|Year : 2021 | Volume
| Issue : 4 | Page : 70-76
Use of pulse oximetry to screen for infant obstructive sleep apnoea
Andy Cheuk-ting Hou, Eric Yat-tung Chan, Ka-li Kwok, Mei-yee Lau, Shuk-yu Leung
Department of Paediatrics, Kwong Wah Hospital, Hong Kong SAR, China
|Date of Submission||31-Mar-2022|
|Date of Decision||27-May-2022|
|Date of Acceptance||13-Jun-2022|
|Date of Web Publication||23-Sep-2022|
Andy Cheuk-ting Hou
Department of Paediatrics, Kwong Wah Hospital, 25 Waterloo Road, Hong Kong SAR
Source of Support: None, Conflict of Interest: None
Introduction: Pulse oximetry is currently used to screen for obstructive sleep apnoea (OSA) in children. However, its use in infant has not yet been well studied. Aim: The aim of this study was to develop a screening criterion using pulse oximetry to identify infant with probable OSA. Materials and Methods: This was a retrospective cross-sectional study including infants <1 year of age with features of upper airway obstruction or requiring home oxygen to find associations between obstructive apnoea hypopnoea index (OAHI) in infant polysomnography (PSG) and parameters in pulse oximetry by Spearman Rho’s correlation. The factor with the strongest correlation is further analysed by receiver-operating characteristic (ROC) curve to identify a cutoff with highest Youden index to screen for probable OSA (OAHI >2 per hour). Results: A total of 27 infants were studied. The index of oxygen desaturation with SpO2 <90% per sampled hour (ODI<90%) had the best correlation with OAHI (r = 0.52, P = 0.005). Using the cutoff of ODI<90% more than 1.3 per hour, the sensitivity and specificity for identifying OAHI >2 per hour was 77% and 71%, respectively. Conclusion: Infant pulse oximetry can be a useful tool to screen for probable infant OSA especially for paediatric units not offering infant PSG service.
Keywords: Infant, obstructive sleep apnoea, oximetry
|How to cite this article:|
Hou AC, Chan EY, Kwok Kl, Lau My, Leung Sy. Use of pulse oximetry to screen for infant obstructive sleep apnoea. Pediatr Respirol Crit Care Med 2021;5:70-6
|How to cite this URL:|
Hou AC, Chan EY, Kwok Kl, Lau My, Leung Sy. Use of pulse oximetry to screen for infant obstructive sleep apnoea. Pediatr Respirol Crit Care Med [serial online] 2021 [cited 2022 Dec 3];5:70-6. Available from: https://www.prccm.org/text.asp?2021/5/4/70/356804
| Introduction|| |
Obstructive sleep apnoea (OSA) has been reported to be associated with sudden infant death syndrome long ago.,, Recent studies reported infants with snoring or OSA had poorer neurocognitive outcomes.,,, Polysomnography (PSG) remains the gold standard to diagnose infant with OSA., Yet the availability of infant PSG in paediatric units is limited due to its sophisticated and labour-intensive nature. Pulse oximetry was introduced as a screening test for OSA in children in 2000. It gains popularity in research and clinical use in the past decades., Recent research further focuses on its simplicity and accuracy., However, publications on pulse oximetry to screen for infant OSA and its clinical use is very limited. The aim of this study was to develop a screening criterion using pulse oximetry to identify infant with probable OSA for earlier referral to diagnosis and intervention.
| Materials and Methods|| |
This was a retrospective cross-sectional study conducted in Kwong Wah Hospital (Hong Kong) from December 2019 to December 2020. The study included infants aged 12 months or below. These subjects simultaneously underwent a PSG and a separate pulse oximetry study. Infants with cyanotic heart diseases or severe hypoxic ischemic encephalopathy were excluded from the study.
