Sleep Technologists Research Articles

Validation of automated detection of REM sleep without atonia using in-laboratory and in-home recordings.

To evaluate the concordance between visual scoring and automated detection of REM sleep without atonia (RSWA) and the validity and reliability of in-home automated-RSWA detection in REM sleep behavior disorder (RBD) patients and a control group (CG). Sleep Profiler signals were acquired during simultaneous in-laboratory polysomnography in 24 isolated RBD patients. Chin and arm RSWA measures visually scored by an expert sleep technologist were compared to algorithms designed to automate RSWA detection. In a second cohort, the accuracy of automated-RSWA detection for discriminating between RBD and CG (n = 21 and 42, respectively) was assessed in multi-night in-home recordings. For the in-laboratory studies, agreement between visual and auto-scored RSWA from the chin and arm were excellent, with intra-class correlations of 0.89 and 0.95, respectively, and substantial, based on Kappa scores of 0.68 and 0.74, respectively. For classification of iRBD patients versus controls, specificities derived from auto-detected RSWA densities obtained from in-home recordings were 0.88 for the chin, 0.93 for the arm, and 0.90 for the chin or arm, while the sensitivities were 0.81, 0.81 and 0.86, respectively. The night-to-night consistencies of the respective auto-detected RSWA densities were good based on intra-class correlations of 0.81, 0.79 and 0.84, however some night-to-night disagreements in abnormal RSWA detection were observed. When compared to expert visual RSWA scoring, automated RSWA detection demonstrates promise for detection of RBD. The night-to-night reliability of chin- and arm-RSWA densities acquired in-home were equivalent.

Time–frequency ridge characterisation of sleep stage transitions: Towards improving electroencephalogram annotations using an advanced visualisation technique

Manual sleep stage scoring of polysomnography recordings is an expensive and time-consuming process, further complicated by inconsistent sleep stage agreement among sleep experts (clinicians and sleep technologists). Hence, development of automated sleep scoring algorithms are an emerging topic of interest. Automation typically mimics the clinical decision path by implementing a series of predefined rules, such as the American Academy of Sleep Medicine’s (AASM) scoring manual. Recently, data driven methods have emerged using machine or deep learning. Both manual and automated methods of scoring have known limitations; primarily, unacceptable variation in agreement between different scorers and algorithms. Within the literature, electroencephalogram (EEG) frequency is an important feature considered by both sleep experts and automated approaches for classifying sleep stages.This study presents a novel approach to sleep stage analysis, by developing a methodology to precisely determine the temporal location of sleep stage transitions. The current gold standard fails to identify such transitional changes, which leads to poor inter-scorer reliability. Therefore, development and implementation of such methodologies is a crucial, but overlooked, step in improving the consistency of scoring within sleep studies. In this work, EEG time–frequency ridge analysis was used to characterise the dominant frequency component of EEG signals in time, at the point of sleep stage transition. An in-depth analysis of N3 → N2 and N2 → N3 transitions in the 2018 PhysioNet challenge “You Snooze, You Win” and the Wisconsin Sleep Cohort (WSC) datasets (n = 994, n = 742; approximately 13,888 h of sleep data) showed consistent time–frequency patterns at the point of transition, from one sleep stage to another. This methodology allows simple and ‘interpretable’ features to be generated in future work, to precisely identify the temporal location of sleep stage transitions with the aim of improving inter-scorer reliability.

Automated explainable wavelet-based sleep scoring system for a population suspected with insomnia, apnea and periodic leg movement

