© 2002 American Thoracic Society
Use of Nasal Cannula for Detecting Sleep Apneas and Hypopneas in Infants and ChildrenService de Physiologie, Hôpital Robert Debré, Université Paris VII, INSERM E9935, Paris, France Correspondence and requests for reprints should be addressed to Ha Trang, M.D., Ph.D., Service de Physiologie Hôpital Robert Debré, 48 boulevard Serurier, 75019 Paris, France. E-mail: ha.trang{at}rdb.ap-hop-paris.fr
We evaluated tolerance of nasal cannula (NC) by 14 infants (median age, 2.6 months) and 16 children (median age, 5.5 years) with suspected obstructive sleep apnea syndrome and compared the efficacy of the NC with that of a nasobuccal thermistor in detecting obstructive apneas (OA) and obstructive hypopneas (OH) on polysomnography traces. The relationship between cannula flow and esophageal pressure was assessed in six patients. Time spent with an uninterpretable flow signal was longer when using a cannula than when using a thermistor in infants (p < 0.05) and children (p < 0.01), and it was longer in the younger patients (p < 0.05). Among the 650 OA–OH detected by either method, only 38% were detected by both, and 58% were detected by the cannula and missed by the thermistor, so that the apnea–hypopnea index was higher with cannula than with thermistor in each age group (p < 0.01). More hypopneas than apneas were detected by the cannula and missed by the thermistor (p < 0.001). Out-of-phase thoracic and abdominal motions and/or changes in the end-tidal CO2 signal shape were associated with 86% of OH identified by cannula. In the six patients whose esophageal pressure was measured, all respiratory events identified using a cannula were associated with increased "airway resistance." Thus, the NC is more likely than the thermistor to detect OA and OH in infants and children, and this superiority is particularly marked for hypopneas.
Key Words: apnea • hypopnea • nasal cannula • child • infant
Obstructive sleep apnea syndrome (OSAS) in children can manifest as brief apneas and hypopneas (1). More frequently, prolonged partial obstruction of the upper airway occurs during sleep, producing severe desaturation and hypercapnia, in the absence of a respiratory event clearly identified using a nasobuccal thermistor (NBT). This breathing pattern has been called "obstructive hypoventilation" (2). Thermistors detect the difference in temperature between inspired and expired air. Because of their long time-constant response, they may fail to reliably detect partial airway obstruction (3). Nevertheless, most pediatric sleep laboratories use NBT to measure airflow (4). Esophageal pressure (Pes) measurement, which is the standard of reference for evaluating respiratory effort, is too invasive for routine sleep studies in children. Over the last few years, a growing number of studies in adult patients have shown that using a nasal cannula (NC) may dramatically improve the detection of sleep-disordered breathing by identifying obstructive apneas (OA) and obstructive hypopneas (OH) that are missed by thermistors (5–11). The cannula is similar to oxygen nasal prongs. It is placed at the anterior nares and is connected to a pressure transducer capable of detecting pressure changes during inspiration and expiration. This method has not been evaluated in pediatric patients. The present study aimed to evaluate the tolerance of NC by infants and children during sleep and to compare the efficacy of the NC with that of the NBT in detecting OA and OH. The relationship between cannula inspiratory flow and Pes was also assessed in a small group of patients.
Fourteen consecutive infants of less than 1 year of age and 16 consecutive children referred to our hospital for suspected OSAS were studied. The appropriate ethics committee approved the study, and informed consent was obtained from the parents of all patients. Polysomnographic traces were recorded during an afternoon nap for infants and during nocturnal sleep for children. Patients slept in their preferred position, except the infants with Pierre Robin sequence, who were studied in the prone position. A trained sleep technician was present throughout the recordings. For all patients, airflow was recorded using both an NC connected to a 2-cm H2O pressure transducer (Protech, Minneapolis, MN) and an NBT (Respironics, Murrysville, PA). Cannula and thermistor sizes were chosen to match the size of the patient's nostrils as closely as possible, and the devices were carefully secured to the face with tape. The classic neurophysiologic and respiratory signals were also recorded: electroencephalogram (C4/A1, C3/A2), electro-oculograms, submental electromyogram (EMG), and diaphragmatic EMG using external electrodes placed in the eight right intercostal space on the anterior and midaxillary lines (12), thoracic and abdominal respiratory motions using strain gauges, oxygen saturation (SaO2) measured by pulse oximeter set at 2-second averaging time (Nellcor, Hayward, CA), end-tidal expired CO2 (PETCO2) measured breath-by-breath using a 5F sampling catheter placed at the anterior nare (Novametrix, Wallingford, CT), and body movements recorded using an actimeter. All data were recorded using a computer-assisted system (Alice; Respironics). In addition, Pes was measured in six patients whose parents gave consent for catheter placement. A flexible, small (1.67 mm outer diameter, size 5F) catheter equipped with a pressure transducer (Gaeltec, Dunvegan, UK) was placed transnasally before the sleep recording and positioned in the lower third of the esophagus. Calibration was achieved at the end of the polysomnographic study by immersing the catheter into a tube 2 cm in diameter. The signal from the pressure transducer was recorded as immersion depth was increased in steps of 5 cm, from 0 to 70 cm (13).