PSG was performed by qualified sleep technologists in sleep laboratory with a digitized system (Siesta, Profusion 3 Software, Compumedics, Australia). The standard infant montage was used and the following parameters were recorded during the study: electroencephalogram (EEG) F3-M2, F4-M1, O1-M2, O2-M1, C4-M1, C3-M2; electrooculogram (EOG); submental, tibial, and intercostal electromyogram (EMG); electrocardiography (ECG); airflow with nasal pressure transducer and oral thermal sensor; oxygen saturation and pulse waveform by build-in pulse oximeter; carbon dioxide level by transcutaneous carbon dioxide monitor (TCM 4, Radiometer, Copenhagen, Denmark); rib cage and abdominal volume changes using respiratory inductance plethysmograph (RIP); sleep characteristics by position sensor, snoring microphone and video monitoring using an infrared video camera. Infant and paediatric sleep staging rules of American Academy of Sleep Medicine (AASM) manual version 2.4 were observed. Infants staging was done in children younger than 2 months and the child staging to older children. AASM criteria for obstructive apnoea and obstructive hypopnoea were employed.
Pulse oximetry study was recorded by Masimo Radical-7 pulse-oximeter. Data were downloaded to PROFOX software for analysis. Periods with low Signal IQ (Signal Identification and Quality indicator) were excluded from interpretation as artefacts. Awake periods when the infant’s eyes were opened or during crying or feeding were manually excluded. This is done by one same independent observer, not involved in PSG interpretation, using 5–10 min, to quickly screen through the video captured by the infrared camera recorded during the pulse oximetry study and PSG.
Several oximetry parameters were analysed and are listed in [Table 1]. These included ODI40: index of oxygen desaturation ≥4% from baseline (ODI4) for events lasting more than 0 s per sampled hour; ODI410: ODI4 for events lasting more than or equal to 10 s per sampled hour; ODI30: index of oxygen desaturation ≥3% from baseline (ODI3) for events lasting more than 0 s per sampled hour; ODI310: ODI3 for events lasting more than or equal to 10 s per sampled hour; ODI<90%: index of oxygen desaturation with SpO2 <90% per sampled hour; cluster of 5: ≥5 drops of SpO2 ≥4% from baseline within 30 min; cluster of 3: ≥3 drops of SpO2 ≥4% from baseline within 30 min; McGill score by Nixon in 2004.
All statistical analyses were performed by using IBM SPSS Statistics for Windows, Version 25.0. Separate Spearman Rho’s correlation for each mentioned oximetry parameters with OAHI were determined. Receiver-operating characteristic (ROC) curves were plotted for the oximetry parameter with the most significant Spearman’s Rho’s correlation to OAHI. The cutoff level with the optimal combination of sensitivity and specificity was calculated using the Youden index. The area under curve (AUC) reflects the accuracy of the oximetry parameters in predicting OAHI >2 per hour and >5 per hour in PSG. A value of P < 0.05 was considered statistically significant.
This study was approved by the Hong Kong Hospital Authority Kowloon Central Cluster Ethics Committee.
| Results|| |
This study included 27 infants consisted of 17 boys and 10 girls, of which 17 were ex-preterm infants. The median chronological age was 122 days and the median corrected age was 105 days. The indications for PSG were moderate-to-severe bronchopulmonary dysplasia (BPD) (n = 13), upper airway obstruction (n = 13), and recurrent clinical apnoea (n = 1).
Median total sleep time, study time, and sleep efficiency of the PSGs were 174 min, 219 min, and 80.5%, respectively. The median obstructive apnoea hypopnoea index (OAHI) of the subjects was 1.8 per hour.
For pulse oximetry, the median study time was 244 min, the median time excluded as artefacts, and awake period was 12 min. The median of ODI40 was 13.2 per hour; ODI410 was 5.1 per hour; ODI30 was 24.0 per hour; ODI310 was 10.0 per hour; ODI<90% was 1.4 per hour; cluster of 5 was 0.8 per hour; and cluster of 3 was 1.2 per hour. Using McGill scoring, 9 infants scored 1; 11 infants scored 2; and 7 infants scored 3. These are summarized in [Table 2].
|Table 2: Demographics; infant polysomnography and pulse oximetry parameters|
Click here to view
The Spearman’s Rho correlation between each oximetry parameter and OAHI (with p-value <0.05) were as follows: 0.498 for ODI40; 0.413 for ODI410; 0.427 for ODI30; 0.520 for ODI<90%; 0.493 for cluster of 5 per hour; and 0.414 for McGill score. The results of ODI310 and cluster of 3 per hour were statistically insignificant. These are summarized in [Table 3]. ODI <90% had the highest correlation with OAHI (r = 0.52, P = 0.005). This is shown in [Figure 1].
