0
Original Research: Sleep Disorders |

The Utility of the Elbow Sign in the Diagnosis of OSAElbow Sign for Diagnosis of OSA FREE TO VIEW

Mark E. Fenton, MD, FCCP; Karen Heathcote, MD; Rhonda Bryce, MD; Robert Skomro, MD, FCCP; John K. Reid, MD, FCCP; John Gjevre, MD, FCCP; David Cotton, MD, FCCP
Author and Funding Information

From the Division of Respirology, Critical Care and Sleep Medicine, and Clinical Research Support Unit, University of Saskatchewan, Saskatoon, SK, Canada.

Correspondence to: Mark E. Fenton, MD, FCCP, Division of Respirology, Critical Care and Sleep Medicine, 5th Floor Ellis Hall, Royal University Hospital, 103 Hospital Dr, Saskatoon, SK, S7N 0W8 Canada; e-mail: mark.fenton@usask.ca


These data were presented, in part, at CHEST 2012, October 24, 2012, Atlanta, GA.

Funding/Support: The authors have reported to CHEST that no funding was received for this study.

Reproduction of this article is prohibited without written permission from the American College of Chest Physicians. See online for more details.


Chest. 2014;145(3):518-524. doi:10.1378/chest.13-1046
Text Size: A A A
Published online

Background:  Multiple questionnaires have been used to predict the diagnosis of OSA. Such models typically have multiple questions requiring cumulative scoring for interpretation. We wanted to determine whether a simple two-part questionnaire has predictive value in the pretest clinical evaluation for OSA.

Methods:  A questionnaire consisting of two questions—(1) Does your bed partner ever poke or elbow you because you are snoring? and (2) Does your bed partner ever poke or elbow you because you have stopped breathing?—was prospectively administered to patients evaluated in a sleep disorders clinic prior to undergoing polysomnography. Age, sex, BMI, and Epworth Sleepiness Scale data were collected.

Results:  Among the 128 patients who had a polysomnogram, answering “yes” to being awakened for snoring increased the OR of an apnea-hypopnea index ≥ 5/h 3.9 times compared with “no.” Answering “yes” to being awakened for apneic spells was associated with an OR of 5.8 for an apnea-hypopnea index ≥ 5/h compared with “no.” These associations did not differ by sex, BMI, Epworth Sleepiness Scale or answering “yes” to the other question. Subjects > 50 years old with OSA were less likely to report a positive elbow sign and had a significantly lower OR for being awakened for apneic spells than those < 50 years old. The sensitivity and specificity of being awakened for apneic spells was 65% and 76%, respectively, with a positive predictive value of 90%. Subgroup analysis revealed that in men with a BMI > 31 a positive elbow sign had a specificity of 96.6% for a diagnosis of OSA.

Conclusions:  Among patients referred to a sleep disorders clinic, a positive response to being elbowed/poked for apneic spells significantly improves the pretest prediction of OSA.

OSA is a common disorder1 that is associated with hypertension and cardiovascular disease,2 increased risk of motor vehicle accidents,3 and increased health-care costs along with increased absenteeism and decreased productivity in the workplace,4 all of which have significant individual and societal consequences. OSA remains undiagnosed in a significant proportion of the population,5 in part related to underrecognition; however access to testing facilities is a major barrier to diagnosis in many jurisdictions. In 2004, Flemons et al6 reported the wait time for polysomnography (PSG) in Canada was 4 to 35 months, 7 to 60 months in the United Kingdom, and 2 to 10 months in the United States. Other reviews suggest that poor access remains a significant issue.7 As a result, adaptations including increased use of level 3 testing8 and even empirical use of CPAP9 have been implemented. It has also meant that waiting lists for PSG are carefully monitored and triaged, often according to health risk (cardiovascular comorbidities) and societal risk (eg, truck drivers).

Multiple clinical prediction models and questionnaires have been published to aid in the diagnosis of OSA.10 However, such models are cumbersome to use in the clinical setting and have not been widely accepted. Several questionnaires,1116 including the Berlin questionnaire and the STOP-Bang questionnaire, have been developed to improve pretest prediction of OSA. The Berlin questionnaire has been validated in the general population but has limited diagnostic specificity. The STOP-Bang questionnaire was developed as a preoperative screening tool to identify high-risk surgical patients who require diagnostic testing for OSA. As such, it achieves a high sensitivity but low specificity with a score under 4 with improved specificity at higher scores in patients awaiting surgery.17 In a review, Abrishami et al18 reported that many questionnaires are specific to certain populations, but overall have limited sensitivity and specificity in the diagnosis of OSA. All of the models and questionnaires developed to date have multiple questions and domains to remember along with scoring systems that make them cumbersome and of limited usefulness in the clinical setting.