Analysis The quality of the airflow signals was examined on the entire polysomnography trace for all patients. An uninterpretable airflow signal was defined as no airflow signal during 30 seconds of normal respiration (i.e., normal flow signal by the other method, unchanged respiratory motions, normal SaO2 and PETCO2 signal, as compared with the 10 preceding respiratory cycles). Time spent with an uninterpretable airflow signal was expressed as percent of total sleep time. Respiratory events were scored on data from the entire nap in infants or from 2 consecutive hours of continuous sleep with no more than 10 minutes of wakefulness and with respiratory signals of good quality in older children. Using a blind procedure, one of us scored respiration twice, once with the cannula signal visible and once with the thermistor signal visible, the other signal being deleted from the computer screen before scoring. The other respiratory variables displayed were thoracic and abdominal respiratory motions, SaO2, PETCO2 signal, and diaphragmatic EMG. The Pes signal was deleted from the computer screen during this phase of scoring. Respiration was scored on 30-second epochs in infants and on 60-second epochs in children. In this study, we considered only obstructive respiratory events (i.e., events with persistent or increased respiratory motions and/or diaphragmatic EMG activity) that lasted at least 3 seconds in infants (16) and 5 seconds in children (17). OA was defined as a complete absence of airflow. OH was defined as (1) a 50% or greater decrease in peak airflow amplitude compared with baseline (4), or (2) a discernible decrease of less than 50% in the airflow amplitude, with a 3% or greater drop in SaO2 and/or with termination by arousal (18). Arousal was defined as an electroencephalogram and/or an EMG and/or a movement arousal longer than 1 second (19). The presence or absence of concomitant changes in the other respiratory parameters during each respiratory event was noted. Particular attention was given to detect out-of-phase respiratory motions of the thorax and abdomen. The amplitude and the configuration of the PETCO2 signal (reduced amplitude, notched signal contour) were of more importance than the PETCO2 absolute value (20). The apnea–hypopnea index was the number of OA–OH (OAH) per hour of sleep analyzed. The duration of respiratory events was measured from one airflow signal amplitude peak to the next.
We analyzed the relationship between NC inspiratory flow and Pes in the six patients whose Pes was measured. Respiratory cycles preceding and following OA and OH detected by cannula were selected for this analysis. Because use of a pressure transducer–mounted catheter is associated with a drift in the zero value that is small but sometimes non-negligible during prolonged recording, absolute Pes values were not considered (21). Pes was expressed in centimeters of H2O, and cannula flow (Flow) in arbitrary units. For a given respiratory cycle,
Statistics
We studied 30 patients: 14 infants (median age, 2.6 months; range 0.5–9.4) and 16 children (median age, 5.5 years; range, 1.4–14.8). The most frequent underlying diseases were craniofacial malformations (n = 15), oropharyngeal abnormalities (n = 5), prematurity with suspected apneas (n = 5), and obesity (n = 3) (see Table E1 in the online data supplement).
Tolerance of NC
Cannula signal unreliability was caused by displacement of the cannula outside the nares or its obstruction by the inner nasal wall, prompting repositioning of the cannula by the technician during sleep recordings. This problem was most common in the youngest infants and in those studied in the prone position. Another frequent cause was occurrence of mouth breathing. Chewing movements may produce artifacts on the cannula signal.
Detection of OA
Detection of OH NC detected all 159 OH observed overall, whereas the thermistor detected only 14% of these events. The proportion of events detected by the cannula and missed by the thermistor was considerably higher for hypopneas than for apneas in each age group (p < 0.001). The ratio of hypopneas to apneas was far higher in children (50%) than in infants (25%). Decrease in cannula signal amplitude was greater than 50% compared with baseline for 97% of hypopneas in infants and 86% in children (Table 2).
In the younger infants, OH were associated with a deep reduction in the cannula signal amplitude with no marked changes in its shape (Figure 2A) . In contrast, in the children and older infants, changes in the cannula signal shape were common during hypopneas, often preceding or accompanying the reduction in flow amplitude; these changes consisted of peaks and plateaus during parts of the inspiration or the entire inspiration (Figure 2B). The decrease in cannula signal amplitude sometimes preceded the decrease in thermistor signal amplitude by one or many respiratory cycles. The profiles of the cannula inspiratory flow shapes were similar in boys and girls.
Concomitant changes in the other respiratory signals were frequently associated with OH. Partial or complete loss of phase of thoracic and abdominal motions and/or changes in the PETCO2 signal shape were noted in 83% of OH in infants and in 89% of OH in children (Table 2).