ROCs were derived from logistic regression models with OAHI >2 per hour and >5 per hour as response variables and ODI<90% as covariate. The AUC for OAHI >2 per hour was 0.77 (P = 0.016). Using the cutoff of ODI<90% more than 1.3 per hour, the sensitivity and specificity for identifying OAHI >2 per hour was 77% and 71%, respectively. This is shown in [Figure 2]A and [Table 4]A. The AUC for OAHI >5 per hour was 0.98 (P = 0.008). Using the cutoff of ODI<90% more than 3.8 per hour, the sensitivity and specificity for identifying OAHI >5 per hour was 100% and 91.7%, respectively. This is shown in [Figure 2]B and [Table 4]B.
|Figure 2: (A) ROC curves with OAHI >2 and ODI<90%. (B) ROC curves with OAHI >5 and ODI<90%|
Click here to view
|Table 4A: Receiver-operating characteristic (ROC) curve analysis of ODI<90% for detecting OAHI>2/h|
Click here to view
|Table 4B: Receiver-operating characteristic (ROC) curve analysis of ODI<90% for detecting OAHI>5/h|
Click here to view
| Discussion|| |
This study showed that video-adjusted infant pulse oximetry had a significant correlation with in-hospital PSG. Index of oxygen desaturation with SpO2 <90% per sampled hour (ODI<90%) had the best correlation with OAHI. Non-video adjusted oximetry with automatic artefacts exclusion may still resulted in SpO2 drops and SpO2 nadir much lower than the baseline leading to questionable preciseness.
Yet there is not a consensus on defining infant OSA. American Thoracic Society (ATS) suggested that an infant found to have an apnoea–hypopnoea index (AHI) greater than 2 per hour should alert physicians as to the probable presence of significant OSA. European Respiratory Society practice statement for obstructive sleep disordered breathing in 1- to 23-month-old children defined OSA with OAHI ≥1 per hour. Kato and Schlüter reported obstructive apnoea was rarely seen in healthy term infants in their publications in early 2000s., However, hypopnoea index (HI) was not reported in both studies. Kato explained that their PSG recording techniques in 1991 to 1993 did not allow for identification of such events. In 2013, 2015 and 2019, Brockmann, Duenas-Meza and Daftary, respectively, published new reference values for infant PSG from ‘healthy’ term subjects.,, Data on HI were included in these three studies. The median OAHI ranged from 0.5 per hour to 7.7 per hour, in which younger infants has higher median OAHI. The significant diversified findings between earlier and later publications were partly contributed by the advancement in technology and equipment sensitivity nowadays compared to the first infant PSG in Jan 1989 by Schlüter. Moreover, there had been changes in the AASM scoring criteria for defining an obstructive apnoea event and an obstructive hypopnoea event throughout the past two decades.,,, Some otorhinolaryngologists has been using OAHI ≥1.5 to define infant OSAS in their publications.,, Up till now, we still lack a universally adopted cutoff for defining OSAS in infants. From current evidence, we can predict the ‘normal’ OAHI in infant is likely represented by an inverse proportion curve, and join as a continuum with the cutoff of OAHI <1 in children. This study hence adopted a fixed cutoff of OAHI >2 to screen for presence of probable OSA among infants.
Ehsan included 38 infants over their 8-year study period. They have shown that ODI from home- or hospital-based oximetry has a significant positive correlation with OAHI in PSG for young infants with symptoms or risk factors for sleep disordered breathing. An average of 3 hours data were excluded due to artefact in each oximetry of their subjects. They reported ODI40 >3 is useful to screen for OAHI>5 in symptomatic infants (noisy breathing/snoring or history of ALTE/BRUE) with a sensitivity of 100%, specificity of 35%. They showed that ODI40 was not a good parameter to screen for milder OSA.