We recognized that many patients presenting to our sleep disorders clinic (SDC) often reported being elbowed or poked by their bed partner because of snoring or witnessed apneic spells. We hypothesized that simply asking about this phenomenon, particularly related to apneic spells, has diagnostic value in identifying patients with OSA.

Ethics approval was obtained from the University of Saskatchewan Biomedical Research Ethics Board (Bio No. 09-173). A simple self-administered questionnaire consisting of the following two questions was developed: (1) Does your bed partner ever poke or elbow you because you are snoring? and (2) Does your bed partner ever poke or elbow you because you have stopped breathing? It was administered prospectively to patients referred with a suspected sleep disorder to the SDC at the University of Saskatchewan. No exclusion criteria were applied, and no specific inquiry as to the existence of a current bed partner was made. Participants’ age, sex, BMI, and Epworth Sleepiness Scale (ESS) score were also collected. At the discretion of the responsible sleep physician, who was blinded to the results of the questionnaire, patients were referred on for further diagnostic testing with either PSG or level 3 testing. Typically, level 3 testing is used for patients without serious comorbidities deemed to have a high pretest probability of OSA. However, such patients may also be referred for PSG depending on other considerations including, but not limited to, occupation and distance to care. The majority of testing within our region is coordinated through our SDC, which offers both attended in-laboratory level 1 PSG and home-based level 3 testing. Our SDC is the only source of level 1 testing locally. However, there are both private (direct cost to patient) and public (indirect cost to patient) avenues for level 3 testing. Patients sent for PSG were placed on a common waiting list. PSG was done according to the usual clinical practice in our SDC independent of individual questionnaire results.

PSG (Sandman software, version 9.1; Mallinckrodt) was done using standard American Academy of Sleep Medicine protocols19 and supervised by registered sleep technicians who were blinded to the questionnaire results. The studies were scored in accordance with the American Academy of Sleep Medicine scoring manual20 by registered sleep technicians and interpreted by sleep physicians, all of whom were blinded to questionnaire results. For the purposes of this study, a PSG result was considered positive for OSA if the apnea-hypopnea index (AHI) was ≥ 5/h in accordance with standard definitions of OSA. Values < 5/h were considered negative. If a split-night protocol was used, only the diagnostic portion of the night was used to determine AHI. Individual management of each patient was left to the discretion of the most responsible physician independent of this study.

Average wait time for PSG in our SDC was approximately 270 days, despite careful triage and implementation of a level 3 care pathway. Because of the unpredictability of the date when all patients would have completed testing, we elected to sample our database 1 year after closing enrollment.

Basic descriptive statistics (mean, median, SD, range, and proportion) were calculated as appropriate to describe the study subjects regarding age, BMI, sex, ESS result, and awakening by sleeping partner for snoring or apnea. χ2 and t test analyses were used to compare categorical and continuous characteristics, respectively, between those with and without a positive PSG result, defining a positive result as an AHI ≥ 5. Positive PSG results were also subdivided by severity into mild, moderate, and severe categories, respectively, defined by AHI values of 5 to 14.9, 15 to 29.9, and ≥ 30. Results were deemed statistically significant at P values < .05, and all analyses were undertaken using SPSS Statistics for Windows, version 20.0 (IBM).

Awakening by partner for snoring and awakening by partner for apnea in relation to PSG outcome were initially examined individually using two separate univariate logistic regression models. To assess the consistency of these associations in different population subgroups, the models were repeatedly evaluated, each time with the inclusion of a different covariate and its interaction with the key snoring or apnea variable. Overall sensitivity, specificity, positive and negative predictive values, and positive and negative likelihood ratios were also calculated. The Mantel-Haenszel χ2 (linear-by-linear association) test was used to assess whether a statistically significant trend toward higher probabilities of partner-initiated awakening could be seen with increasing OSA severity.

Additional consideration was given to the utility of evaluating partner-initiated awakening for snoring or apnea in conjunction with other well-established OSA risk factors, specifically for the purpose of developing an easily recognizable pretest profile of high probability subjects in the context of this referred population. Only the variables of BMI, age, sex, and partner-initiated awakening for snoring or apnea were considered as potential characteristics due to their relatively simple clinical assessment. Again, for ease in real-world application, continuous variables were categorized only utilizing two categories; for effective grouping, various cutpoints were evaluated for each variable, determining divisions with strong positive predictive values (PPVs) and at least a moderate degree of sensitivity. The variables were entered into a logistic regression model, and statistically significant variables were further examined in all possible two- to four-term combinations. Among these, a subgroup of statistically significant characteristics producing the model with greatest estimate precision (narrowest CIs) was sought to maximize reliability of the subsequently generated predicted probabilities of OSA. For clinical utility (ie, to identify a reasonably large number of patients in whom the diagnosis is nearly certain), the final model selected was also required to offer a minimum specificity of 95% in patients possessing the included characteristics and a minimum sensitivity of 20%.