Sleep is an integral and vital component of human life, contributing significantly to overall health and well-being, but a considerable number of people worldwide experience sleep disorders. Sleep disorder diagnosis heavily depends on accurately classifying sleep stages. Traditionally, this classification has been performed manually by trained sleep technologists that visually inspect polysomnography records. However, in order to mitigate the labor-intensive nature of this process, automated approaches have been developed. These automated methods aim to streamline and facilitate sleep stage classification. This study aims to classify sleep stages in a dataset comprising subjects with insomnia, PLM, and sleep apnea. The dataset consists of PSG recordings from the multi-ethnic study of atherosclerosis (MESA) cohort of the national sleep research resource (NSRR), including 2056 subjects. Among these subjects, 130 have insomnia, 39 suffer from PLM, 156 have sleep apnea, and the remaining 1731 are classified as good sleepers. This study proposes an automated computerized technique to classify sleep stages, developing a machine-learning model with explainable artificial intelligence (XAI) capabilities using wavelet-based Hjorth parameters. An optimal biorthogonal wavelet filter bank (BOWFB) has been employed to extract subbands (SBs) from 30 seconds of electroencephalogram (EEG) epochs. Three EEG channels, namely: Fz_Cz, Cz_Oz, and C4_M1, are employed to yield an optimum outcome. The Hjorth parameters extracted from SBs were then fed to different machine learning algorithms. To gain an understanding of the model, in this study, we used SHAP (Shapley Additive explanations) method. For subjects suffering from the aforementioned diseases, the model utilized features derived from all channels and employed an ensembled bagged trees (EnBT) classifier. The highest accuracy of 86.8%, 87.3%, 85.0%, 84.5%, and 83.8% is obtained for the insomniac, PLM, apniac, good sleepers and complete datasets, respectively. Using these techniques and datasets, the study aims to enhance sleep stage classification accuracy and improve understanding of sleep disorders such as insomnia, PLM, and sleep apnea.

Validation of Tracheal Sound-Based Respiratory Effort Monitoring for Obstructive Sleep Apnoea Diagnosis.

Background: Respiratory effort is considered important in the context of the diagnosis of obstructive sleep apnoea (OSA), as well as other sleep disorders. However, current monitoring techniques can be obtrusive and interfere with a patient's natural sleep. This study examines the reliability of an unobtrusive tracheal sound-based approach to monitor respiratory effort in the context of OSA, using manually marked respiratory inductance plethysmography (RIP) signals as a gold standard for validation. Methods: In total, 150 patients were trained on the use of type III cardiorespiratory polygraphy, which they took to use at home, alongside a neck-worn AcuPebble system. The respiratory effort channels obtained from the tracheal sound recordings were compared to the effort measured by the RIP bands during automatic and manual marking experiments. A total of 133 central apnoeas, 218 obstructive apnoeas, 263 obstructive hypopneas, and 270 normal breathing randomly selected segments were shuffled and blindly marked by a Registered Polysomnographic Technologist (RPSGT) in both types of channels. The RIP signals had previously also been independently marked by another expert clinician in the context of diagnosing those patients, and without access to the effort channel of AcuPebble. The classification achieved with the acoustically obtained effort was assessed with statistical metrics and the average amplitude distributions per respiratory event type for each of the different channels were also studied to assess the overlap between event types. Results: The performance of the acoustic effort channel was evaluated for the events where both scorers were in agreement in the marking of the gold standard reference channel, showing an average sensitivity of 90.5%, a specificity of 98.6%, and an accuracy of 96.8% against the reference standard with blind expert marking. In addition, a comparison using the Embla Remlogic 4.0 automatic software of the reference standard for classification, as opposed to the expert marking, showed that the acoustic channels outperformed the RIP channels (acoustic sensitivity: 71.9%; acoustic specificity: 97.2%; RIP sensitivity: 70.1%; RIP specificity: 76.1%). The amplitude trends across different event types also showed that the acoustic channels exhibited a better differentiation between the amplitude distributions of different event types, which can help when doing manual interpretation. Conclusions: The results prove that the acoustically obtained effort channel extracted using AcuPebble is an accurate, reliable, and more patient-friendly alternative to RIP in the context of OSA.

An optimized framework for processing multicentric polysomnographic data incorporating expert human oversight.

Polysomnographic recordings are essential for diagnosing many sleep disorders, yet their detailed analysis presents considerable challenges. With the rise of machine learning methodologies, researchers have created various algorithms to automatically score and extract clinically relevant features from polysomnography, but less research has been devoted to how exactly the algorithms should be incorporated into the workflow of sleep technologists. This paper presents a sophisticated data collection platform developed under the Sleep Revolution project, to harness polysomnographic data from multiple European centers. A tripartite platform is presented: a user-friendly web platform for uploading three-night polysomnographic recordings, a dedicated splitter that segments these into individual one-night recordings, and an advanced processor that enhances the one-night polysomnography with contemporary automatic scoring algorithms. The platform is evaluated using real-life data and human scorers, whereby scoring time, accuracy, and trust are quantified. Additionally, the scorers were interviewed about their trust in the platform, along with the impact of its integration into their workflow. We found that incorporating AI into the workflow of sleep technologists both decreased the time to score by up to 65 min and increased the agreement between technologists by as much as 0.17 κ. We conclude that while the inclusion of AI into the workflow of sleep technologists can have a positive impact in terms of speed and agreement, there is a need for trust in the algorithms.