In this study, ODI<90%, McGill score and cluster of 5 per hour are good predictors of probable OSA. Why ODI<90% is chosen for further analysis is because of its highest correlation coefficient and its simplicity in data collection. This study showed that ODI<90% of >1.3 per hour, can detect OAHI >2 per hour with a sensitivity and specificity of 77% and 71%, respectively. The strength of this current study is standardization, simultaneousness and conveniency. All our infant PSGs were performed and scored by a standardized protocol in accordance to AASM manual version 2.4. The interpretation is done by the same group of qualified Paediatric Respiratory Medicine specialist with a decade or more experience in sleep medicine. Both PSG and oximetry of each subject were performed simultaneously but independently. Our oximetry recording technically started earlier and ended later than PSG; this is represented by the discrepancy in the reported study time. In daily practice, most pulse oximetry machines can automatically count for the number of times that SpO2 drops below 90%. Thus, it is convenient to use the suggested ODI<90% cutoff as a screening tool.
There are several limitations in this study. First, the small sample size of 27 infants. Second, our PSG recorded a total sleep time close to the lower margin of the median range (2.5 to 6 hours) reported in those normal value studies;,,, however, each of our PSG has included portions of REM and non-REM (active and quiet) sleep. Third, we only recruited infants <1 year of age with features of upper airway obstruction or requiring home oxygen in this study, this may cause selection bias in our findings. Hence generalization of its utility to healthy infants may not be congruent.
Galway reported that 59% of children screened positive by pulse oximetry for OSA on at least one of the three nights compared with 38% if only one night had been performed. If oximetry is abnormal on the first night there is no requirement to do further recordings on subsequent nights. They also found that reducing the threshold duration for technically adequate oximetry traces to a minimum of 4 h increased the number of patients who would have been screened positive for OSA by McGill classification.
For screening infant at risk of OSA, we suggest a video-taped pulse oximetry of 3–4 hours with accurate artefact exclusion. Obvious awake periods where the infant is crying or feeding should be manually excluded by quick screening through the recorded video. Repeat pulse oximetry in screened negative but clinically high-risks patient may help to reduce the possible ‘first night effect’ though this phenomenon has not been studied in infants., Any screened positive infants should be referred to have a formal infant PSG for OSA confirmation and further management.
Home oximetry screening for OSA in Hong Kong children is not as popular as foreign countries. Infant home oximetry screening is even rare. Unobserved pulse oximetry in young infants usually posted us with great difficulty to distinguish motion artefacts from genuine desaturation events., The aim of this study was to provide a solution to paediatric units without infant PSG service to identify babies at risk of OSA by employing observed in-hospital oximetry. A video-taped home oximetry allowing accurate artefact exclusion may be considered as an alternative, although further study with similar methodology should be carried out in home setting for verification. Directly applying the result of this study for unobserved oximetry to screen for OSA in infants may not be accurate.
In conclusion, video-assisted pulse oximetry using ODI<90% is a simple parameter to predict probable OSA with a relatively good sensitivity and specificity. Although more effort is spent in editing a video adjusted oximetry, less-sophisticated tools and time is required in performing and reporting an infant PSG.
Financial support and sponsorship
This study was supported by the Tung Wah Group of Hospitals Research Fund 2020/2021.
Conflicts of interest
The authors declare that they have no conflicts of interests.
The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Hong Kong Hospital Authority Kowloon Central Cluster Ethics Committee.
Not applicable to this study.
| References|| |
Guilleminault C, Ariagno R, Korobkin R, Nagel L, Baldwin R, Coons S, et al
. Mixed and obstructive sleep apnea and near miss for sudden infant death syndrome: 2. Comparison of near miss and normal control infants by age. Pediatrics 1979;64:882-91.
Kahn A, Groswasser J, Rebuffat E, Sottiaux M, Blum D, Foerster M, et al
. Sleep and cardiorespiratory characteristics of infant victims of sudden death: A prospective case-control study. Sleep 1992;15:287-92.