A total of 902 patients were seen in the SDC, 438 (48.6%) of whom consented to participate in this study between October 1, 2009, and September 30, 2010. Three hundred twelve men (71.2%) and 126 women (28.8%) elected to participate. Forty patients (9.1%) were not referred for further testing at the SDC (typically patients with insomnia or patients already on CPAP and one who had already had a diagnostic PSG), and three were excluded because they had only treatment/titration PSG data available. Of the remaining 399 subjects, 129 (29.5%) had a level 1 polysomnogram by the time of data sampling in October 2011. The remaining 269 were either still waiting for testing at the time of sampling the dataset or had alternative (typically level 3) testing done. Only PSG data were used in this analysis, and only one subject was ineligible due to an indeterminate PSG result.

A total of 94 men (73.4%) and 34 women (26.6%) with an average age of 50.8 years, average BMI of 33.9 kg/m2, and median ESS of 12 of 24 underwent PSG. Ninety-seven (75.8%) were found to have an AHI ≥ 5/h. There was no statistically significant difference in ESS score when comparing positive and negative groups (Table 1). There was a significantly higher proportion of men among positive studies (83.5%) compared with negative (41.9%) (P < .001). Those with positive studies were also heavier (BMI, 34.9 vs 30.8; P = .01) and showed a trend to being older (age, 51.8 years vs 47.6 years; P = .053) compared with those with negative studies (Table 2).

Table Graphic Jump Location
Table 1 —Subject Characteristics, Combined and Stratified by OSA

ESS = Epworth Sleepiness Scale; IQR = interquartile range; Max = maximum; Min = minimum; PSG = polysomnography.

a 

Subsequent P values from t test unless indicated.

b 

P value from the Mann-Whitney U test for comparison of median values.

Table Graphic Jump Location
Table 2 —Subject Characteristics, Combined and Stratified by OSA

Data are given as No. (%). See Table 1 legend for expansion of abbreviations.

a 

Subsequent P values from the χ2 test; number of subjects within specific variables may not sum to outcome total due to missing values.

Snoring

A total of 99 subjects reported being poked or elbowed for snoring, 81 of whom had a positive PSG, resulting in a much higher proportion than in the negative PSG group. The positive and negative predictive values were 82% and 46%, respectively. Answering “yes” to being awakened for snoring was associated with an OR of 3.9 for having a positive PSG compared with answering “no.” This association did not show a definitive difference by sex, BMI, ESS score, or answering “yes” to being awakened for apneic spells. χ2 analysis demonstrated that as disease severity increased there was an increased tendency to be awakened for snoring (P = .0004) (Table 3). The sensitivity and specificity for an AHI ≥ 5/h of being awakened for snoring was 84% and 42%, respectively, with positive and negative likelihood ratios of 1.45 and 0.37, respectively (Table 4). Corresponding values for AHI ≥ 15/h and ≥ 30/h are included in Table 4.

Table Graphic Jump Location
Table 3 —Frequency of Partner-Initiated Awakening by Disease Severity

Data are given as No. (%).

a 

χ2 linear-by-linear association for trend.

Table Graphic Jump Location
Table 4 —Sensitivity, Specificity, and Likelihood Ratios of Partner-Initiated Awakening by Disease Severity
Apneic Spells

A total of 68 of 123 subjects who answered the question reported being poked or elbowed for apneic spells, 61 of whom had a positive PSG. This was again higher compared with those with a negative PSG. Answering “yes” to being awakened for apneic spells was associated with an OR of 5.8 for having a positive PSG, and again did not show a clear difference by sex, BMI, ESS score, and response to the snoring question. When comparing this association between younger and older subjects, as defined by ≥ 50 years vs ≤ 50 years, it became apparent that elbowing for apnea is a weaker predictor in the older group compared with the younger group (OR, 18.9 vs 2.5; Breslow-Day; P = .047), due to both decreased elbowing for apnea being reported in the older subjects with OSA and increased reporting for the same in older subjects without OSA. However, χ2 analysis showed an increased likelihood of being awakened for apneic spells as disease severity increased (P < .0001) (Table 3). The overall sensitivity and specificity was 65% and 76%, respectively, with a PPV of 90% and negative predictive value of 40%. The positive likelihood ratio associated with answering “yes” to apneic spells was 2.69 and negative likelihood ratio 0.46. Corresponding values for AHI ≥ 15 and ≥ 30, respectively, are included in Table 4.

