
A titration polysomnogram was obtained using the same polysomnogram parameters as described. The goal of the titration polysomnogram was to improve gas exchange and to reduce the AHI to normal or near normal levels through PAP at a level tolerated by the patient. PAP was started at single level pressure of 5 cmH2O and increased by increments of 2 cmH2O when needed. Patients who seemed uncomfortable or who required single level pressures of >15 cmH2O were switched to bilevel pressures, beginning at 10/5 with ≥5 cmH2O difference between inspiratory and expiratory pressures.
Once optimal pressures were determined, the family received a follow-up telephone call from a dedicated sleep nurse to review study results, PAP pressures, and instructions regarding home health PAP set-up and clinic follow-up. Humidifiers were used for all of the patients. A representative of the home health care company who visited the patients in their home offered various masks for the best fit, and follow-up appointments in the sleep clinic were arranged 2 to 4 weeks into PAP therapy.
Assessment of PAP Effectiveness
Objective polysomnogram findings, including AHI, arterial Spo2 nadir, mean ETCO2 levels, and presence of obstructive hypoventilation (defined as the presence of ETCO2 of ≥50 mmHg for >10% of TST), were compared before PAP and after optimal PAP settings were initiated on repeat polysomnogram. The repeat polysomnogram was performed on PAP. In addition, clinical symptoms based on history, including nocturnal (snoring, witnessed apneas, choking episodes, and enuresis) and daytime symptoms (observed daytime sleepiness, hyperactivity, and school performance), were compared before and after home PAP use. A symptom score was computed on the basis of the total number of 4 nocturnal symptoms and daytime somnolence reported, for a total possible symptom score of 5.
Assessment of PAP Adherence
Both objective (clock-counter/meter) and subjective (parental report) data were obtained only from the patients with meter readings to determine the mean number of hours of PAP use per night and the frequency of PAP use during the week (number of nights per week in which PAP was used). Patients were considered to be adherent if the meter readings indicated PAP use of >4 hours per night of use and ≥5 nights per week, similar to the calculation of adherence in some adult studies.19,32
Nonparametric tests (eg, χ2 tests and Mann-Whitney U tests) were used to measure the significance of univariate associations between adherence to PAP use (for the 27 patients with meter readings) and each of the clinical outcomes, presence and absence of individual symptoms, total number of symptoms, and change in symptoms, as well as gender, race, age, and BMI. Spearman correlations measured associations among continuous variables, such as age, average hours of PAP use per night, total number of symptoms before and after PAP, and change in symptoms. We illustrated the agreement between parent-reported and meter readings of mean hours per night of PAP use (for patients who used clock-counter meters) using a Bland-Altman plot,33 which plots the differences between the 2 measures against the averages of the 2 measures. Wilcoxon signed-rank tests were used to measure the significance of changes in each of the symptoms and in the total number of symptoms measured before and after home PAP use. We also ran a repeated-measures analysis of variance to test whether the change in the total number of symptoms was different between adherent and nonadherent patients.
Forty-six patients who were seen in the St Louis Children's Hospital Sleep Clinic from January 2000 through December 2004 met the inclusion criteria (Table 1). No patients were excluded from the study on the basis of the cognitive-impairment exclusion criterion. The mean age of these patients was 13.6 years (SD: 3.1 years), and mean BMI was 39.8 kg/m2 (SD: 15.2 kg/m2), ranging from 16.7 to 76.4 kg/m2 (BMI z scores ranged from −1.52 to 2.41). As shown in Table 1, more than half of the sample was <15 years of age. Thirty-one patients (67%) had obesity as an associated comorbidity of OSAS. Mean AHI at baseline was 28.4 events per hour. Of the 46 patients, 2 patients (4%), aged 11 and 19 years, completely refused PAP therapy despite further education and follow-up. These 2 patients had very mild OSA on the basis of a baseline AHI of 1.5 and 6.4 events per hour, respectively. Of the 44 patients on PAP, 30 (68%) were on single pressure with a mean continuous PAP pressure of 8.8 cmH2O (SD: 2.1 cmH2O); and 14 (32%) were on bilevel pressures, with a mean inspiratory pressure of 14.9 (SD: 3.5) and a mean expiratory pressure of 8.4 (SD: 2.6). Eleven patients reported problems during PAP use. The most common problem involved mask fit (8 of 11); other problems included skin breakdown, dry nose, and stuffy ears, each reported by 1 patient.
