Article Text
PDF
Abstract
Objective To define reference ranges for the 3% oxygen desaturation index (DI3) in healthy children under 12 years old during sleep.
Design Observational.
Setting Home.
Subjects Healthy children aged 6 months to 12 years of age.
Intervention Nocturnal pulse oximetry at home. Parents documented sleep times. Visi-Download software (Stowood Scientific) analysed data with artefact and wake periods removed.
Main outcome measures The following oximetry parameters used in the assessment of sleep-disordered breathing conditions were measured: 3% (DI3) and 4% (DI4) oxygen desaturation indices—the number of times per hour where the oxygen saturation falls by at least 3% or 4% from baseline, mean saturations (SAT50), minimum saturations (SATmin), delta index 12 s (DI12s), and percentage time with saturations below 92% and 90%.
Results Seventy-nine children underwent nocturnal home pulse oximetry, from which there were 66 studies suitable for analysis. The median values for DI3 and DI4 were 2.58 (95% CI 1.96 to 3.10) and 0.92 (95% CI 0.73 to 1.15), respectively. The 95th and 97.5th centiles for DI3 were 6.43 and 7.06, respectively, which inform our cut-off value for normality. The mean values for SAT50 and SATmin were 97.57% (95% CI 97.38% to 97.76%) and 91.09% (95% CI 90.32% to 91.86%), respectively.
Conclusion In children aged 6 months to 12 years, we define normality of the 3% oxygen desaturation index as <7 using standalone, motion-resistant pulse oximeters with short averaging times.
- sleep
- physiology
Data availability statement
All data relevant to the study are included in the article or uploaded as online supplemental information. Raw data are available from the corresponding author contactable by email.
Statistics from Altmetric.com
What is already known on this topic?
A 4% oxygen desaturation index greater than 4 is widely used as a criterion of abnormality.
Changes in the definition of sleep apnoea (American Academy Sleep Medicine) has resulted in a shift to report 3% rather than 4% oxygen desaturations.
Data are lacking on reference ranges for 3% oxygen desaturation indices using modern oximeters able to exclude motion artefact and which use short averaging times.
What this study adds?
This study defines 3% oxygen desaturation index reference ranges for children aged 6 months to 12 years using motion-resistant oximeters with short averaging times.
This study supports existing data on commonly reported oximetry variables used in clinical practice for children.
Younger children demonstrate higher 3% oxygen desaturation indices, but further study is warranted to establish normative ranges in children under 2 years of age.
Introduction
Pulse oximetry is a non-invasive method of estimating blood oxygen saturations and is frequently used in children to investigate sleep-disordered breathing (SDB). SDB conditions may be broadly split into two groups—obstructive sleep apnoea (OSA) and central sleep apnoea (CSA). Urschitz et al previously published normative oximetry data in primary school age children.1 They published mean oxygen saturations (SAT50) and the 4% desaturation index (DI4)—the number of times per hour where the oxygen saturation falls by at least 4% from baseline. These oximetry parameters are frequently used to detect abnormality. Based on their data, a DI4 of >4 is regarded as abnormal and has been widely used as a criterion of abnormality.
In 2012, the American Academy of Sleep Medicine updated the scoring criteria for SDB such that apnoeas and hypopnoeas were defined according to oxygen desaturation of 3% (DI3) from baseline rather than 4%.2 Alongside this, the 2019 Australasian Sleep Association guidelines on pulse oximetry recommended reporting 3% rather than 4% desaturations when defining cut-offs for normality.3
We previously reported normative oximetry parameters including the DI3 for healthy infants under 4 months of age using modern oximeters. These are able to extract motion artefact and have short averaging times and can therefore accurately detect brief periods of desaturation often seen in children with SDB.4 Currently, no data exist on reference ranges for DI3 for older children using these oximeters. This study aimed to provide reference ranges for DI3 in healthy children between the ages of 6 months and 12 years using Masimo oximetry set at a short averaging time.
Methods
Healthy children of health professionals at University Hospital Southampton NHS Foundation Trust below the age of 12 years were recruited for study. Basic demographics, medical history data for the child and written informed consent to participate in the study were obtained from the parents. Parents were also asked to complete a sleep history questionnaire for their child. Participants were excluded from the study if they had known respiratory or cardiac conditions or conditions predisposing to SDB. Recordings were undertaken when the child was clinically well and without intercurrent illness.