Subgroup Analysis: Elbow Sign Use
In Modification of OSA Risk Stratification:

The overall OSA prevalence in this selected population was 75.8%, which generates pretest odds of 3.13 (97 positive PSG/31 negative PSG). We then used the pretest odds to calculate posttest odds in three selected subgroups of subjects in our population corresponding to low risk, intermediate risk, and high risk, respectively (Table 5). Low-risk individuals in this dataset were represented by women younger than 50 years old with a BMI < 30. Intermediate-risk cases were represented by women aged 50 and 59 years old with a BMI under 30. Men younger than 50 years old with a BMI between 30 and 34.9 were recognized as a high-risk group. Applying the likelihood ratios in these groups did not result in meaningful differences between pretest risk and posttest risk in the low- and high-risk groups. In contrast, in the intermediate-risk group, a positive response to being awakened for apneic spells increased the pre-PSG risk from 50% to 73%, which likely has clinical implications in the allocation of diagnostic testing.

Table Graphic Jump Location
Table 5 —Impact of Elbow Sign for Apnea on Stratification of OSA Risk Across Low-, Intermediate-, and High-Risk Groups

Test in this table refers to the question of partner-initiated awakening of apnea. Pretest risk/odds refers to the likelihood of having OSA prior to asking the question. Posttest odds/risk refers to the subsequent probability of OSA after a positive or negative response is received.

To Establish/Confirm OSA Diagnosis:

A logistic regression model was developed to predict OSA (see Materials and Methods section for details). Based on this model, male subjects with a BMI of 31 or higher and a history of partner-initiated awakening for apnea had a 95.1% predicted probability of OSA (Table 6). In our sample, this profile fit 39 of the 128 subjects (29.4%) that received PSG testing, 38 of whom had an AHI ≥ 5/h (PPV = 97.4%, 95% CI [86.5%, 99.9%]). Among the 123 subjects who responded to the question of partner-initiated awakening for apnea, the specificity and sensitivity for these combined characteristics were 96.6% and 40.4%, respectively. In other words, application of this model could have established the diagnosis of OSA (with acceptable certainty) in 38 of our subjects without PSG.

Table Graphic Jump Location
Table 6 —Logistic Regression Model for Clinical Pretest OSA Prediction

PPV = positive predictive value. See Table 1 legend for expansion of other abbreviation.

a 

Model-based probability of positive PSG if all three characteristics present.

This is not the first study to attempt to use clinical variables to predict OSA,1117 and similar to those studies, we used PSG as the reference standard for comparison. However, it differs from previous models and questionnaires in several important respects. Other questionnaires consist of multiple questions and domains that require cumulative scoring to determine the risk of OSA (eg, Berlin questionnaire, STOP-Bang). In contrast, our questionnaire is simple and easy to remember without an associated scoring system. Other questionnaires have been developed as screening tools for either the general population (Berlin) or specific populations such as preoperative patients. The STOP-Bang questionnaire has high sensitivity (84%) and relatively low specificity (56%) for an AHI ≥ 5/h in patients awaiting surgery. This compares favorably to being awakened for snoring in our population, suggesting that asking this question may be used as an alternative to the STOP-Bang questionnaire to screen for OSA in clinic patients. Farney et al21 analyzed the STOP-Bang questionnaire using validated STOP-equivalent questions in a population referred for PSG, a population similar to our population. Three analytical methods were constructed to analyze the data; however, ultimately, a linear model best estimated the probability of OSA according to the categories of none, mild, moderate, and severe based on AHI. As with other approaches, the questionnaire needs to be interpreted by a cumulative scoring method. Their conclusion was that the STOP-Bang questionnaire may be useful in estimating the severity of OSA prior to PSG and, hence, useful in triaging the waitlist for PSG.

In contrast, the elbow sign is simple and in a population already known to have a high pretest probability of OSA (similar to that of Farney et al21) achieves an acceptable degree of specificity (76%) overall, albeit with a low sensitivity (65%) for an AHI ≥ 5/h. Despite the acceptable specificity, using the elbow sign alone will result in missing approximately one-third of patients with OSA. Taking this a step further, men with a BMI of 31 or higher and a positive elbow sign have a profile with high specificity (96.6%) for OSA. In other words, in nearly one-third of our subjects, the elbow sign could be used as an alternative to PSG to establish the diagnosis of OSA. Although the elbow sign appears to be a better predictor in younger subjects, which would be in keeping with previous studies,22 its presence within the criteria here offered an improvement in specificity for the older subgroup (from 85.7% to 100%). This value was on par with specificity seen within the younger subgroup (93.3%), suggesting that individuals who fit the profile are highly likely to have OSA regardless of age. The overall sensitivity of the criteria within the older group did, however, drop from 49.1% to 30.9% in our sample when a positive elbow sign was included, suggesting that the profile will fail to identify more older subjects with OSA than it would younger ones.