Twenty-seven patients (59%) had clock-counter/meter readings. At the time the data were collected by chart review, the average length of home PAP use among the 27 patients using PAP with meters had been 18.1 months (SD: 11.8 months; range: 3–43 months; median: 15.0 months).
As shown in Tables 2 and 3, changes in AHI, Spo2 nadir, and clinical symptoms showed significant improvement regardless of PAP mode (single or bilevel pressure). Polysomnogram parameters showed significant improvements in AHI, Spo2 nadir, mean ETCO2 level, and the presence of obstructive hypoventilation after PAP (P < .001).
After home PAP use, there was a significant decrease in nocturnal symptoms reported at the follow-up clinic visit (each P < .001). Significant improvements in daytime somnolence and school performance were also observed (each P < .001); improvement in hyperactivity was not statistically significant (P = .053).
The frequency of use and adherence could be calculated only for the 27 patients for whom meter readings were available. Of these 27 patients, 19 (70%) were determined to be adherent, using PAP a mean 73% (SD: 28%) of the week. Twenty-three patients (85%) used PAP >4 hours per night, with mean of 7.0 hours per night (SD: 2.1 hours per night) of use; the mean parent report for these 27 patients was 7.7 hours per night (SD: 1.7 hours per night). Figure 1 shows histograms of the number of patients by hours per night of PAP use based on parent report (Fig 1A) and meter readings (Fig 1B). As the Bland-Altman plot (Fig 2) illustrates, parental report showed good agreement with meter readings for 24 of the 27 patients having both parent and meter readings (agreement would not be considered “good” for the points above or below the 95% confidence limit). Overall, the limits of agreement (−1.91 and 3.30) were small enough to indicate that, whereas not perfect, parental report showed good agreement with meter readings. However, as shown in Table 4, parents of patients with meter readings of ≤4 hours per night of use overestimated the mean hours of PAP use per night by ∼2 hours compared with the actual meter readings, whereas for patients with meter readings >4 hours per night, parents overestimated the mean hours of PAP use per night by only half an hour.
Neither the presence of subjective OSA symptoms, either alone or in total, nor the severity of OSA as defined by baseline AHI, Spo2 nadir, mean ETCO2 levels, or the presence of obstructive hypoventilation correlated with meter measures of hours of PAP use per night in our patients. However, when the frequency of use was also taken into account, patients who were adherent had higher baseline AHI (P = .022) and showed greater improvement in AHI on PAP (P = .015) compared with nonadherent patients. The proportion of adherent patients did not differ significantly by age, gender, BMI, or race. The mode of PAP therapy (single versus bilevel pressure) was not significantly associated with adherence (Fisher's exact test, P = .145), although 80% (16 of 20) of patients on single pressure were found to be adherent compared with 43% (3 of 7) of patients on bilevel pressure. For the 27 patients with meter readings, the average number of hours of PAP use per night, either by parental report (P = .004) or meter reading (P = .005), correlated significantly with the frequency of use (ie, the percentage of nights that PAP was used during the week).