Data were collected overnight at home using a Masimo Rad-8 (USA) pulse oximeter with a 2 s averaging time. Parents were asked to record times of sleep onset, morning wakening, periods of wake throughout the night and how the sleep quality compared with a usual night’s sleep. Data were collected for the following parameters: mean oxygen saturations (SAT50), minimum oxygen saturations (SATmin), number of oxygen desaturation events >3% (DI3) and >4% (DI4) from baseline per hour, delta index (DI12s), and percentage of time with saturations below 92% and 90%. McGill Oximetry Scores were calculated for participants over the age of 12 months as a screening tool for OSA.5
Data analysis was performed using Visi-Download software (Stowood Scientific, UK). The software automatically extracted artefacts caused by low signal quality and movement. Periods of awakening were removed according to data provided by the caregiver and review of the study by a physiologist skilled at interpreting oximetry traces. Previous inter-rater analysis in our laboratory has demonstrated excellent inter-rater reliability for determining artefact-free recording time (AFRT).4 Analysis was performed on studies with at least 5 hours of AFRT,1 which must have included at least two periods of presumed active or rapid eye movement sleep as determined by increased heart rate variability. Each sleep report was reviewed by a paediatric respiratory consultant to ensure there were no abnormalities suggestive of clinical disease that warranted follow-up.
Data analyses were performed using SPSS (V.24) and CI analysis software packages. Normally distributed variables were reported as means and 95% CIs and non-normally distributed variables as medians and 95% CIs.
Results
Seventy-nine children who were recruited to the study underwent nocturnal home pulse oximetry (see figure 1). Two participants were excluded as they did not return the pulse oximeter before the end of the data collection period. During the study collection period, an increase in the evidence base demonstrated that young infants have rapid changes in their desaturation indices with maturation in the first 4 months of life.4 Since only two infants were recruited under 4 months of age, a decision was made to exclude them from further analysis. Overall, nine participants were excluded for failing to collect at least 5 hours of AFRT, despite attempts to repeat studies (figure 1).
Flowchart showing recruitment. AFRT, artefact-free recording time.
The median age at recording of the 66 successful participants was 94 months (95% CI 83 to 107 months). The age distribution of participants is displayed in figure 2.
Histogram showing the age distribution of the study participants.
There were 31 male (47%) and 35 female (53%) participants. Body mass index was recorded in 60 participants with a mean z score of −0.03 (95% CI −0.29 to 0.24). The mean duration of AFRT was 532 min (95% CI 511 to 553 min). Answers from the sleep history questionnaire revealed 17% of participants were known to snore, which is within the reported normal prevalence in children, and only 3% of participants reported symptoms of SDB. All children over the age of 12 months had a McGill score of 1 which is considered negative or inconclusive for OSA.3 5
For the entire study group, the mean values for SAT50 and SATmin were 97.57% (95% CI 97.38% to 97.76%) and 91.09% (95% CI 90.32% to 91.86%), respectively (table 1). The median values for DI3 and DI4 were 2.58 (95% CI 1.96 to 3.10) and 0.92 (95% CI 0.73 to 1.15), respectively. The 95th and 97.5th centiles for DI3 were 6.43 and 7.06, respectively, which inform our cut-off value for normality (table 2). The mean DI12s was 0.35 (95% CI 0.33 to 0.38).
Descriptive statistics for oxygen saturation baseline parameters for the entire study group (n=66): minimum oxygen saturations (SATmin) and mean oxygen saturations (SAT50)
Descriptive statistics for oximetry desaturation events and indices for the entire study group (n=66): number of oxygen desaturations >4% and >3% from baseline per hour (DI4 and DI3), delta index 12 s (DI12s) and percentage time with saturations below 92% and 90%
A subgroup analysis of participants aged between 6 months and 24 months (n=4) demonstrated higher DI3 (4.1, range 2.93 to 6.61) compared with the older age subgroup of 24 months to 12 years (2.51, 95% CI 1.93 to 2.93). Using the Mann-Whitney U test, the difference in DI3 between the age subgroups was found to be statistically significant (p=0.04). Comparison of DI4 between the subgroups was also found to be significantly higher in the younger age subgroup (p=0.04).
Discussion
Motion-resistant oximeters with short averaging times are increasingly recommended in the investigation of children with SDB since they are highly sensitive for detecting brief changes in oxygen saturation frequently seen in these patients.6 7 In 2012, the American Academy of Sleep Medicine guidelines were amended such that scoring criteria for OSA and CSA were based on a desaturation of 3% from baseline rather than 4%.2 Thus, the number of 3% desaturation events per hour may now be considered more relevant than 4%. While data exist for many parameters of oxygen saturation, little is known about normal limits for 3% desaturations in children using standalone oximeters with short averaging times able to exclude motion artefact. This is the first study reporting reference ranges for 3% oxygen desaturation indices for children older than 6 months using this technology.