Similar to previous studies of witnessed apneas,23,24 we found that women are less likely to report partner-initiated awakening for apneic spells (33.3% vs 70.9%, P = .005), although this difference was only borderline significant for snoring (66.7% vs 87.7%, P = .05) in our data. Shepertycky et al23 reported that women are less likely to be aware of witnessed apneic spells, and Valipour et al24 reported that women with an AHI < 15/h were also less likely to report witnessed apneas. However, when we examined the predictive nature of a positive elbow sign by comparing male and female subjects with OSA to those without OSA, no statistically significant difference was seen between sexes in the degree to which the odds of OSA increase in those with a positive elbow sign when compared with sex-matched counterparts without a positive elbow sign.

The practical application of using the elbow sign in our center is to identify those patients in whom diagnostic testing can be obviated and treatment initiated. Auto-CPAP is an accepted and proven method of CPAP titration8 and can be done to determine individual pressure needs without waiting for diagnostic testing. The downstream effects of this strategy include faster time to treatment of the patient in question and less pressure on the PSG waiting list with resultant decreases in wait times for other patients requiring PSG in whom the diagnosis is less certain or other diagnoses are being entertained.

The economic impact of this approach has not been studied. However, a reduction in direct cost (cost of PSG) and indirect cost (eg, reduction in health-care cost) can be anticipated. The economic impact of OSA has been estimated as similar to the impact of diabetes ($132 billion/y).21 Untreated OSA has been shown to increase health-care utilization; treatment with CPAP reduces this and its associated costs.7,25 Therefore, we anticipate that using this model in our center will result in lower direct and indirect costs by reducing the time to starting treatment and improvements in access overall.

PSG is the accepted gold standard for the diagnosis and staging of OSA26 and was, therefore, deemed to be the best comparator to establish the efficacy of the elbow sign. This is similar to studies of other well-established questionnaires.16,17

There are some obvious limitations to this study. It is a relatively small, single-center study of a population with a high pretest probability for OSA, which limits its application to the general population or even a general medical population. However, the results are likely generalizable to other sleep specialist clinics–a statement that needs to be further evaluated.

The results of this prospective study suggest that among patients referred to an SDC, simply asking if a bed partner awakens a patient because of apneic spells has significant diagnostic value. In a select group of patients, it may obviate the need for PSG, an expensive, time-consuming test with limited availability. Thus, the elbow sign appears to have major implications for the investigation and management of patients with OSA and importantly the management of waiting lists for sleep testing.

Author contributions: Dr Fenton had full access to the data in this study and takes responsibility for the integrity of the data and accuracy of data analysis.

Dr Fenton: contributed to the study design, analysis of data, and writing of the manuscript.

Dr Heathcote: contributed to the study design and manuscript review.

Dr Bryce: contributed to data analysis, writing of the manuscript, and manuscript review.

Dr Skomro: contributed to the revision of the manuscript with important intellectual content.

Dr Reid: contributed to the revision of the manuscript with important intellectual content.

Dr Gjevre: contributed to the revision of the manuscript with important intellectual content.

Dr Cotton: contributed to the revision of the manuscript with important intellectual content.

Financial/nonfinancial disclosures: The authors have reported to CHEST that no potential conflicts of interest exist with any companies/organizations whose products or services may be discussed in this article.

Other contributions: The authors thank Michael Smith, MSc, and Hyun Lim, PhD, for their assistance.

AHI

apnea-hypopnea index

ESS

Epworth Sleepiness Scale

PPV

positive predictive value

PSG

polysomnography

SDC

sleep disorders clinic

Young T, Peppard PE, Gottlieb DJ. Epidemiology of obstructive sleep apnea: a population health perspective. Am J Respir Crit Care Med. 2002;165(9):1217-1239. [CrossRef] [PubMed]
 
McNicholas WT, Bonsigore MR; Management Committee of EU COST ACTION B26. Sleep apnoea as an independent risk factor for cardiovascular disease: current evidence, basic mechanisms and research priorities. [published correction appears inEur Respir J.2007;29(3):614]. Eur Respir J. 2007;29(1):156-178. [CrossRef] [PubMed]
 
