Home continuous positive airway pressure therapy in infants: a single-center experience

Abstract

Study Objectives:

There are limited data on indications and outcomes of home continuous positive airway pressure (CPAP) therapy in the first year of life. We aimed to analyze the clinical, demographic, and polysomnographic characteristics of a cohort of children initiated on home CPAP for treatment of sleep-disordered breathing and as respiratory support in the first year of life.

Methods:

Children started on CPAP in the first year of life at the Queensland Children’s Hospital were retrospectively evaluated for clinical and demographic parameters, underlying diagnoses, respiratory support, airway surgical intervention, and polysomnography results at baseline and on CPAP.

Results:

Twenty-nine infants (median age [interquartile range] at CPAP initiation, 182 days [126–265.5 days]) were included. The underlying etiology included Trisomy 21 (n = 6), craniofacial syndromes (n = 5), hypotonia (n = 8; 5 with noncraniofacial syndrome), airway malacia (n = 5), skeletal dysplasia (n = 2), nonsyndromic upper airway obstruction (n = 2), and chronic neonatal lung disease (n = 1). The median (interquartile range) obstructive apnea-hypopnea index was 14 events/h (6.2–31 events/h) at CPAP initiation, which improved on CPAP to 3.4 events/h (1.4–6.4 events/h). The median (interquartile range) transcutaneous CO₂ max remained unchanged on CPAP (56.6 mm Hg [49–66.5 mm Hg] pre-CPAP vs 54.9 mm Hg [47–62 mm Hg] on CPAP). Fifteen children needed surgical airway intervention (11 pre-CPAP and 4 post-CPAP). CPAP therapy could be successfully stopped in 9 children, 2 children needed tracheostomy, and 1 child died during the follow-up period.

Conclusions:

Home CPAP as respiratory support is an effective long-term therapy in infancy, and these patients can be weaned from CPAP therapy even if it was initiated early. Prospective studies with predefined criteria for CPAP initiation and cessation would help ascertain long-term outcomes in this poorly researched group.

Citation:

Joshi SS, Sivapalan D, Leclerc M-J, Kapur N. Home continuous positive airway pressure therapy in infants: a single-center experience. J Clin Sleep Med. 2023;19(3):473–477.

BRIEF SUMMARY

Current Knowledge/Study Rationale: The present study was done in a cohort of children who were initiated on home continuous positive airway pressure therapy in the first year of life. The clinical, demographic, and polysomnographic characteristics along with treatment outcomes were analyzed.

Study Impact: The results support the consideration for this modality of treatment as respiratory support even during infancy. Further prospective studies are required to ascertain the long-term outcome of sleep-disordered breathing in infants and assess the advantage of early treatment on neurocognitive development in this poorly researched age group.

INTRODUCTION

Sleep-disordered breathing (SDB) patterns in children younger than age 12 months include periodic breathing, apnea of prematurity, congenital central hypoventilation syndrome, central apnea, and obstructive sleep apnea.^1–3 Obstructive sleep apnea (OSA) is by far the most well studied of this group and is characterized by partial or complete airway obstruction that affects normal ventilation during sleep and/or disrupts the regular sleep pattern.³

While the reported prevalence of SDB in children varies depending on the definitions used, OSA is reported in approximately 1%–5% in children and affects children of all age groups.^3,4 The predisposing factors that contribute to OSA in children in the first year of life differ from those in older children, in whom airway obstruction usually stems from adenotonsillar hypertrophy. In infants, airway obstruction may be developmental due to neurological immaturity or secondary to other medical conditions such as craniofacial anomalies and skeletal and neuromuscular syndromes.⁵ Infants have a greater propensity for airway obstruction and gas exchange abnormalities, both anatomically and physiologically. In addition, a highly compliant chest wall predisposes to gas exchange abnormalities during even brief periods of airway obstruction.^6,7 The role of adverse ventilatory mechanics, especially of upper airways and the chest wall, the function of chemoreceptors, and altered arousal responses in infants contribute to breathing disturbances during sleep.^8,9 A predominant rapid eye movement (REM) sleep state distribution causing loss of muscle tone also exacerbates obstruction in infants.^10,11

Because of this interplay of structural and functional causes of airway obstruction in infancy, the surgical treatment options for OSA are limited at this age. Surgery is usually not recommended in this cohort due to the high risk of postoperative complications. Continuous positive airway pressure (CPAP) can therefore be used as a transition therapy to delay surgery while waiting for the infant to grow. Positive airway pressure therapy in the form of nasal CPAP via mask has been described as a viable modality in infants at risk for sudden infant death syndrome due to increased upper airway resistance.¹² Despite this research, there are meager data available documenting the use of positive airway pressure therapy in the treatment of SDB in infants.¹³

Our study aimed to describe the clinical, demographic, and polysomnographic (PSG) characteristics and outcomes of a cohort of children who were initiated on home CPAP therapy as respiratory support in the first year of life.