0563 Exploring Trends of Oronasal versus Nasal Mask Use by Sleep Technologists

0563 Exploring Trends of Oronasal versus Nasal Mask Use by Sleep Technologists

Validation of a sleep staging classification model for healthy adults based on two combinations of a single-channel EEG headband and wrist actigraphy.

The aim of this study was to develop a sleep staging classification model capable of accurately performing on different wearable devices. Twenty-three healthy participants underwent a full-night type I polysomnography and used two device combinations: (A) flexible single-channel electroencephalogram (EEG) headband + actigraphy (n = 12) and (B) rigid single-channel EEG headband + actigraphy (n = 11). The signals were segmented into 30-second epochs according to polysomnographic stages (scored by a board-certified sleep technologist; model ground truth) and 18 frequency and time features were extracted. The model consisted of an ensemble of bagged decision trees. Bagging refers to bootstrap aggregation to reduce overfitting and improve generalization. To evaluate the model, a training dataset under 5-fold cross-validation and an 80-20% dataset split was used. The headbands were also evaluated without the actigraphy feature. Participants also completed a usability evaluation (comfort, pain while sleeping, and sleep disturbance). Combination A had an F1-score of 98.4% and the flexible headband alone of 97.7% (error rate for N1: combination A = 9.8%; flexible headband alone = 15.7%). Combination B had an F1-score of 96.9% and the rigid headband alone of 95.3% (error rate for N1: combination B = 17.0%; rigid headband alone = 27.7%); in both, N1 was more confounded with N2. We developed an accurate sleep classification model based on a single-channel EEG device, and actigraphy was not an important feature of the model. Both headbands were found to be useful, with the rigid one being more disruptive to sleep. Future research can improve our results by applying the developed model in a population with sleep disorders. Registry: ClinicalTrials.gov; Name: Actigraphy, Wearable EEG Band and Smartphone for Sleep Staging; URL: https://clinicaltrials.gov/study/NCT04943562; Identifier: NCT04943562. Melo MC, Vallim JRS, Garbuio S, etal. Validation of a sleep staging classification model for healthy adults based on 2 combinations of a single-channel EEG headband and wrist actigraphy. J Clin Sleep Med. 2024;20(6):983-990.

Partial endorsement of: “Video-polysomnography procedures for diagnosis of rapid eye movement sleep behavior disorder (RBD) and the identification of its prodromal stages: Guidelines from the International RBD Study Group” by the World Sleep Society

Updated guidelines for the video-polysomnography (vPSG) procedures for diagnosing rapid eye movement sleep behavior disorder (RBD) and the identification of its prodromal stages have recently been proposed by the Neurophysiology Working Group of the International RBD Study Group (IRBDSG). These guidelines were selected for review by a World Sleep Society (WSS) Parasomnias Task Force and the WSS International Sleep Medicine Guidelines Committee. A survey was completed by sleep society leaders and prominent sleep clinicians and researchers in 31 WSS member countries across six continents, focused on sleep technologist training and certification; extent of public/private health insurance coverage for the vPSG evaluation of RBD; extent of hospital-based sleep-technologist-attended overnight vPSG studies; availability of video during PSG studies; and sufficient specification of PSG machines to record and analyze REM sleep without atonia. The findings from this survey indicated that most health systems and medical communities across WSS member countries would not be capable of implementing the proposed more stringent guidelines, which would then strongly interfere with the diagnosis of RBD in a large portion of patients who would not be able to receive the required (often repeated) vPSG evaluation. Therefore, the WSS can only partially endorse the updated guidelines and concludes that the current International Classification of Sleep Disorders-3rd edition diagnostic criteria for RBD should still be retained as the standard reference for the diagnosis of RBD, and that further discussion across all members of the IRBDSG should take place to ensure the feasibility of any future proposed changes.