Similar to previous studies in adults6,7 and children,8–12,20 our study showed PAP to be associated with improvement in clinical symptoms and objective improvement in polysomnogram parameters. Although our study showed no significant change in reported hyperactivity or behavior problems, this may be because of the small number of patients reporting this symptom, because only 8 (18%) of 45 patients had hyperactivity or behavior problems before initiation of the PAP treatment. That 65% of patients had obesity as a comorbid condition deserves attention, given the growing problem of obesity in the United States34 and growing interest in obesity as a risk factor for OSAS in children.35
Contrary to earlier observations in children, which did not use objective meter readings,9,10 but comparable to observations in adults,17,32 85% of our sample used PAP >4 hours per night based on clock-counter/meter reports (Table 4). Moreover, similar to findings in adults,18 we found that mean hours of PAP use per night correlated positively with frequency of PAP use during the week. In contrast to 1 study showing regular PAP use in only 46% of adult patients,19 70% of our sample used PAP >4 hours per night for an average of 73% of the week. In a recent pediatric study, one third of the children dropped out before 6 months.20 In contrast, we collected follow-up data for all of our patients after PAP was set up at home, with only 2 patients refusing PAP therapy. However, our study was a retrospective review of patients' charts, whereas the Marcus et al study20 used a prospective design. Although we asked patients about problems that they experienced with PAP use, their reasons for adherence to PAP therapy were not identified. It is possible that patients who were poorly motivated to use PAP at the outset did not keep their clinic appointments to have their initial polysomnogram or to have PAP started and, therefore, were not included in the chart review. Moreover, our sample consisted of school-aged children and adolescents with OSAS-associated comorbidities and persistence of OSA symptoms despite adenotonsillectomy. This persistence of OSA, in itself, may have motivated these patients to continue with the recommended PAP therapy. However, we cannot attribute the persistence of OSAS as a cause of adherence to PAP because of the limitations of our retrospective study design.
Our results suggest that PAP adherence can be achieved among school-aged children and adolescents, a group deemed to be least compliant in earlier studies.9,10 This finding is notable given that adherence to medication among adult patients with chronic conditions has been reported to range from only 43% to 78% and drops dramatically after the first 6 months of therapy.36 Similarly, ∼50% of adolescents with chronic conditions do not comply with care recommendations.37 Among adult patients accepting home PAP, 12% to 25% can be expected to discontinue PAP use by 3 years.26 Long-term use of PAP by school-aged children and adolescents has not been reported previously; patients in our study had used PAP for a mean duration of 18.1 months at the time data were collected by chart review, ranging from 3 to 43 months of use among the 27 patients with meter readings.
Factors that influence adolescents' adherence to chronic disease treatments have been studied with inconclusive findings.37 The severity of OSA as defined by baseline AHI, Spo2 nadir, and symptom scores was not significantly correlated with meter measures of hours of PAP use per night in our patients, similar to a recent report on adherence to PAP in children 2 to 16 years of age.20 As reported previously in adult studies, however, the mean number of hours of use per night alone may not be an accurate reflection of actual PAP use and adherence.18 Using a more restrictive definition of adherence, including frequency of nights used per week, as well as mean number of hours of PAP use per night, we found that baseline AHI and change in AHI on PAP differed significantly between adherent and nonadherent PAP users, similar to findings in adults.26,28 On the basis of our findings, we submit that a more conservative definition of adherence that includes both the number of hours and frequency of use would provide a more clinically meaningful definition of adherence than merely the mean number of hours per night used.
We found no significant associations between PAP adherence and gender, age, or BMI in our pediatric sample, similar to other reports in adults18,32 and in children.20 In addition, there was no significant difference between PAP mode (single versus bilevel pressure) and either PAP effectiveness or adherence among the 27 patients with meter readings, which previous studies also reported.20,38 Our sample size, however, may be too small to identify statistically significant associations between PAP adherence and these other variables (with increased risk of a type II error). The calculation was based on 27 patients with meter readings, with 20 patients on single and only 7 patients on bilevel pressures, and the patients were not randomly assigned to treatment group. Thus, a prospective randomized trial to compare adherence using single and bilevel PAP in a larger sample of school-aged children and adolescents is recommended.