METHODS

From a cohort of children initiated on home nasal CPAP therapy at the Queensland Children’s Hospital from November 2014–October 2019, a subgroup of children in whom CPAP was initiated in the first year of life was retrospectively studied. This cohort included all infants started on long-term home CPAP in the entire state of Queensland. Data extracted from electronic chart review included age at CPAP initiation, sex, anthropometric parameters, underlying etiology, details of airway assessment, respiratory support prior to CPAP therapy, details of airway surgical intervention if any, and PSG parameters at baseline and on CPAP treatment.

The protocol in our unit for CPAP initiation includes the fitting of an appropriate interface to minimize mask leak. For infants not already initiated on CPAP in the intensive care unit, the mask fitting is done a day prior or the morning of the study for all infants referred from the inpatient unit or on the evening of the study for those referred from outpatient settings. Infants are studied using a full PSG setup, and pressures are titrated to attempt to overcome airway obstruction and normalize gas exchange. Some children undergo single-night PSG study (split PSG) in which an initial diagnostic polysomnogram is followed by a CPAP titration study the same night. The split-night study is considered valid if at least 2 REM sleep periods are captured in the diagnostic segment of the study. For this study, PSG parameters for the split study were analyzed separately as “diagnostic” for the time not on CPAP and as “CPAP initiation” while infants were on therapy. The positive airway pressure apnea-hypopnea index (AHI) is a composite AHI of all trialled pressures during the course of the polysomnogram.

Statistical analysis

Descriptive statistics (mean, median, standard deviation, and interquartile range [IQR]) were used for analysis and reporting of results. The respiratory and sleep variables from all infants were combined and averaged according to sleep state and event type. These were compared as per the type of sleep study: either diagnostic or CPAP titration. The differences between data from diagnostic studies and CPAP studies were analyzed using the paired t test. The study was approved by the Institutional Human Research Ethics committee (LNR/2020/QCHQ/60938).

RESULTS

Twenty-nine infants (median age, 182 days [IQR, 126–265.5 days]) were initiated on CPAP between November 2014 and October 2019 and were included in the analysis. Trisomy 21 was the most common underlying diagnosis (n = 6; 21%) with 5 children diagnosed with a known craniofacial syndrome. Central hypotonia (n = 8; 28%) was the most common associated comorbidity (Table 1).

Table 1.

Associated comorbid condition.

Underlying Condition	Number (%)
Trisomy 21	6 (20%)
Craniofacial syndrome	5 (17%)
Pfeiffer’s syndrome	1
Goldenhar syndrome	1
Crouzon syndrome	1
Pierre–Robin syndrome	1
Poland–Moebius syndrome	1
Central hypotonia*	8 (28%)
Smith–Magenis syndrome	1
Phelan–McDermaid syndrome	1
Rubenstein–Taybi syndrome	1
NECP2 syndrome	1
Trisomy 12p	1
No known syndrome	3
Large airway malacia†	5 (17%)
Laryngomalacia	2
Tracheomalacia/tracheo-bronchomalacia	2
Laryngo-tracheomalacia	1
Skeletal dysplasia	2 (7%)
Choanal stenosis	1 (3.5%)
Chronic neonatal lung disease	1 (3.5%)
Micrognathia	1 (3.5%)

*All children with Trisomy 21 also had central hypotonia but are not part of this group.

†Syndromic children with associated airway malacia are classified under the specific syndrome.

Prior to CPAP initiation, 18 (62%) infants were dependent on respiratory support (supplemental O₂ in 14 [48%], nasopharyngeal tube in 1 [3.5%], and both in 3 [10%]). Fifteen (52%) children underwent surgical airway intervention (11 pre-CPAP and 4 post-CPAP) for the upper airway obstruction. This included adenoidectomy in 5 (2 with aryepiglottoplasty), aryepiglottoplasty in 4, adenotonsillectomy in 2, tracheal granuloma removal in 2, and 1 each with tonsillectomy and choanal atresia dilatation along with stent insertion.

Twenty-four infants had a diagnostic polysomnogram before the CPAP titration study, 7 of whom were in split studies. Five infants (1 each with Trisomy 21, chronic neonatal lung disease, skeletal dysplasia, craniofacial syndrome, and syndromic [non–Trisomy 21] hypotonia) were initiated on CPAP in the pediatric intensive care unit without a pre-CPAP diagnostic polysomnogram. Post-CPAP paired PSG comparison was hence available in 24 children. Overall, the cohort had a median AHI of 21.55 events/h (IQR, 6.7–39.4 events/h) prior to CPAP initiation, which improved to 4.75 events/h (IQR, 2.9–7.7 events/h) on CPAP (P < .001) (Table 2).

