C. Frank Ryan

Diagnosis and Initial Management of Obstructive Sleep Apnea without Polysomnography: A Randomized Validation Study

Context Overnight polysomnography in a sleep laboratory is normal practice for diagnosing obstructive sleep apnea (OSA) but it is expensive and can delay diagnosis. Contribution The authors combined standard clinical scales and overnight home oximetry to ensure a pretest probability of OSA of 90% or greater. Sixty-eight patients were randomly assigned to usual care (polysomnography obtained before continuous positive airway pressure [CPAP]) or ambulatory management (start CPAP without doing polysomnography). After 3 months, the 2 groups had the same results on overnight polysomnography. Cautions The study was done in a single tertiary care center. Implications Most patients with a probability of OSA of 90% or greater do not require polysomnography before starting CPAP. The Editors Symptomatic obstructive sleep apnea is a common, underdiagnosed condition that occurs in 4% of men and in 2% of women (1). Patients with obstructive sleep apnea have considerable comorbid conditions, including excessive daytime sleepiness; concentration difficulties; and an increased risk for motor vehicle accidents, hypertension (2), coronary artery disease, and strokes. Obstructive sleep apnea may result in a systemic inflammatory state that predisposes the patient to cardiac and cerebrovascular conditions (3). Continuous positive airway pressure (CPAP) is an effective treatment that is commonly prescribed for symptomatic patients with obstructive sleep apnea: It is cost-effective (4) and reduces daytime sleepiness, rates of motor vehicle accidents, and blood pressure. The American Thoracic Society (5) and the American Academy of Sleep Medicine (6) recommend supervised polysomnography (PSG) in the sleep laboratory over 2 nights for diagnosis of obstructive sleep apnea and initiation of CPAP. This approach to a highly prevalent condition results in inevitable discrepancies between the demand for services and the current capacity of sleep laboratories (7). Various strategies have been proposed to expedite diagnosis and treatment for obstructive sleep apnea (8). Predictive algorithms (915) and widespread use of overnight home monitoring, such as oximetry, have improved access to diagnostic testing (1622). Several algorithms have been used to determine optimal CPAP (23, 24). Ambulatory use of autotitrating CPAP machines is effective in determining therapeutic CPAP (25) and as a treatment method. To our knowledge, this is the first study to examine a combined ambulatory diagnostic and treatment algorithm without PSG in initial management of obstructive sleep apnea. To do so, we performed 2 parallel validation studies: 1) a cross-sectional study in patients screened by the diagnostic algorithm to test its use in identifying high-probability patients using PSG as the gold standard, and 2) a randomized trial of PSG versus ambulatory CPAP titration in high-probability patients identified by the diagnostic algorithm. Our study asked whether the conventional PSG approach was superior to the ambulatory approach in terms of controlling obstructive sleep apnea as measured by the apneahypopnea index (AHI) on CPAP after 3 months of treatment. We also wanted to determine whether there was any difference between the 2 management strategies in terms of sleepiness, quality of life, treatment adherence, and CPAP after 3 months of treatment. Methods Study Design We designed a randomized, controlled, open-label clinical trial to compare PSG with an ambulatory algorithm for titration of effective CPAP in patients with a high probability of moderate to severe obstructive sleep apnea. Patient Selection Participants were recruited from among adult patients referred from the catchment area of the Sleep Disorders Program at University of British Columbia Hospital, Vancouver, British Columbia, Canada, between May 2004 and November 2005, for assessment of suspected obstructive sleep apnea. Consecutive patients who, on the basis of a routine clinical evaluation by their sleep physician, were suspected of having moderate to severe obstructive sleep apnea, met the inclusion criteria, and were referred for possible participation in the trial, were considered for recruitment. Eligible patients had a high pretest probability of moderate to severe obstructive sleep apnea, were medically stable, and were not taking any sedative medications. We excluded patients who were pregnant or who had abnormal results on spirometry (predicted forced vital capacity or FEV1 <70%); a known cause of daytime sleepiness; a major psychiatric disorder; a life-threatening comorbid illness, such as unstable coronary artery disease or chronic lung disease; a motor vehicle accident attributable to hypersomnolence in the preceding 5 years; previous treatment for obstructive sleep apnea; a contraindication to nasal CPAP therapy; or the inability to provide informed consent. Each eligible patient provided written informed consent, and our institutional ethics review committees approved the protocol. Patients were enrolled by the research coordinator who collected the baseline data. Determining Pretest Probability High-probability patients were identified by sequential application of the Epworth Sleepiness Scale (ESS), Sleep Apnea Clinical Score (SACS) (11), and overnight oximetry in the home. The range of possible scores on the ESS is 0 to 24. In a retrospective case series of all patients (n= 798) referred to our sleep clinic with suspected obstructive sleep apnea who had a full diagnostic overnight PSG between 1 January 2001 and 31 December 2001 (26), the prevalence of moderate to severe obstructive sleep apnea (AHI >15/h) was 49% among patients with an ESS score of 10 or greater. The SACS is a screening tool based on snoring, witnessed episodes of apnea, neck circumference, and systemic hypertension that can be used to calculate likelihood ratios for the presence of obstructive sleep apnea (11). A score of 15 or greater gives a likelihood ratio of 4.45 of having moderate to severe sleep apnea (11). The Remmers Sleep Recorder (SagaTech Electronics Inc., Calgary, Alberta, Canada) is an easy-to-use, multichannel portable device that measures oxygen saturation, respiratory effort, airflow, snoring, and leg movements. This device calculates a respiratory disturbance index (RDI) on the basis of oxygen desaturation events. An RDI of 15/h or greater measured by the Remmers Sleep Recorder has a sensitivity of 98% and a specificity of 88% for diagnosis of moderate to severe obstructive sleep apnea, giving a likelihood ratio of 8.1 (18). Starting with pretest odds of approximately 1:1 on the basis of an ESS score of 10 or greater, we used Bayes theorem and assumed no interaction between components of the algorithm. The combined likelihood ratio conditional on a SACS of 15 or greater and an oximetry RDI of 15/h or greater yielded an estimated pretest probability of moderate to severe obstructive sleep apnea greater than 95%. Protocol Before randomization, all recruited patients completed the Sleep Apnea Quality of Life Index (SAQLI)a comprehensive survey with a high degree of internal consistency, face validity, and construct validity designed to measure outcomes of clinical trials in sleep apnea (27, 28). All patients received a one-on-one orientation and education session from a dedicated CPAP coordinator who provided information in a standardized fashion regarding nasal CPAP therapy. This session included mask fitting and a trial of CPAP while the patient was awake to ensure that he or she could tolerate the device (Appendix). The CPAP coordinator randomly assigned patients to standard PSG or the ambulatory algorithm (Figure 1), using a stratified block randomization with a block size of 20. The stratification factors were an ESS score less than 15 versus an ESS score 15 or greater and oximetry RDI less than 30 episodes per hour versus RDI 30 episodes per hour or greater. Four large envelopes were prepared, 1 for each block. There were 20 folded cards in each envelope, 10 for the PSG group and 10 for the ambulatory group. Patients who consented to participate were assigned to a block according to their ESS and RDI scores. Each patient picked 1 card from the designated envelope; the CPAP coordinator noted the treatment allocation and destroyed the card. Blinding patients to their treatment allocation was not possible, but all patients were treated in an identical manner, including follow-up, aside from the interventions being studied. The ResMed AutoSet Spirit (ResMed Inc., Sydney, Australia) autotitrating CPAP machine was used in all patients. Together with the ResMed Autoscan software, this device is capable of storing and downloading data on compliance; mask leak; and CPAP, including the 95th percentile pressure (29)the pressure at or below which the patient spent 95% of the time. Final CPAP was determined according to treatment group. Figure 1. Design of the clinical study. CPAP = continuous positive airway pressure; ESS = Epworth Sleepiness Scale; PSG = polysomnography; SAQLI = Sleep Apnea Quality of Life Index. Polysomnography Group A trained technologist supervised PSG. Obstructive sleep apnea was confirmed during a regularly scheduled overnight PSG in the sleep laboratory at our hospital. Final CPAP was determined according to a standard protocol during a CPAP titration PSG performed on the following night (Appendix). There was no subsequent adjustment to the fixed CPAP in the PSG group. Ambulatory Group The AutoSet Spirit was set to autotitrate at pressures between 4 and 20 cm H2O. After being used for 1 week, the ResMed Autoscan software was interrogated for efficacy data, including CPAP, mask leak, residual respiratory events, and use. The 95th percentile pressure was taken as the initial effective pressure if no residual sleep-disordered breathing was identified. The patient continued treatment at this pressure in fixed CPAP mode for another week. On days 6 and 13, overnight oximetry using