Karimi M, Eder DN, Eskandari D, Zou D, Hedner JA, Grote L. Impaired vigilance and increased accident rate in public transport operators is associated with sleep disorders. Accid Anal Prev. 2013;51:208-214. [CrossRef] [PubMed]
 
Teng AY, Won C. Implications of OSA on work and work disability including drivers. Clin Chest Med. 2012;33(4):731-744. [CrossRef] [PubMed]
 
Lurie A. Obstructive sleep apnea in adults: epidemiology, clinical presentation, and treatment options. Adv Cardiol. 2011;46:1-42. [PubMed]
 
Flemons WW, Douglas NJ, Kuna ST, Rodenstein DO, Wheatley J. Access to diagnosis and treatment of patients with suspected sleep apnea. Am J Respir Crit Care Med. 2004;169(6):668-672. [CrossRef] [PubMed]
 
Leger D, Bayon V, Laaban JP, Philip P. Impact of sleep apnea on economics. Sleep Med Rev. 2012;16(5):455-462. [CrossRef] [PubMed]
 
Skomro RP, Gjevre J, Reid J, et al. Outcomes of home-based diagnosis and treatment of obstructive sleep apnea. Chest. 2010;138(2):257-263. [CrossRef] [PubMed]
 
Skomro RP, Cotton DJ, Gjevre JA, et al. An empirical continuous positive airway pressure trial for suspected obstructive sleep apnea. Can Respir J. 2007;14(3):159-163. [PubMed]
 
Flemons WW, Whitelaw WA, Brant R, Remmers JE. Likelihood ratios for a sleep apnea clinical prediction rule. Am J Respir Crit Care Med. 1994;150(5 pt 1):1279-1285. [CrossRef] [PubMed]
 
Weatherwax KJ, Lin X, Marzec ML, Malow BA. Obstructive sleep apnea in epilepsy patients: the Sleep Apnea scale of the Sleep Disorders Questionnaire (SA-SDQ) is a useful screening instrument for obstructive sleep apnea in a disease-specific population. Sleep Med. 2003;4(6):517-521. [CrossRef] [PubMed]
 
Kapuniai LE, Andrew DJ, Crowell DH, Pearce JW. Identifying sleep apnea from self-reports. Sleep. 1988;11(5):430-436. [PubMed]
 
Netzer NC, Stoohs RA, Netzer CM, Clark K, Strohl KP. Using the Berlin Questionnaire to identify patients at risk for the sleep apnea syndrome. Ann Intern Med. 1999;131(7):485-491. [CrossRef] [PubMed]
 
Young T, Palta M, Dempsey J, Skatrud J, Weber S, Badr S. The occurrence of sleep-disordered breathing among middle-aged adults. N Engl J Med. 1993;328(17):1230-1235. [CrossRef] [PubMed]
 
Haraldsson PO, Carenfelt C, Knutsson E, Persson HE, Rinder J. Preliminary report: validity of symptom analysis and daytime polysomnography in diagnosis of sleep apnea. Sleep. 1992;15(3):261-263. [PubMed]
 
Chung F, Yegneswaran B, Liao P, et al. STOP questionnaire: a tool to screen patients for obstructive sleep apnea. Anesthesiology. 2008;108(5):812-821. [CrossRef] [PubMed]
 
Chung F, Subramanyam R, Liao P, Sasaki E, Shapiro C, Sun Y. High STOP-Bang score indicates a high probability of obstructive sleep apnoea. Br J Anaesth. 2012;108(5):768-775. [CrossRef] [PubMed]
 
Abrishami A, Khajehdehi A, Chung F. A systematic review of screening questionnaires for obstructive sleep apnea. Can J Anaesth. 2010;57(5):423-438. [CrossRef] [PubMed]
 
Kushida CA, Littner MR, Morgenthaler T, et al. Practice parameters for the indications for polysomnography and related procedures: an update for 2005. Sleep. 2005;28(4):499-521. [PubMed]
 
Iber C, Ancoli-Israel S, Chesson AL, Quan SF. The AASM Manual for the Scoring of Sleep and Associated Events: Rules, Terminology and Technical Specifications.21st ed. Westchester, IL: American Academy of Sleep Medicine; 2007.
 