Table 2.

PSG variables for diagnostic and CPAP studies.

Variables	Diagnostic Study (n = 24)	CPAP Study (n = 24)	P*
Sleep efficiency	76.5 (19.5)	86 (12)	≤.0001
% REM sleep	39 (12.25)	37 (9.75)	.753
Overall AHI	21.55 (32.73)	4.75 (4.78)	.0009
Obstructive AHI	13.85 (24.8)	3.4 (5)	.002
REM sleep AHI	44.5 (48.65)	9.8 (9.43)	.0005
NREM sleep AHI	3.15 (8.6)	1.6 (2.8)	.56
Nadir SpO2	78 (12)	83 (6.75)	.03
Mean SpO2	97.9 (1.7)	97.8 (1.8)	.5
TcCO2 max	56.6 (17.6)	54.9 (14.55)	.37
Mean TcCO2	48.2 (7.2)	46.4 (10.33)	.502

The data are presented as median (interquartile range). *Paired t test was used to calculate statistical significance for only 24 pairs in which two polysomnograms were available. AHI = apnea-hypopnea index, CPAP = continuous positive airway pressure, NREM = non-rapid eye movement, PSG = polysomnographic, REM = rapid eye movement.

A similar improvement was also noted for REM sleep stage AHI (median, 44.5 events/h (IQR, 16.05–64.7 events/h) in the diagnostic study vs 9.8 events/h (IQR, 5.8–15.2 events/h) on CPAP (P = .0005). However, there was no significant change in the non-REM sleep stage AHI between the diagnostic (median, 3.15 events/h [IQR, 1.18–9.78 events/h]) and the titration study (median, 1.6 events/h [IQR, 0.6–3.5 events/h]; P = .56). Significant improvement was also seen in obstructive AHI (pre-CPAP median, 13.85 events/h [IQR, 6.2–31 events/h] vs post-CPAP median, 3.4 evvents/h [IQR, 1.4–6.4 evvents/h]; P = .002).

A median CPAP pressure of 5.7 cm H₂O (range, 4–8 cm H₂O) was required. The most commonly used masks for delivering CPAP therapy in this age group were all nasal masks: the Mask Medic size 2 in 11 (39%) infants, the Respironics Child Profile Lite in 6 (21%) infants, and the Wisp Pediatric Bubble Mask in 7 (24%) infants. After discharge, 27 of 29 infants tolerated the CPAP through the nasal mask well at home. Two infants were unable to tolerate the CPAP therapy at home and were subsequently managed using supplemental subnasal O₂.

At censoring for the study, 15 (52%) infants were still using CPAP therapy with a median (range) duration of CPAP use of 26 (10–58) months with no major adverse effects reported. Nine (31%) were successfully weaned off CPAP by the median (range) age of 27 (7–45) months (3 with Trisomy 21, 2 of whom had adenoidectomy; 1 each with choanal stenosis, chronic neonatal lung disease, tracheomalacia associated with tracheo-esophageal fistula, Rubinstein–Taybi syndrome, nonsyndromic hypotonia, and MECP2 syndrome -methyl-CpG-binding protein 2) after a median (range) CPAP usage of 21 (4–35) months. Three infants continued to be symptomatic on CPAP and needed tracheostomy; one of these infants underwent jaw advancement surgery.

DISCUSSION

Our retrospective study on 29 infants initiated on home CPAP as respiratory support reports significant improvement in PSG parameters. The underlying etiology was varied, with Trisomy 21 being the most common underlying etiology; a high proportion of children also had associated central hypotonia. Our data support the use of home CPAP therapy as a viable option for respiratory support even during the first year of life.

Upper airway obstruction in infants differs from that in older children due to differences in airway structure and tone, lung mechanics (low functional residual capacity), REM sleep predominance, and immature variable ventilatory control. There are limited data related to the causes, management, and outcomes of OSA in this age group. In 92 infants with OSA, Adeleye et al¹⁴ reported Trisomy 21 as the most common comorbidity in this age group. In another retrospective study in 97 infants with OSA, Ramgopal and colleagues¹⁵ reported laryngomalacia (24%) and craniofacial abnormalities (16%) as commonly associated comorbid conditions of infantile OSA, with genetic abnormalities (53%) such as Trisomy 21 common.