Association of Perceived Stress Scale with the Severity of Obstructive Sleep Apnea in Pakistani Population

Background: Obstructive Sleep Apnea (OSA) is a prevalent sleep disorder characterized by recurrent episodes of partial or complete upper airway obstruction during sleep, leading to disruptions in breathing patterns and, consequently, reduced oxygen level. OSA has traditionally been viewed as a sleep-related physiological disorder, there is growing recognition that psychological factors, such as stress, may play a role in its onset and progression. Objective: To evaluate the association of Perceived Stress Scale with the severity of Obstructive Sleep Apnea in Pakistani Population. Study Design: Prospective cross-sectional study. Settings: untreated OSA individuals at Pulmonary Vascular and Sleep Disorders division of Dow University Hospital in Karachi, Pakistan. Duration: Study was done during a period of six months from May 2022 to October 2022. Methods: Patients OSA diagnosis through overnight Polysomnography with AHI of more than 5, and OSA participants above 18 years of age were included. A stress questionnaire was filled out when the individual came for overnight Polysomnography. The Sleep technologist measures the anthropometric measurements like height, and weight through height scales and weight machines, the neck circumference through the inch tape, BMI was computed through a formula which is weight in kg/height in meter2, and stress was evaluated through the Perceived Stress Scale (PSS). Data analysis was performed by using IBM SPSS statistics 26. Results: A total of 167 OSA patients participated in the study. There were 100 (60%) male and 67 (40%) female individuals. The mean age of females who perceived stress was 53.7 ± 10.5years which is higher from males 47.9 ± 13.1 years. Low stress was found in moderate and severe OSA as compare to mild OSA, similarly 49% moderate stress in Moderate OSA and 52% moderate stress in Severe OSA patients respectively. The percentage of high stress was significantly prominent in Mild OSA in comparison to Moderate and Severe OSA. The result of the study shows that the perceived stress scale is significantly correlated with Obstructive sleep apnea (p<0.001). Conclusion: The outcomes of this study reveal a direct significant correlation between OSA and the Perceived Stress Scale. The findings underscore the importance of ongoing investigation to delve into potential underlying mechanisms.

0783 Sleep Continuity and Quality in Adolescents with Functional Abdominal Pain

Abstract Introduction Functional Abdominal Pain Disorders (FAPD), in which chronic abdominal pain occurs independently of organic pathology, are common in adolescents. FAPD can lead to significant impairment in daytime function and may overlap with sleep disorders. Though several studies have explored the nature of sleep disturbances experienced by adolescents with FAPD, none have characterized differences in sleep architecture in this population. Methods Adolescents (age 11-18 years) with FAPD and age/gender-matched controls were prospectively recruited to undergo polysomnography (PSG). PSG was performed according to AASM guidelines, scored by a registered sleep technologist, and interpreted by a board-certified sleep physician. The first four hours of artifact-free NREM sleep were analyzed using fast Fourier transformation with Welch tapered windows of 1024 samples (and 50% overlap) averaged over 30-second epochs. Total power was calculated for very low (0.3 – 1 Hz), delta (1 – 4 Hz), theta (4 – 8 Hz), and alpha (8-12 Hz) frequencies. Comparisons between groups were made using Wilcoxon rank sum tests or t-tests for non-normal and normal data, respectively. Results A total of 41 subjects were recruited (FAPD n=21, control n=20). The FAPD group had a greater arousal index and a lower percentage of NREM 3 sleep when compared to the control group (FAPD arousals: mean=10.08/min ± 4.89 min, control arousals mean=7.31/min ± 3.61/min, p = 0.046), FAPD NREM 3: mean=20.13% ± 5.47%, control NREM 3: mean=24.77% ± 6.94%, p =0.023). Delta power was significantly lower in the FAPD group compared to controls (FAPD Delta mean=488.67 µV2 ± 282.82 µV2, control Delta mean=1193.14 µV2 ± 663.89 µV2, p < 0.001). Conclusion This study demonstrates significant differences in sleep architecture of adolescents with FAPD compared to their peers. In particular, decreased proportions of NREM 3 sleep and lower Delta power are indicative of diminished slow wave sleep (SWS). Reductions in SWS may be indicative of common neurotransmitter pathways shared between sleep and chronic pain syndromes, or secondary to sleep disruption associated with pain. Further research is indicated. Support (if any) This research is supported by the Cincinnati Children’s Research Foundation and the Summer Medical Student Respiratory Research Fellowship program at Cincinnati Children’s.