Farney RJ, Walker BS, Farney RM, Snow GL, Walker JM. The STOP-Bang equivalent model and prediction of severity of obstructive sleep apnea: relation to polysomnographic measurements of the apnea/hypopnea index. J Clin Sleep Med. 2011;7(5):459-65B. [PubMed]
 
Young T, Shahar E, Nieto FJ, et al; Sleep Heart Health Study Research Group. Predictors of sleep-disordered breathing in community-dwelling adults: the Sleep Heart Health Study. Arch Intern Med. 2002;162(8):893-900. [CrossRef] [PubMed]
 
Shepertycky MR, Banno K, Kryger MH. Differences between men and women in the clinical presentation of patients diagnosed with obstructive sleep apnea syndrome. Sleep. 2005;28(3):309-314. [PubMed]
 
Valipour A, Lothaller H, Rauscher H, Zwick H, Burghuber OC, Lavie P. Gender-related differences in symptoms of patients with suspected breathing disorders in sleep: a clinical population study using the sleep disorders questionnaire. Sleep. 2007;30(3):312-319. [PubMed]
 
AlGhanim N, Comondore VR, Fleetham J, Marra CA, Ayas NT. The economic impact of obstructive sleep apnea. Lung. 2008;186(1):7-12. [CrossRef] [PubMed]
 
Practice parameters for the indications for polysomnography and related procedures. Polysomnography Task Force, American Sleep Disorders Association Standards of Practice Committee. Sleep. 1997;20(6):406-422. [PubMed]
 

Figures

Tables

Table Graphic Jump Location
Table 1 —Subject Characteristics, Combined and Stratified by OSA

ESS = Epworth Sleepiness Scale; IQR = interquartile range; Max = maximum; Min = minimum; PSG = polysomnography.

a 

Subsequent P values from t test unless indicated.

b 

P value from the Mann-Whitney U test for comparison of median values.

Table Graphic Jump Location
Table 2 —Subject Characteristics, Combined and Stratified by OSA

Data are given as No. (%). See Table 1 legend for expansion of abbreviations.

a 

Subsequent P values from the χ2 test; number of subjects within specific variables may not sum to outcome total due to missing values.

Table Graphic Jump Location
Table 3 —Frequency of Partner-Initiated Awakening by Disease Severity

Data are given as No. (%).

a 

χ2 linear-by-linear association for trend.

Table Graphic Jump Location
Table 4 —Sensitivity, Specificity, and Likelihood Ratios of Partner-Initiated Awakening by Disease Severity
Table Graphic Jump Location
Table 5 —Impact of Elbow Sign for Apnea on Stratification of OSA Risk Across Low-, Intermediate-, and High-Risk Groups

Test in this table refers to the question of partner-initiated awakening of apnea. Pretest risk/odds refers to the likelihood of having OSA prior to asking the question. Posttest odds/risk refers to the subsequent probability of OSA after a positive or negative response is received.

Table Graphic Jump Location
Table 6 —Logistic Regression Model for Clinical Pretest OSA Prediction

PPV = positive predictive value. See Table 1 legend for expansion of other abbreviation.

a 

Model-based probability of positive PSG if all three characteristics present.

References

Young T, Peppard PE, Gottlieb DJ. Epidemiology of obstructive sleep apnea: a population health perspective. Am J Respir Crit Care Med. 2002;165(9):1217-1239. [CrossRef] [PubMed]
 
McNicholas WT, Bonsigore MR; Management Committee of EU COST ACTION B26. Sleep apnoea as an independent risk factor for cardiovascular disease: current evidence, basic mechanisms and research priorities. [published correction appears inEur Respir J.2007;29(3):614]. Eur Respir J. 2007;29(1):156-178. [CrossRef] [PubMed]
 
Karimi M, Eder DN, Eskandari D, Zou D, Hedner JA, Grote L. Impaired vigilance and increased accident rate in public transport operators is associated with sleep disorders. Accid Anal Prev. 2013;51:208-214. [CrossRef] [PubMed]
 
Teng AY, Won C. Implications of OSA on work and work disability including drivers. Clin Chest Med. 2012;33(4):731-744. [CrossRef] [PubMed]
 
Lurie A. Obstructive sleep apnea in adults: epidemiology, clinical presentation, and treatment options. Adv Cardiol. 2011;46:1-42. [PubMed]
 
Flemons WW, Douglas NJ, Kuna ST, Rodenstein DO, Wheatley J. Access to diagnosis and treatment of patients with suspected sleep apnea. Am J Respir Crit Care Med. 2004;169(6):668-672. [CrossRef] [PubMed]
 
Leger D, Bayon V, Laaban JP, Philip P. Impact of sleep apnea on economics. Sleep Med Rev. 2012;16(5):455-462. [CrossRef] [PubMed]
 
Skomro RP, Gjevre J, Reid J, et al. Outcomes of home-based diagnosis and treatment of obstructive sleep apnea. Chest. 2010;138(2):257-263. [CrossRef] [PubMed]
 
Skomro RP, Cotton DJ, Gjevre JA, et al. An empirical continuous positive airway pressure trial for suspected obstructive sleep apnea. Can Respir J. 2007;14(3):159-163. [PubMed]
 