Kato I, Groswasser J, Franco P, Scaillet S, Kelmanson I, Togari H, et al
. Developmental characteristics of apnea in infants who succumb to sudden infant death syndrome. Am J Respir Crit Care Med 2001;164:1464-9.
Montgomery-Downs HE, Gozal D. Snore-associated sleep fragmentation in infancy: Mental development effects and contribution of secondhand cigarette smoke exposure. Pediatrics 2006;117:e496-502.
Piteo AM, Kennedy JD, Roberts RM, Martin AJ, Nettelbeck T, Kohler MJ, et al
. Snoring and cognitive development in infancy. Sleep Med 2011;12:981-7.
Smith CB, Walker K, Badawi N, Waters KA, MacLean JE. Impact of sleep and breathing in infancy on outcomes at three years of age for children with cleft lip and/or palate. Sleep 2014;37:919-25.
Bandyopadhyay A, Harmon H, Slaven JE, Daftary AS. Neurodevelopmental outcomes at two years of age for premature infants diagnosed with neonatal obstructive sleep apnea. J Clin Sleep Med 2017;13:1311-7.
Crowell DH, Brooks LJ, Colton T, Corwin MJ, Hoppenbrouwers TT, Hunt CE, et al
. Infant polysomnography: Reliability. Collaborative home infant monitoring evaluation (Chime) steering committee. Sleep 1997;20:553-60.
Crowell DH, Kulp TD, Kapuniai LE, Hunt CE, Brooks LJ, Weese-Mayer DE, et al
; CHIME Study Group. Infant polysomnography: Reliability and validity of infant arousal assessment. J Clin Neurophysiol 2002;19:469-83.
Brouillette RT, Morielli A, Leimanis A, Waters KA, Luciano R, Ducharme FM. Nocturnal pulse oximetry as an abbreviated testing modality for pediatric obstructive sleep apnea. Pediatrics 2000;105:405-12.
Álvarez D, Alonso-Álvarez ML, Gutiérrez-Tobal GC, Crespo A, Kheirandish-Gozal L, Hornero R, et al
. Automated screening of children with obstructive sleep apnea using nocturnal oximetry: An alternative to respiratory polygraphy in unattended settings. J Clin Sleep Med 2017;13:693-702.
Crespo A, Álvarez D, Kheirandish-Gozal L, Gutiérrez-Tobal GC, Cerezo-Hernández A, Gozal D, et al
. Assessment of oximetry-based statistical classifiers as simplified screening tools in the management of childhood obstructive sleep apnea. Sleep Breath 2018;22:1063-73.
Vaquerizo-Villar F, Álvarez D, Kheirandish-Gozal L, Gutiérrez-Tobal GC, Barroso-García V, Crespo A, et al
. Wavelet analysis of oximetry recordings to assist in the automated detection of moderate-to-severe pediatric sleep apnea-hypopnea syndrome. PLoS One 2018;13:e0208502.
Jiménez-García J, Gutiérrez-Tobal GC, García M, Kheirandish-Gozal L, Martín-Montero A, Álvarez D, et al
. Assessment of airflow and oximetry signals to detect pediatric sleep apnoea-hypopnoea syndrome using adaboost. Entropy (Basel) 2020;22:670.
Ehsan Z, He S, Huang G, Hossain MM, Simakajornboon N. Can overnight portable pulse oximetry be used to stratify obstructive sleep apnea risk in infants? A correlation analysis. Pediatr Pulmonol 2020;55:2082-8.
Kline C, Krupski T. Infant and toddler polysomnography. Respir Care Clin N Am 2006;12:1-10.
Berry RB, Brooks R, Gamaldo C, Harding SM, Lloyd RM, Quan SF, et al
. Aasm scoring manual updates for 2017 (version 2.4). J Clin Sleep Med 2017;13:665-6.
Nixon GM, Kermack AS, Davis GM, Manoukian JJ, Brown KA, Brouillette RT. Planning adenotonsillectomy in children with obstructive sleep apnea: The role of overnight oximetry. Pediatrics 2004;113:e19-25.
Katz ES, Mitchell RB, D’Ambrosio CM. Obstructive sleep apnea in infants. Am J Respir Crit Care Med 2012;185:805-16.