Diagnosis of Nocturnal Hypoventilation in Pediatric Neuromuscular Disorders: A Survey of Clinical Practice in Australia and New Zealand

Objectives: Nocturnal hypoventilation is a complication of neuromuscular disorders. There are various recommendations for measuring pCO<sub>2</sub> during polysomnography and numerous national and international definitions of hypoventilation that could contribute to significant variations in clinical practice. We therefore aimed to determine clinical practices implemented by Australasian pediatric sleep physicians.Methods: Pediatric sleep physicians completed an electronic survey for information regarding pCO<sub>2</sub> measurements and definitions of hypoventilation that are followed for children with neuromuscular disorders.Results: It was found that transcutaneous measurement of pCO<sub>2</sub> was performed in all centers, with 25% of the centers simultaneously performing capnography. Twelve definitions of hypoventilation were used, including published definitions from the American Academy of Sleep Medicine (AASM) manual and recommendations of the pediatric Australasian Sleep Association/Australasian Sleep Technologists Association. The most commonly used definition of hypoventilation (9/17, 53%) was the 2012 pediatric AASM definition (pCO<sub>2</sub> >50 mmHg for >25% of the total sleep time). There was a discrepancy between centers and individuals within the same center when defining hypoventilation. Answers stating the use of the Australasian definitions (rise in pCO<sub>2</sub> ≥10 mmHg from wake to sleep, average rise in pCO<sub>2</sub> ≥3 mmHg from non rapid eye movement to rapid eye movement sleep) were more frequent when asked specifically via a checkbox (yes/no) compared to free text.Conclusions: These results confirm the heterogeneity and lack of standardization of clinical practice within Australasia when measuring pCO<sub>2</sub> during polysomnography and defining hypoventilation. The Australasian definitions were not used as frequently as anticipated.

Benchmarking performance of an automatic polysomnography scoring system in a population with suspected sleep disorders.

The current gold standard for measuring sleep disorders is polysomnography (PSG), which is manually scored by a sleep technologist. Scoring a PSG is time-consuming and tedious, with substantial inter-rater variability. A deep-learning-based sleep analysis software module can perform autoscoring of PSG. The primary objective of the study is to validate the accuracy and reliability of the autoscoring software. The secondary objective is to measure workflow improvements in terms of time and cost via a time motion study. The performance of an automatic PSG scoring software was benchmarked against the performance of two independent sleep technologists on PSG data collected from patients with suspected sleep disorders. The technologists at the hospital clinic and a third-party scoring company scored the PSG records independently. The scores were then compared between the technologists and the automatic scoring system. An observational study was also performed where the time taken for sleep technologists at the hospital clinic to manually score PSGs was tracked, along with the time taken by the automatic scoring software to assess for potential time savings. Pearson's correlation between the manually scored apnea-hypopnea index (AHI) and the automatically scored AHI was 0.962, demonstrating a near-perfect agreement. The autoscoring system demonstrated similar results in sleep staging. The agreement between automatic staging and manual scoring was higher in terms of accuracy and Cohen's kappa than the agreement between experts. The autoscoring system took an average of 42.7 s to score each record compared with 4,243 s for manual scoring. Following a manual review of the auto scores, an average time savings of 38.6 min per PSG was observed, amounting to 0.25 full-time equivalent (FTE) savings per year. The findings indicate a potential for a reduction in the burden of manual scoring of PSGs by sleep technologists and may be of operational significance for sleep laboratories in the healthcare setting.

Deep learning for sleep analysis on children with sleep-disordered breathing: Automatic detection of mouth breathing events

IntroductionSleep-disordered breathing (SDB) can range from habitual snoring to severe obstructive sleep apnea (OSA). A common characteristic of SDB in children is mouth breathing, yet it is commonly overlooked and inconsistently diagnosed. The primary aim of this study is to construct a deep learning algorithm in order to automatically detect mouth breathing events in children from polysomnography (PSG) recordings.MethodsThe PSG of 20 subjects aged 10–13 years were used, 15 of which had reported snoring or presented high snoring and/or high OSA values by scoring conducted by a sleep technologist, including mouth breathing events. The separately measured mouth and nasal pressure signals from the PSG were fed through convolutional neural networks to identify mouth breathing events.ResultsThe finalized model presented 93.5% accuracy, 97.8% precision, 89% true positive rate, and 2% false positive rate when applied to the validation data that was set aside from the training data. The model's performance decreased when applied to a second validation data set, indicating a need for a larger training set.ConclusionThe results show the potential of deep neural networks in the analysis and classification of biological signals, and illustrates the usefulness of machine learning in sleep analysis.