Flemons WW, Whitelaw WA, Brant R, Remmers JE. Likelihood ratios for a sleep apnea clinical prediction rule. Am J Respir Crit Care Med. 1994;150(5 pt 1):1279-1285. [CrossRef] [PubMed]
 
Weatherwax KJ, Lin X, Marzec ML, Malow BA. Obstructive sleep apnea in epilepsy patients: the Sleep Apnea scale of the Sleep Disorders Questionnaire (SA-SDQ) is a useful screening instrument for obstructive sleep apnea in a disease-specific population. Sleep Med. 2003;4(6):517-521. [CrossRef] [PubMed]
 
Kapuniai LE, Andrew DJ, Crowell DH, Pearce JW. Identifying sleep apnea from self-reports. Sleep. 1988;11(5):430-436. [PubMed]
 
Netzer NC, Stoohs RA, Netzer CM, Clark K, Strohl KP. Using the Berlin Questionnaire to identify patients at risk for the sleep apnea syndrome. Ann Intern Med. 1999;131(7):485-491. [CrossRef] [PubMed]
 
Young T, Palta M, Dempsey J, Skatrud J, Weber S, Badr S. The occurrence of sleep-disordered breathing among middle-aged adults. N Engl J Med. 1993;328(17):1230-1235. [CrossRef] [PubMed]
 
Haraldsson PO, Carenfelt C, Knutsson E, Persson HE, Rinder J. Preliminary report: validity of symptom analysis and daytime polysomnography in diagnosis of sleep apnea. Sleep. 1992;15(3):261-263. [PubMed]
 
Chung F, Yegneswaran B, Liao P, et al. STOP questionnaire: a tool to screen patients for obstructive sleep apnea. Anesthesiology. 2008;108(5):812-821. [CrossRef] [PubMed]
 
Chung F, Subramanyam R, Liao P, Sasaki E, Shapiro C, Sun Y. High STOP-Bang score indicates a high probability of obstructive sleep apnoea. Br J Anaesth. 2012;108(5):768-775. [CrossRef] [PubMed]
 
Abrishami A, Khajehdehi A, Chung F. A systematic review of screening questionnaires for obstructive sleep apnea. Can J Anaesth. 2010;57(5):423-438. [CrossRef] [PubMed]
 
Kushida CA, Littner MR, Morgenthaler T, et al. Practice parameters for the indications for polysomnography and related procedures: an update for 2005. Sleep. 2005;28(4):499-521. [PubMed]
 
Iber C, Ancoli-Israel S, Chesson AL, Quan SF. The AASM Manual for the Scoring of Sleep and Associated Events: Rules, Terminology and Technical Specifications.21st ed. Westchester, IL: American Academy of Sleep Medicine; 2007.
 
Farney RJ, Walker BS, Farney RM, Snow GL, Walker JM. The STOP-Bang equivalent model and prediction of severity of obstructive sleep apnea: relation to polysomnographic measurements of the apnea/hypopnea index. J Clin Sleep Med. 2011;7(5):459-65B. [PubMed]
 
Young T, Shahar E, Nieto FJ, et al; Sleep Heart Health Study Research Group. Predictors of sleep-disordered breathing in community-dwelling adults: the Sleep Heart Health Study. Arch Intern Med. 2002;162(8):893-900. [CrossRef] [PubMed]
 
Shepertycky MR, Banno K, Kryger MH. Differences between men and women in the clinical presentation of patients diagnosed with obstructive sleep apnea syndrome. Sleep. 2005;28(3):309-314. [PubMed]
 
Valipour A, Lothaller H, Rauscher H, Zwick H, Burghuber OC, Lavie P. Gender-related differences in symptoms of patients with suspected breathing disorders in sleep: a clinical population study using the sleep disorders questionnaire. Sleep. 2007;30(3):312-319. [PubMed]
 
AlGhanim N, Comondore VR, Fleetham J, Marra CA, Ayas NT. The economic impact of obstructive sleep apnea. Lung. 2008;186(1):7-12. [CrossRef] [PubMed]
 
Practice parameters for the indications for polysomnography and related procedures. Polysomnography Task Force, American Sleep Disorders Association Standards of Practice Committee. Sleep. 1997;20(6):406-422. [PubMed]
 
NOTE:
Citing articles are presented as examples only. In non-demo SCM6 implementation, integration with CrossRef’s "Cited By" API will populate this tab (http://www.crossref.org/citedby.html).

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging & repositioning the boxes below.

Find Similar Articles
CHEST Journal Articles
PubMed Articles
  • CHEST Journal
    Print ISSN: 0012-3692
    Online ISSN: 1931-3543