Studies in adults have suggested that intense patient education is associated with improved adherence.17,25 Our patients and their parents were all educated about OSAS and PAP by a pediatric, board-certified sleep specialist. Information alone, however, may not be enough to promote behavioral change in adolescents. Anxiety has been found to be related to nonadherence to PAP therapy in adults.39 Whether anxiety is associated with nonadherence to PAP therapy in adolescents is unknown. For chronically ill adolescents, attitudes, as well as the personal meaning and significance of an illness and its treatment, have been cited as an important factor affecting adherence.37 In addition, the importance of a good relationship between the adolescent patient and health care staff cannot be overemphasized.37 We believe that the relatively high level of adherence to PAP in our study may have been related to the continuity in care and follow-up of our patients, but an experimental design would be necessary to test this hypothesis as well. How their adherence to PAP may change over time also remains to be seen.
Much has been written about the differences in adolescents' cognitive development as they go through early (11–14 years), middle (15–17 years), and late (17–21) adolescence.23 More than 70% of our patients were in the early and midadolescent age groups, when formal operative thinking has yet to develop and children still have limited ability to consider long-range health risks.40 Nevertheless, we observed no significant differences in adherence among the 4 age groups, the youngest age group being <11 years of age. Moreover, only 2 (4%) of the 46 patients refused PAP; 1 was 11 years old (early adolescence) and the other was 19 years old (late adolescence). This low level of refusal to use PAP is similar to the report by McArdle et al in adults,26 but less than the 8% to 15% reported by others.13,14 Janson et al41 identified the lack of subjective effect of treatment as a common reason for nonadherence among his adult patients, and lack of perceived benefit might have been the reason for nonuse of PAP in the 2 patients who refused PAP, because both of these patients had very mild OSA at baseline (AHI of 1.5 and 6.4 events per hour).
Similar to recent studies,20,21 we only included children and adolescents who had undergone otolaryngology evaluation and adenotonsillectomy, a potential confounding variable and the treatment of choice in the pediatric population.4 If present after adenotonsillectomy, OSAS must be viewed as a chronic illness and PAP as long-term treatment. Future research might be designed to evaluate differences in adherence as a function of school-age and adolescent cognitive development using a prospective study design. Better understanding of adolescents' perceptions of OSAS and how their perceptions affect adherence to PAP therapy are warranted. Moreover, whether adolescents who are adherent with PAP therapy continue to be adherent as adults remains to be seen.
This study has limitations inherent to its small sample size and retrospective design. Intensive support27 and behavior intervention42 have been recommended to improve PAP adherence. However, although we think our educational intervention and regular follow-up of patients constituted an effective and comprehensive support mechanism, we cannot infer that our intervention “caused” adherence to PAP resulting in improved outcomes, because we analyzed existing data, and the intervention was offered to all of the patients. Also, whereas parent-reported hours of use per night showed good agreement with meter readings in this study (but for a few points outside the confidence bounds in Fig 2), studies have shown that self-reported use in adults overestimated PAP use by an average of 1.0 hours when compared with the more objective meter readings.43 A recent study of PAP adherence in children also showed parental assessment overestimated actual use,20 and we found that parents of patients with meter readings of ≤4 hours per night overestimated the mean hours of PAP use per night by ∼2 hours (Table 4). An important difference between the study by Marcus et al20 and our study is that we reported better adherence among our school-aged children and adolescents, with more patients in our study registering >4 hours of PAP use per /night on the meter than registering ≤4 hours per night. But with 27 patients in all (and only 4 in the group registering ≤4 hours per night on the meter), our finding of good agreement between parent-reported use and meter readings may be a “false-negative” finding and may be attributed to increased β error (low power to detect a statistically significant difference). Regardless, the contrast in findings between the Marcus et al study20 and ours speaks to the necessity of additional research using an experimental design to better assess the external validity of findings from each of the 2 studies. In addition, in this retrospective study, not all of the patients were blinded to the presence of a clock-counter meter. It remains to be seen whether having knowledge of the presence of a clock-counter meter correlates with better PAP adherence.