Keep it simple: A novel technique for measuring airflow using a wireless patch

Despite the growing use of home sleep testing for obstructive sleep apnea (OSA), there are significant barriers that make it difficult for patients. This study aimed to evaluate a new set of wireless patches for the detection of OSA as compared to polysomnography (PSG) respiratory signals. Safety was also evaluated. Patients (n=25; mean age=51.3; SD=15.6) undergoing standard PSG evaluation for OSA wore 2 device patches applied on the chest and abdomen to provide surrogates of respiratory effort and derived airflow. Two sets of respiratory data (with common SpO2 and heart rate as a reference from PSG) were created, one from PSG and one from the device patches. Data were scored by a Registered Polysomnographic Technologist. Linear regression analyses and Bland-Altman plots were used to compare the two derived apnea-hypopnea indices (AHI). Bench testing was performed to determine whether the device patch respiratory signal accuracy remained consistent for different body characteristics. There was a significant correlation for AHI between the device patch and PSG (r=0.979; P < .001; 95% CI 0.952, 1.00). There were also significant correlations between PSG and the device patch for supine AHI (r=0.866, P < .001, 95% CI 0.610, 0.965) and nonsupine AHI (r=0.983; <0.001, 95% CI 0.954, 1.00). No differences in respiratory signal accuracy were found with bench testing comparison groups. These preliminary results suggest that the device patch may be an effective way to measure respiratory dynamics, including derived airflow, to evaluate OSA. Results show promise as a new innovative method for home sleep testing.

P010 Success and failure rates of ambulatory home sleep studies (Level 2) among Indigenous patients

Abstract Introduction/Aim Level 2 sleep studies are being increasingly performed worldwide in order to minimise the cost and access to level1 in-lab polysomnography. However, their accuracy and validity remain disputed. For populations residing in regional and remote localities, access to level1 sleep studies are further compromised, such as among Indigenous people of Australia, where 80% reside in remote communities. There is limited information in the literature regarding the efficacy of level2 sleep studies among Indigenous patients. Hence, in this study we evaluated the efficacy of level2 sleep studies amongst an Indigenous population. Methods 552 level2 sleep studies from Indigenous patients recorded between 2011-2020 were included. Sleep studies were analysed and scored manually by registered polysomnographic technologist as per standard guidelines. The studies were also assessed to determine the soundness of the study. Demographic and clinical information of patients was also collected. Results 50% of the patients were female, and 51% resided in remote or very remote locations. Of the 552 studies, 47(8.5%) were failed. There was no significant differences in demographic factors between failed and successful studies. The most common reason for failure was intolerance of the sensors which resulting in removing the device. Conclusion Level 2 sleep studies showed high success rate amongst this Indigenous population with just 8.5% failure rate. As there is limited access to level 1 sleep studies among the Indigenous Australian population due to remoteness, this study shows that level 2 sleep studies are a feasible alternative in the diagnosis for the presence of sleep disorders.

P137 Diagnosis of nocturnal hypoventilation in children with neuromuscular disorders by polysomnography – a survey of clinical practice in Australia and New Zealand

