0450 Role of Oxygen Saturation During Sleep in Identifying Obesity Hypoventilation Syndrome and Its Correlation With Supine Wake End-Tidal Pco2

Abstract

type III home monitors are difficult for patients to set up and wearing the monitors is disruptive to patient's typical sleep pattern. The diagnostic value of an inexpensive easy-to-use light-weight flexible skin-adhesive patch (SomnaPatch) that minimally affects sleep was evaluated in this study. Methods: Simultaneous polysomnography (PSG) and the diagnostic patch recordings were made in 179 subjects (mean age 54.0 ± 13.6 y, 55% male) selected from the databases of patients previously tested with PSG to ensure even representation of the clinically important apnea-hypopnea index (AHI) ranges. The skin-adhesive diagnostic patch weighs less than one ounce and records nasal pressure, blood oxygen saturation, pulse rate, respiratory effort, sleep time and body position (S 3 C 4 O 2 P 2 E 3 R 2 category). To compare the apnea-hypopnea index of the diagnostic patch with polysomnography, all recordings were auto-scored with the Somnolyzer software (Respironics). BlandAltman analysis was performed. Sensitivity, specificity and accuracy were calculated and receiver operating characteristic (ROC) curves were constructed for six AHI thresholds (5, 10, 15, 20, 25 and 30 events per hour). The rate of clinical agreement and positive likelihood ratio were calculated. Results: Overnight recordings from 174 subjects were included in the final analysis. All six ROC curves had area under the curve of over 0.9. Sensitivity, specificity and accuracy for the optimal threshold of AHI≥15 were 0.86, 0.83 and 0.85 respectively. Positive likelihood ratio (LR+) was 7.4. Bland-Altman analysis showed that the bias was 0.9 events per hour and the limits of agreement were 18.1 and -16. Introduction: Hypermobility syndromes can be an important risk factor for Sleep Disordered Breathing (SDB). The risk factors for SDB, such as obesity, male gender, or post-menopausal status, are frequently absent in hypermobile populations, yet the impact of SDB on the quality of these patients' lives can be profound. Methods: Charts of 49 consecutive patients with a Beighton hypermobility score ≥5 from a Neurology Institute were retrospectively reviewed. All ± values reflect the standard error of the mean. Results: Eighty-four percent of these patients had sleep complaints. Thirty underwent Polysomnography (PSG), though PSG was requested for all. Five were denied by insurance, and 6 refused PSG. PSGs demonstrated SDB in 29 patients: 19 had Obstructive Sleep Apnea (G47.33) and 10 had Sleep Apnea -Unspecified (Upper Airway Resistance Syndrome) (G47.30). The average age was 33.4 ± 2.3 years, the average BMI was 26.9 ± 1.6, and the average Epworth Sleepiness Scale Score was 12.1 ± 0.98. Twenty-one of these 30 patients received an MRI of the cervical spine for symptoms and signs of cervical radiculopathy. On a T1-weighted, mid-sagittal section, the following measurements were made of the narrowest dimensions in centimeters: Retropalatal (0.47 ± 0.06), Retrolingual (0.86 ± 0.05), and Retroepiglottal (0.47 ± 0.04), and Epiglottal length (2.22 ± 0.06). These measureme...

Evaluation and Management of Obesity Hypoventilation Syndrome. An Official American Thoracic Society Clinical Practice Guideline

“…One of the four studies included a retrospective and prospective sample and provided diagnostic accuracy data for both samples at the cutoff of 27 mmol/L (40). Therefore, the pooled analysis includes five samples, with a total Nine studies evaluated hypoxemia by pulse oximetry (41,42,44,45,(47)(48)(49)(50)(51) and had prevalence of OHS between 17% and 66%. We identified 14 different measures of hypoxemia in these studies, including: awake oxygen saturation by pulse oximetry (Sp O 2 ) below a defined threshold (92-98%), sleep Sp O 2 below 90% for various percentages of sleep time (10-46%), nadir Sp O 2 during sleep below a defined threshold (76% or 80%), mean Sp O 2 during sleep below 90%, or a combination of Sp O 2 and bicarbonate level.…”

“…We identified 31 studies of interest and excluded 18 studies because they did not report the data required to answer the posed question. Of the remaining 13 articles we reviewed, 9 were full manuscripts ( 40 – 48 ) and 4 were in abstract form ( 49 – 52 ). Six were prospective ( 42 , 43 , 45 , 47 , 51 , 52 ), six were retrospective, and one had a retrospective and a prospective design ( 40 ).…”