In addition, whereas adherence had been defined in terms of hours and frequency of use, sleep patterns and habits also may be important, given reports that OSAS is predominantly a rapid eye movement phenomenon in children.44 Because children tend to consolidate rapid eye movement sleep toward the latter half of the night and the early morning, when most PAP sessions might end, consideration as to what constitutes regular use might need to include not only the number of hours used but, specifically, during which hours of sleep it is used.
This study shows that most of the school-aged children and adolescents who received a dedicated education and follow-up program at our institution were adherent to PAP therapy. Patients with a higher baseline AHI and a greater change in AHI on PAP were more likely to be adherent. Adherence was not associated with gender, race, age, or BMI in our sample of pediatric patients. Identification of other factors that may be associated with PAP adherence by school-aged children and adolescent patients requires additional study.
We thank Yan Yan, MD, PhD, for constructing the Bland-Altman plot and providing other statistical analytical support.
- Accepted May 7, 2007.
- Address correspondence to Donna B. Jeffe, PhD, Siteman Cancer Center, Washington University School of Medicine, 4444 Forest Park Ave, Suite 6700, St Louis, MO 63108. E-mail:
The authors have indicated they have no financial relationships relevant to this article to disclose.
- ↵Reuveni H, Simon T, Tal A, Elhayany A, Tarasiuk A. Health care services utilization in children with obstructive sleep apnea syndrome. Pediatrics.2002;110 :68– 72
- ↵American Academy of Pediatrics, Section on Pediatric Pulmonology, Subcommittee on Obstructive Sleep Apnea Syndrome. Clinical practice guideline: diagnosis and management of childhood obstructive sleep apnea syndrome. Pediatrics.2002;109 :704– 712
- ↵Sullivan CE, Issa FG, Berthon-Jones M, Eves L. Reversal of obstructive sleep apnoea by continuous positive airway pressure applied through the nares. Lancet.1981;1 (8225):862–865
- ↵Massa F, Gonsalez S, Laverty A, Wallis C, Lane R. The use of nasal continuous positive airway pressure to treat obstructive sleep apnoea. Arch Dis Child.2002;87 :438– 443
- ↵Guilleminault C, Nino-Murcia G, Heldt G, Baldwin R, Hutchinson D. Alternative treatment to tracheostomy in obstructive sleep apnea syndrome: nasal continuous positive airway pressure in young children. Pediatrics.1986;78 :797– 802
- ↵Marcus CL, Rosen G, Ward SL, et al. Adherence to and effectiveness of positive airway pressure therapy in children with obstructive sleep apnea. Pediatrics.2006;117(3) . Available at: www.pediatrics.org/cgi/content/full/117/3/e442
- ↵Follansbee DS. Assuming responsibility for diabetes management: what age? What price? Diabetes Educ.1989;15 :347– 353
- ↵Christie D, Viner R. Adolescent development. BMJ.2005;330 :301– 304
- ↵Needlman RD. Growth and development: middle childhood. In: Behrman RE, Kliegman RM, Jenson HB, eds. Nelson Textbook of Pediatrics. 17th ed. Philadelphia, PA: Saunders; 2004:51– 53
- ↵Popescu G, Latham M, Allgar V, Elliott MW. Continuous positive airway pressure for sleep apnoea/hypopnoea syndrome: usefulness of a 2 week trial to identify factors associated with long term use. Thorax.2001;56 :727– 733
- ↵Bland JM, Altman DG. Measuring agreement in method comparison studies. Stat Methods Med Res.1999;8 :135– 160
- ↵Gutgesell ME, Payne N. Issues of adolescent psychological development in the 21st century. Pediatr Rev.2004;25 :79– 85
- Copyright © 2007 by the American Academy of Pediatrics