Abstract Background Diagnosis of nocturnal hypoventilation requires measurement of pCO2 during sleep, which can occur via exhaled gas (capnography) or transcutaneous methods. Different national and international definitions of hypoventilation exist, which may contribute to significant variation in clinical practice. Methods An electronic survey was completed by pediatric sleep physicians in Australia and New Zealand to determine how pCO2 was measured during polysomnography, what pCO2 data was available to reporting physicians/included in the report and which definitions of hypoventilation physicians use. Results More than two thirds (70%) of physicians indicated that in their centre pCO2 was collected via transcutaneous measurement only and 30% by both transcutaneous and end-tidal measurement. Twelve definitions of hypoventilation were used, including published definitions from the American Academy of Sleep Medicine manual and the pediatric Australasian Sleep Association/Australasian Sleep Technologists Association recommendations. The most commonly used definition was an increase in pCO2 &amp;gt;50mmHg for &amp;gt;25% total sleep time. The percentage of total sleep time with pCO2 &amp;gt;50mmHg was routinely included in the report and/or easily available in the raw data for 70% of physicians. Discussion There was significant variation and lack of standardisation within Australia and New Zealand when measuring and reporting pCO2 during polysomnography as well as definitions of hypoventilation used by pediatric sleep physicians. The Australasian Sleep Association/Australasian Sleep Technologists Association recommendations were not frequently used, possibly due to relevant information not being available to reporting physicians, recommendations being irrelevant and/or outdated and excessive cognitive load associated with remembering multiple definitions.

P113 A detailed analysis of multicentric sleep staging inter-rater variabilities

Abstract Introduction Correct identification of sleep stages is important in sleep disorder diagnosis; but is subject to high inter-scorer variability. Thus, we aimed to systematically analyse sleep staging agreements between ten scorers in a multicentric environment to better understand areas that possess high scoring variations. Method Polysomnographic recordings of 50 individuals (mean (±SD) AHI 12.0±13.2) were independently scored by ten sleep technologists from seven sleep centres ( Europe six, Australia one). The epoch-based majority sleep stage was identified from all scorings; then the mean scorer agreement with this majority score was calculated. We also investigated how sleep-state transitions affect the mean scorer agreement. Results In total, 48556 epochs were analysed. The mean (±SD) scorer agreement with majority score was 86.3±2.9% (Randolph’s kappa κ=0.74) overall; and 86.3±9.5% (κ=0.74), 65.6±13.5% (κ=0.39), 86.1±6.2% (κ=0.74), 88.8±11.2% (κ=0.78), and 91.7±4.5% (κ=0.83) for Wake, N1, N2, N3, and REM respectively. Across the ten scorers, 20.0% of epochs were unanimously scored as N2, and the most frequent disagreement was between N2 and N3 (19.7% of epochs). Only 0.3% of epochs were unanimously scored as N1. There was a negative association between the sleep stage transition frequency of the majority score and the mean scorer agreement (r=-0.61). Discussion High mean scorer agreements were obtained with the majority score except for N1. Surprisingly, considerable disparities were observed among scorers in distinguishing N2 and N3; this may reflect both the discrepancy associated with N2/N3 transitions and potentially different scoring perceptions. Greater scoring discrepancies were identified in individuals with more frequent sleep-state transitions.

Toward Sleep Study Automation: Detection Evaluation of Respiratory-Related Events.

The diagnosis of sleep disordered breathing depends on the detection of respiratory-related events: apneas, hypopneas, snores, or respiratory event-related arousals from sleep studies. While a number of automatic detection methods have been proposed, their reproducibility has been an issue, in part due to the absence of a generally accepted protocol for evaluating their results. With sleep measurements this is usually treated as a classification problem and the accompanying issue of localization is not treated as similarly critical. To address these problems we present a detection evaluation protocol that is able to qualitatively assess the match between two annotations of respiratory-related events. This protocol relies on measuring the relative temporal overlap between two annotations in order to find an alignment that maximizes their F1-score at the sequence level. This protocol can be used in applications which require a precise estimate of the number of events, total event duration, and a joint estimate of event number and duration. We assess its application using a data set that contains over 10,000 manually annotated snore events from 9 subjects, and show that when using the American Academy of Sleep Medicine Manual standard, two sleep technologists can achieve an F1-score of 0.88 when identifying the presence of snore events. In addition, we drafted rules for marking snore boundaries and showed that one sleep technologist can achieve F1-score of 0.94 at the same tasks. Finally, we compared this protocol against the protocol that is used to evaluate sleep spindle detection and highlighted the differences.