In the present study, nasal CPAP was successfully used in 93% of infants needing CPAP support. In a similar study by McNamara and Sullivan,¹³ CPAP via nasal mask was used successfully in 18 of 24 infants, with pressures ranging from 4–6 cm H₂O. A previous study on 80 children (aged ≤ 15 years, including infants) by Waters et al¹⁶ reported nasal CPAP to be effective in treating OSA in children with anatomic facial abnormalities. Administration of nasal CPAP at a mean pressure of 8 cm H₂O for an average of 15 months eliminated the symptoms, signs, and PSG abnormalities of OSA in 90% of these children.

Children with craniofacial conditions are generally at increased risk for the development of SDB,¹⁷ but this population is complex because of its heterogeneity. Small oropharyngeal airway size, neuromotor deficits impairing the ability to maintain patent airway during sleep, increased upper airway collapsibility,¹⁸ and low muscle tone all contribute to the high prevalence of SDB in this group.¹⁹

Adenotonsillectomy is generally the first modality of treatment for clinically significant upper airway obstruction, which may improve upper airway patency enough to ameliorate or resolve the obstruction.¹⁷ Positive airway pressure therapy is also considered in patients with minimal adenotonsillar tissue or when a nonsurgical approach is preferred. Children with craniofacial conditions and those in the first year of life have multifactorial OSA because the obstruction is due to factors other than adenotonsillar hypertrophy alone and may occur at multiple sites within the airway. In addition to adenotonsillectomy, these children may need adjuvant surgical procedures such as supraglottoplasty, tongue base/reduction procedures, expansion sphincter/lateral pharyngoplasty, and mandibular distraction osteogenesis. Tracheotomy is reserved for children with severe OSA who have failed to respond to other treatment approaches. In 126 infants with OSA, Leonardis and colleagues²⁰ reported that younger infants (aged 0–3 months) were more likely to be managed by surgical interventions compared to the older infants; these procedures included supraglottoplasty in severe laryngomalacia, mandibular distraction for micrognathia, and adenoidectomy (with or without tonsillectomy).

Data on long-term outcomes (clinical and adherence) in pediatric cohorts on CPAP therapy are limited.²¹ In this cohort, one-third of the infants were successfully weaned off CPAP therapy by age 2 years. While there are sparse published data on the indication and timing of CPAP cessation in children, our group has previously reported a cohort of 24 children in whom CPAP could be successfully ceased.²¹ Improvement of upper airway tone combined with widening of the upper airway with age and improvement in airway malacia may have relieved the obstructive symptoms in some of these children.²² All 9 children in whom CPAP was electively ceased had subsequent polysomnograms.

The decision about the timing of the follow-up polysomnogram is made by the individual treating sleep physician. Based on symptoms and underlying etiology, the physician decides whether the study needs to be started as a split diagnostic study to understand whether there is ongoing need for CPAP support. Often these studies are then continued as diagnostic if no indication for CPAP reinitiation is found during the conduct of the study. Pediatric OSA is a distinct entity from adult OSA, highlighted by differences in reasons for upper airway obstruction, response to surgery, and growth-related improvement in airway caliber, making extrapolation from the adult literature difficult. Moreover, healthy infants in the first 30 days of life can have high rates of respiratory events on polysomnography.²³ Daftary and colleagues²³ reported an AHI of 14.9 events/h in a group of normal infants, a level that would otherwise classify as abnormal in an older child. This trend creates an added complexity in defining OSA in this age group.

The limitations of this study include single-center data, a small sample size, and the retrospective nature of study design. Details on adherence data were not available. Further, diagnostic polysomnography was not available for 5 infants, hence for paired statistical analysis, only 24 pairs were included. Despite these limitations, ours is one of the few studies to report the feasibility and outcomes of home CPAP therapy in infancy. Our results suggest a high acceptance rate and the need for close clinical follow-up to monitor changes with age and growth.

CONCLUSIONS

Based on our single-center experience on a cohort of infants on home CPAP therapy, we conclude that long-term CPAP therapy is a viable therapeutic option for respiratory support at home at this age. Further work is needed to determine the relationship between PSG results, outcomes, and the impact of treatment on a long-term basis. Prospective studies with predefined inclusion criteria and random allocation of treatment pathways will provide robust evidence in this underresearched area.

DISCLOSURE STATEMENT

All authors have seen and approved the manuscript. The authors report no conflicts of interest.

ABBREVIATIONS

AHI

apnea-hypopnea index

CPAP

continuous positive airway pressure

IQR

interquartile range

OSA

obstructive sleep apnea

PSG

polysomnographic

REM

rapid eye movement

SDB

sleep-disordered breathing

SpO₂

Saturation of peripheral oxygen

TcCO₂

total carbon dioxide