Residual sleep apnea on polysomnography after 3 months of CPAP therapy: Clinical implications, predictors and patterns

OBJECTIVE We sought to determine the clinical implications, predictors and patterns of residual sleep apnea on continuous positive airway pressure (CPAP) treatment in patients with moderate-to-severe obstructive sleep apnea (OSA). METHODS We performed a post hoc secondary analysis of data from a previously reported randomized trial. Sleepy patients with a high risk of moderate-to-severe OSA identified by a diagnostic algorithm were randomly assigned to standard CPAP titration during polysomnography (PSG) or ambulatory titration using auto-CPAP and home sleep testing. We observed them for 3 months and measured apnea-hypopnea index (AHI) on CPAP, Epworth sleepiness scale (ESS), sleep apnea quality of life index (SAQLI), CPAP pressure and objective CPAP compliance. RESULTS Sixty-one patients were randomized, 30 to the PSG group and 31 to the ambulatory group. Fifteen patients (25%) had residual sleep apnea (AHI > 10/h on CPAP) with similar proportions in the PSG (7/30) and ambulatory (8/31) groups. Baseline variables including age, body mass index (BMI), ESS, SAQLI, respiratory disturbance index (RDI) and CPAP pressure did not differ between the groups. Outcomes including compliance were worse in patients with residual sleep apnea. Periodic breathing was prevalent among patients with residual sleep apnea. CONCLUSIONS Residual sleep apnea is common in patients with moderate-to-severe OSA, despite careful CPAP titration, and is associated with worse outcomes.