Saito et al 8 investigated the role of standalone pulse oximetry in determining indications for adenotonsillectomy in children with SDB. The study included 25 healthy children as controls against 225 participants of interest. In the control group, they reported a mean SATmin value of 94.56% (SD ±2.38) while in our population it was 91.09% (SD ±3.20). Their reported mean values for DI4 and DI3 were 0.21 (SD ±0.29) and 0.74 (SD ±0.65), respectively, which are both significantly lower than our data. These discrepancies from our data are most likely explained by the difference in pulse oximeter averaging times used. Saito et al used oximeters with a longer averaging time of 5 s where pickup of desaturations is often blunted by the increased interval between averaging points.8 9 The impact of different technologies has the potential to affect management plans. For example, in this study, the authors concluded that children with a DI3 of more than 3.5 should undergo adenotonsillectomy given the sensitivity for a successful surgery was 94% at this cut-off value. However, this proposed DI3 cut-off value for surgery in SDB would fall well within our reported normative range (mean 2.90 SD ±1.65).
Other studies have reported DI3 incorporated into polysomnography (PSG). The largest of these published by Scholle et al 10 recruited 209 healthy children for the generation of normative cardiorespiratory parameters using PSG. The study merged results from the Nonin oximeter incorporated into the Embla system and Masimo data obtained from the Alice system. However, unpublished data from our laboratory have previously shown poor limits of agreement between integrated Nonin and standalone Masimo oximeters. Details on averaging times for the oximeters were not reported in this paper. The participants were stratified into eight groups each containing fewer than 35 children. The study is therefore limited by small sample sizes in each group and an absence of reference ranges for the group as a whole. Similarly, Wang et al 11 again merged data from two separate oximeters and do not report averaging times. Finally, a study by Verhulst et al 12 used a Palco pulse oximeter that has not been validated for continuous monitoring in children. These studies highlight how differences in oximeter technology and study protocol are a key limitation in comparing normative values for desaturation indices derived from oximetry or polysomnographic studies within the literature currently.13
The decision to recruit a community-based study population was based on the fact that most clinical pulse oximetry traces are undertaken at home. Therefore, the aim was to produce normative data which would be relevant to standard clinical practice. The limitations of this technique are that interpretation of the sleep–wake state is constrained and dependent on parental reporting. Within our laboratory, the physiologists are skilled at interpreting these studies and previous inter-rater analysis has demonstrated excellent inter-rater reliability.4
Similar to the Wang et al 11 and Verhulst et al 12 papers, this study recruited children of hospital employees. This may have resulted in recruitment of families who were highly invested in research outcomes. However, the study recruited families from a wide range of healthcare backgrounds creating a diverse group of participants. In addition, our reported standard oximetry parameters are comparable with those generated from polysomnographic studies.10
Young infants demonstrate a higher frequency of desaturation events.4 A subgroup analysis of participants aged between 6 months and 24 months demonstrated higher DI3 compared with the older age subgroup of 24 months to 12 years, an observation explained by the underlying immaturity of respiratory control mechanisms during sleep in normal healthy infants.14 In our study, the difference in DI3 between the age subgroups was found to be statistically significant (p=0.04), despite the fact that we were only able to recruit four infants into the younger age group (figure 2). Further study is warranted to establish normative desaturation indices stratified by age using oximeters with motion-resistant technology and short averaging times throughout infancy.
In conclusion, this study reports for the first time 3% oxygen desaturation indices in children aged 6 months to 12 years using a standalone pulse oximeter with short averaging times able to exclude motion artefact. Based on our data, we define normality as DI3 <7. Our study protocol is in line with the latest technical recommendations for the investigation of SDB in children. We believe that these results can form the basis of reference ranges on which a judgement of normality or otherwise can be determined.
Data availability statement
All data relevant to the study are included in the article or uploaded as online supplemental information. Raw data are available from the corresponding author contactable by email.
Ethics statements
Ethics approval
Southampton REC (16/LO/2254).
Footnotes
Contributors All authors contributed to the planning and conduct of the work as per the ICMJE recommendations. JWYO, JCG and HJE were responsible for writing and review of the manuscript.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.