Kaditis AG, Alonso Alvarez ML, Boudewyns A, Abel F, Alexopoulos EI, Ersu R, et al
. ERS statement on obstructive sleep disordered breathing in 1- to 23-month-old children. Eur Respir J 2017;50:1700985.
Kato I, Franco P, Groswasser J, Kelmanson I, Togari H, Kahn A. Frequency of obstructive and mixed sleep apneas in 1,023 infants. Sleep 2000;23:487-92.
Schlüter B, Buschatz D, Trowitzsch E. Polysomnographic reference curves for the first and second year of life. Somnologie 2001;5:3-16
Brockmann PE, Poets A, Poets CF. Reference values for respiratory events in overnight polygraphy from infants aged 1 and 3months. Sleep Med 2013;14:1323-7.
Duenas-Meza E, Bazurto-Zapata MA, Gozal D, González-García M, Durán-Cantolla J, Torres-Duque CA. Overnight polysomnographic characteristics and oxygen saturation of healthy infants, 1 to 18 months of age, born and residing at high altitude (2,640 meters). Chest 2015;148:120-7.
Daftary AS, Jalou HE, Shively L, Slaven JE, Davis SD. Polysomnography reference values in healthy newborns. J Clin Sleep Med 2019;15:437-43.
Hirshkowitz M. Polysomnography: Understanding this technology’s past might guide future developments. Ieee Pulse 2014;5:26-8.
Sleep-related breathing disorders in adults: Recommendations for syndrome definition and measurement techniques in clinical research. The Report of an American Academy of Sleep Medicine Task Force. Sleep 1999;22:667-89.
Wiater A, Niewrth HJ, Pediatric Task Force in the German Sleep Society (DGSM). Polysomnographic standards for infants and children. Somnologie 2000;4:39-42.
Iber C, Ancoli-Israel S, Chesson AL, Quan SF. for the American Academy of Sleep Medicine. The AASM Manual for the Scoring of Sleep and Associated Events: Rules, Terminology and Technical Specifications. 1st ed. Westchester, IL: American Academy of Sleep Medicine; 2007.
Berry RB, Budhiraja R, Gottlieb DJ, Gozal D, Iber C, Kapur VK, et al
; American Academy of Sleep Medicine. Rules for scoring respiratory events in sleep: Update of the 2007 AASM manual for the scoring of sleep and associated events. Deliberations of the sleep apnea definitions task force of the American academy of sleep medicine. J Clin Sleep Med 2012;8:597-619.
Muntz H, Wilson M, Park A, Smith M, Grimmer JF. Sleep disordered breathing and obstructive sleep apnea in the cleft population. Laryngoscope 2008;118:348-53.
Robison JG, Wilson C, Otteson TD, Chakravorty SS, Mehta DK. Increased eustachian tube dysfunction in infants with obstructive sleep apnea. Laryngoscope 2012;122:1170-7.
Leonardis RL, Robison JG, Otteson TD. Evaluating the management of obstructive sleep apnea in neonates and infants. Jama Otolaryngol Head Neck Surg 2013;139:139-46.
Galway NC, Maxwell B, Shields M, O’Donoghue D. Use of oximetry to screen for paediatric obstructive sleep apnoea: Is one night enough and is 6 hours too much? Arch Dis Child 2021;106:58-61.
Agnew HW Jr, Webb WB, Williams RL. The first night effect: An EEG study of sleep. Psychophysiology 1966;2:263-6.
Scholle S, Scholle HC, Kemper A, Glaser S, Rieger B, Kemper G, et al
. First night effect in children and adolescents undergoing polysomnography for sleep-disordered breathing. Clin Neurophysiol 2003;114:2138-45.
Fletcher J, Page M, Jeffery HE. Sleep states and neonatal pulse oximetry. Sleep 1998;21:305-10.
Wellington G, Elder D, Campbell A. 24-hour oxygen saturation recordings in preterm infants: Editing artefact. Acta Paediatr 2018;107:1362-9.
[Figure 1], [Figure 2]
[Table 1], [Table 2], [Table 3], [Table 4], [Table 5]