0787 Healthcare Utilization in Initiation of Oral Appliance vs Positive Airway Pressure Therapy for Sleep Apnea

IntroductionThe COVID-19 pandemic and related supply chain issues have created shortages in integral components of Positive Airway Pressure (PAP) devices, the gold standard treatment for sleep apnea. Concurrently, patients have delayed care and are returning in increasing numbers. With these overlapping pressures, alternative treatments are needed. Custom-fabricated Oral Appliances (OA) are uniquely poised as a solution. However, it is unknown if initiation and treatment cost and healthcare utilization are similar to PAP or will create further disruptions at scale.MethodsPatients who initiated PAP or OA therapy 2018-2020 were included. Matched visits 2017-2021 were referenced. Patients with multiple treatment initiations were excluded. Healthcare utilization quantified number visits, stratified by provider type (Physician, Physician Assistant (PA), American Board of Dental Sleep Medicine (ABDSM) Accredited Dentist, or Registered Polysomnographic Technologist (RPSGT)). Contractual amounts for CPT codes were averaged to estimate cost.Results5172 patients, 374 received OA (7.2%). Prior to initiation, OA therapy utilized more visits on average than PAP (4.5±1.7(±SD) vs 3.5±1.9, p<0.0001). Following initiation, OA therapy utilized fewer visits than PAP (4.1±3.9 vs 5.5±4.6, p<0.0001). Specialized provider visits, i.e. dentist for OA, were lower compared to RPSGT for PAP therapy, both before and after initiation (1.4±0.8 vs 2.0±1.4 before, 1.9±1.2 vs 2.6±1.8 after, both p<0.0001). Further, prior to initiation, Physician and PA utilization was similar between OA and PAP therapies (1.4±0.8 vs 1.5±1.0 Physician, 1.1±0.8 vs 1.2±0.8 PA, both p>0.057). However, following initiation, OA therapy utilized fewer Physician visits than PAP (1.7±1.1 vs 2.1±1.7, p<0.0001) but similar PA visits (1.9±1.5 vs 2.1±1.4, p>0.5). Together, with OA dental visits estimated to be the least expensive associated visit, this analysis estimates that the provider cost of initiation of OA therapy is lower than that of PAP.ConclusionOverall, OA therapy requires less healthcare utilization, especially of providers with highest reimbursement rates. While OA requires more initial appointments, PAP therapy requires more follow up visits with specialized providers and physicians, thereby increasing cost for patients. Additional cost burden of these visits could impact patient willingness to initiate treatment. This analysis provides supportive evidence for OA as an alternative to PAP with lower treatment cost and healthcare utilization, which may provide an advantage for the already over-burdened healthcare system.Support (If Any)

0548 Behavioral Validation of the University of Michigan REM Behavior Disorder Questionnaire in the Synucleinopathies

Abstract Introduction Questionnaires that inquire about REM sleep behavior disorder are typically validated by reference to polysomnographic findings of REM sleep without atonia. We sought validation for such a questionnaire by examining observed REM-based dream enactment over two nights in a sleep laboratory. Methods Participants with alpha-synucleinopathic disease underwent two consecutive nights of polysomnography attended by a single sleep technologist who was instructed to observe for and document any abnormal behaviors occurring out of sleep. We evaluated University of Michigan REM Behavior Disorder Questionnaire responses completed by either the participant or bed-partner prior to polysomnographic recording and analyzed associations between questionnaire responses and observational data in the sleep laboratory. We subsequently cross-validated the technologist’s observations with hypnogram tracings from nights when dream enactment was observed to determine whether the timing of observed behavior aligned with the timing of REM sleep. Results We studied 92 participants [idiopathic Parkinson’s disease, n = 67; Dementia with Lewy bodies, n = 14; Multiple Systems Atrophy, n = 3; Idiopathic REM sleep behavior disorder, n = 8]. Of 92 participants studied, 30 (32.61%) exhibited dream enactment during REM sleep observed on either night of polysomnographic recording (11 on both nights, 19 on one night). There was a statistically significant difference (p = 0.041 by 2-tailed test) in mean questionnaire scores between participants with observed dream enactment during REM sleep (mean score 0.43; [95% CI mean 0.34-0.52]) and those without dream enactment on either night (mean score 0.32; [95% CI mean 0.26-0.38]). When limiting the sample to participants whose questionnaire responses were derived from the bed-partner (n = 74), the difference approached, but did not reach, statistical significance (p = 0.066). Conclusion These data provide further validation of the University of Michigan REM Behavior Disorder Questionnaire as a useful tool in evaluating REM-based dream enactment behavior in patients with alpha-synucleinopathic disease. Support (If Any)