Detection of OAH
NC Flow and Pes Relationship Pes recordings were obtained in six patients: four infants and two children. In these patients, all respiratory events identified by the cannula were associated with an increase in the "airway resistance." Figure 5 shows an OA in an infant, with a reduction in  Pes and a considerably larger reduction in Flow; airway resistance increased 12-fold as compared with the previous respiratory cycles. Meanwhile, Pes increased during an OH in a child. The crescendo Pes pattern terminated with an arousal associated with large variations in Pes waves and increased breaths (see also Figure E1 in the online data supplement).
In the present study, we found that the NC can be used in infants and children during sleep and is more likely than the NBT to detect obstructive respiratory events, with the difference being more marked for OH than for OA. The NC was accepted by all patients during sleep, with a signal of good quality in most of them. However, the time spent with an uninterpretable airflow signal was significantly longer when using a cannula than when using a thermistor, both in infants and children. Difficulties in maintaining the cannula in proper position occur occasionally, requiring repositioning of the cannula by the technician. Another frequent cause is the occurrence of mouth breathing. It is likely that this may, at least in part, account for prolonged loss in the cannula signal observed in some of our patients, despite the presence of a trained sleep technician. This is a potential limitation of airflow measurement should the cannula be used alone, and may affect respiratory event recognition. However, given previous evidence that repetitive periods of mouth breathing and absent or reduced nasal airflow may be related to increased upper-airway resistance and associated with clinical abnormalities (1), some authors suggested that this limitation is likely counterbalanced by the more accurate airflow measurement provided by this technique (6). Others advocate systematic use of an additional oral thermistor. As previously reported in adults, the NC recording in our pediatric patients also detected many OA and OH missed by the thermistor. Nearly 60% of OAH identified by the cannula were missed by the thermistor, and this proportion rose to 86% when only OH were considered. The cannula showed clear evidence of OH during periods that would have been labeled "obstructive hypoventilation" on the basis of thermistor data only. Therefore, the thermistor used alone may result in underdiagnosis or underestimation of sleep-disordered breathing, with the potential high risk for severe complications (1). In the small group of patients who also underwent Pes measurement, all respiratory events detected by the cannula were associated with changes in Pes swings and an increase in the airway resistance. This indicates that they were not overdiagnosed by the cannula. In the others, events detected by the cannula were generally associated with concomitant changes in other respiratory parameters (out-of-phase thoracic and abdominal motions, reduced amplitude and notched contour of PETCO2 signal) or were accompanied by desaturation or termination by arousal. The usefulness and limitations of each of these techniques for diagnosing apneas have been discussed elsewhere (23). The relationships between Pes and airflow measured by the NC were very similar to those previously described in pediatric patients using other techniques (24–26). We also found that the airway resistance can increase severalfold during OH. This increase is probably larger because the nonlinearity of NC traces results in overestimation of airflow amplitude as compared with that measured with a true pneumotachograph (8). Pes changes between inspiration and expiration can decrease or vary during obstructed respiratory cycles in infants. In contrast, children exhibit a pattern much closer to that observed in adult patients, with crescendo Pes variations, and peaks and plateaus on the cannula inspiratory flow trace. The breathing patterns observed in our patients are in accordance with the current concept of a greater upper airway collapsibility in infants than in children (1, 27, 28). Previous studies showed that infants had higher closing pressure and higher compliance of the pharynx at birth that progressively decrease with increasing age, whereas children have a relatively more stable upper airway. This may explain why infants are more prompt in developing apneas than in developing hypopneas as compared with children, and why children may present hypopneas with a flattened inspiratory flow shape. Results of the present study indicate that the NC is more likely than the thermistor to detect obstructive respiratory events in infants and children during sleep. However, there are important considerations before recommending routine use of NC in pediatric patients. It is of interest to determine whether this device can increase airway obstruction per se, particularly in the youngest with very small nares or narrow upper airways due to oropharyngeal or craniofacial malformations. In a previous study, the cannula has been shown to produce a small increase in nasal resistance in normal adults during wakefulness (29) but is believed not to have much influence on the breathing of adult patients with sleep problems. This evaluation needs to be performed in pediatric patients. Furthermore, large multicenter studies are required to obtain normative values of OA and OH identified on NC traces in infants and children and to determine the optimal cutoff values for separating patients with and without OSAS. Another important issue is whether the NC is useful in pediatric patients for identifying more subtle respiratory events, such as those associated with upper-airway resistance syndrome (25) and those called respiratory effort–related arousals in adult patients (7, 11). In conclusion, the present study demonstrates that the NC is effective in infants and children for detecting OA and OH. To avoid overdiagnosis, the authors recommend that respiratory events identified on NC traces be supported by concomitant changes in other respiratory parameters and/or by the concomitant occurrence of desaturation or arousal.
This article has an online data supplement, which is accessible from this issue's table of contents online at www.atsjournals.org Received in original form October 31, 2001; accepted in final form May 21, 2002
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
INTRODUCTION