Oropharyngeal crowding and obesity as predictors of oral appliance treatment response to moderate obstructive sleep apnea.

BACKGROUND Oral appliances are increasingly prescribed for patients with moderate obstructive sleep apnea (OSA) instead of nasal CPAP. However, the efficacy of oral appliances varies greatly. We hypothesized that oral appliances were not efficacious in patients with moderate OSA who were obese with oropharyngeal crowding. METHODS Japanese patients with moderate OSA were prospectively and consecutively recruited. The Mallampati score (MS) was used as an estimate of oropharyngeal crowding. Follow-up polysomnography was performed with the adjusted oral appliance in place. Responders were defined as subjects who showed a follow-up apnea-hypopnea index (AHI) of < 5 with > 50% reduction in baseline AHI. RESULTS The mean baseline AHI was reduced with an oral appliance from 21 ± 4 to 9.8 ± 8 in 95 subjects. Thirty-five patients were regarded as responders. Logistic regression analyses revealed that both MS and BMI could individually predict the treatment outcome. When the cutoff value of BMI was determined to be 24 kg/m2 based on a receiver operating characteristic curve, 53 obese patients (ie, BMI > 24 kg/m2) with an MS of class 4 were indicative of treatment failure with a high negative predictive value (92) and a low negative likelihood ratio (0.28). CONCLUSIONS We conclude that patients with moderate OSA who are obese with oropharyngeal crowding are unlikely to respond to oral appliance treatment. This simple prediction can be applied without the need for any cumbersome tools immediately after the diagnosis of OSA.

Functional contribution of mandibular advancement to awake upper airway patency in obstructive sleep apnea

In the narrowed upper airway of patients with obstructive sleep apnea (OSA), a neuromuscular compensatory mechanism augments the activity of the upper airway dilator muscles in defense of upper airway patency, particularly during inspiration. We hypothesized that mechanical enlargement of the upper airway by a mandibular advancement oral appliance would permit a reduction in this neuromuscular compensation during wakefulness. To test this hypothesis, we focused on changes in the cross-sectional (CS) area of the upper airway before and after emplacement of a ventrally titrated oral appliance in 12 awake OSA patients. The CS areas at the end of tidal expiration (CS area-EET) and at the nadir of intraluminal pressure during inspiration (CS area-IN) were obtained using videoendoscopy. The median apnea–hypopnea index decreased with mandibular advancement. Before mandibular advancement, there was no difference between CS area-EET and CS area-IN in the velopharynx, oropharynx, and hypopharynx. This indicates that upper airway dilator muscle activity increased during inspiration to counteract the intraluminal negative pressure of the upper airway. After mandibular advancement, CS area-EET increased in the velopharynx, oropharynx, and hypopharynx, but CS area-IN was unchanged at any level and was less than CS area-EET in the velopharynx and oropharynx. These findings suggest that mandibular advancement enlarges the upper airway and may reduce upper airway dilator muscle activity during inspiration. We conclude that oral appliances act to return the upper airway towards a normal configuration and pattern of muscle function in OSA patients.

Obstructive sleep apnoea and frequency of occupational injury

Abstract We sought to determine whether patients with obstructive sleep apnoea (OSA) are at increased risk of occupational injury (OI). Patients referred to the University of British Columbia Hospital Sleep Laboratory for suspected OSA (May 2003 to July 2011 were recruited and rates and types of validated OI (that caused at least 1 day of disability) in the 5 years prior to polysomnography were calculated. In a sample of 1236, patients with OSA were twice as likely (OR=1.93, 95% CI 1.06 to 3.50, p=0.03) to suffer at least one OI compared with patients without OSA. This association was attenuated (OR=1.76, CI 0.86 to 3.59, p=0.12) after controlling for confounders. In a secondary analysis, patients with OSA were almost three times more likely (OR=2.88, CI 1.02 to 8.08, p=0.05) to suffer from an injury more likely related to reduced vigilance (eg, a fall or commercial motor vehicle crash) when compared with patients without OSA, and this again was attenuated after controlling for confounders (OR=2.42, CI 0.085 to 6.93, p=0.10).