Who needs a sleep test? The value of the history in the diagnosis of obstructive sleep apnea.

Introduction

Basing a diagnosis of obstructive sleep apnea (OSA) and its milder form, upper airway resistance syndrome (UARS), solely on a history of loud snoring is similar to diagnosing myocardial ischemia solely on a history of chest pain. Although there are some patients who can be diagnosed with OSA or UARS on the basis of history alone, most patients present more of a diagnostic dilemma.

The history might be able to suggest the presence or absence of sleep disordered breathing, but are its sensitivity and specificity adequate? Because definitive testing--such as overnight polysomnography (PSG) performed in a sleep laboratory or a home sleep evaluation--can be expensive and inconvenient, there is a need for an accurate algorithm that would help the physician diagnose or rule out OSA/UARS based on the history alone.

In one sense, the solution to the diagnostic dilemma is easy: Simply perform a screening PSG on every patient who has a history of loud snoring and sleepiness. Many authors agree that this is the ideal approach, but it requires no judgment and begs the question. [1] Obviously, the physical examination, particularly the examination of the upper airway, adds much to making the diagnosis.

I have no perfect algorithm, but I will cover the points that I consider to be the most important in making the diagnosis of OSA/UARS by history alone.

Key components of the history

Loud snoring. Snoring of any intensity raises the possibility of OSA. Intuitively, the louder the snoring, the more likely that there is significant upper airway obstruction during sleep, although this is not always the case. I define excessively loud snoring as snoring that disturbs the bed partner's sleep.

Witnessed apneas. Apneas witnessed by the bed partner, particularly those of prolonged duration ([greater than]10 sec), are one of the most important parts of the history. The number of these events per night is also important, because the greater the number of events, the more likely that significant OSA exists. The history from the bed partner, if available, is important and should be sought.

Daytime somnolence. The end result of OSA is usually excessive daytime somnolence. A detailed history can be obtained fairly quickly, particularly when a standard form such as the Epworth sleepiness scale is used. [2] This scale provides some quantitation of the likelihood of OSA or UARS. The problem is that many of us who do not have OSA are tired during the day as well, particularly in the late afternoon. More significant is a history of falling asleep while driving or working with machinery, especially an episode that resulted in an accident or near accident. A person who has experienced this type of event has a more severe problem that stresses the need for a workup.

Obesity. Although factors concerning weight are usually considered as part of the physical examination, a history of weight gain and excessive weight ([greater than or equal to]20% above the insurance table for age and height) suggests the presence of OSA. A large collar size is also suggestive.

Male gender. OSA is more common in men than in women.

History of hypertension. Because hypertension is one of the sequelae of OSA, its presence can be helpful in making the diagnosis.

History of coronary artery disease. Awareness of a history of coronary artery disease is important in making sure that the diagnosis of OSA is not missed. The combination of OSA and hypoxia is more likely to lead to a serious cardiac event in a patient who has coronary artery disease than in a patient whose heart is normal.

History of nasal obstruction or mouth breathing. Loud snoring usually requires that the mouth be open during sleep. This also might contribute to airway obstruction by the tongue base in the prone position.

Other factors. There are four less important clues that might be helpful when gleaned from the history: 1) a slow awakening from a general anesthetic, which suggests central apnea; 2) a parent's report of the presence of nightmares and enuresis in children or a history of narcolepsy or restless legs syndrome, any of which might lead to the diagnosis of other sleep disorders; 3) morning headache, a nonspecific symptom that is regularly found in patients with OSA; and 4) a history of alcohol ingestion in the evening or the use of short-acting hypnotics or sedatives at bedtime, which can be contributing causes of OSA.

In my experience, the profile of the typical GSA patient is a slightly obese, middle-aged man who snores, who is mildly tired later in the day, and who has a normal or slightly elevated Epworth score and a normal PSG. The diagnosis is "older, tired, male person" (OTMP); no ICD-9 code is available.

Treatment without testing

Several effective office surgical procedures are available to treat snoring, including laser-assisted uvulopalatoplasty (LAUP), uvulectomy, and radiofrequency palate-tightening procedures. The elimination of snoring resolves a major symptom of obstructive sleep apnea, but at the same time it can give the patient and physician a false sense of security. The absence of snoring does not necessarily mean the absence of sleep apnea; in fact, apnea only becomes more difficult to detect. The elimination of daytime sleepiness and nighttime snoring suggests that a "cure" was achieved, but this might not be the case unless the cure is validated by testing.

Is the relief of symptoms enough reason not to test? What about other practitioners--dentists, family physicians, etc.--who prescribe antisnoring devices without conducting a sleep study? These types of circumstances support the argument that every patient should undergo a sleep study before any attempt is made to correct snoring.

I might be in the minority, but I believe that with a careful history and physical examination, there is no need to study every patient who snores. However, those patients who are not studied should be informed that they do have a risk of OSA or UARS. Furthermore, the surgeon should be aware that postoperative swelling after a snoring correction procedure such as a LAUP can temporarily worsen a preexisting sleep apnea condition. [3] Patients whom I might not screen are those who do not have any witnessed apneas, who do not have excessive daytime sleepiness or heart disease, who are not obese, and whose physical evaluation is not consistent with OSA--that is, there is no large tongue, long uvula, recessivejaw, or short, fat neck, etc.

At this time, I believe that we do not have an ideal screening test for sleep apnea that could be conducted with simple equipment, that would allow a person to be tested at home with minimal effort and cost, and that would be very likely to rule out sleep apnea. In another article in this special series on sleep disorders, which will appear in a subsequent issue of EAR, NOSE & THROAT JOURNAL, Davidson et al describe a home PSG unit that they feel could provide just such a test, and more. [4] Time and experience will tell.

Certainly, one can screen for hypoxia as well as certain cardiac arrhythmias at home with a recording pulse oximeter, but oximetry tests for only part of the OSA problem. Is it enough? There is disagreement in the literature, but the general feeling is that oximetry is not adequate. I think it is better than no testing at all, as long as we understand what it is and is not measuring. I believe the diagnosis of nocturnal hypoxia is an important part of screening. If our test criteria are such that the sensitivity is high, say near 100% (i.e., a low false-negative rate), and the specificity is lower (i.e., a higher false-positive rate), we should be able to diagnose any significant OSA with hypoxia before we send the patient for definitive preoperative testing.

Series et al have described home oximetry criteria that provide a sensitivity of 100% and a specificity of 39% for OSA compared with PSG. [5] When overnight oxygen saturation is abnormal or when the patient has significant bradycardia or tachycardia, then more detailed testing should be performed. Using Series' criteria, a normal home oximetry finding rules out hypoxia that produces OSA, but it does not rule out a sleep disorder that can contribute to fragmented sleep and daytime sleepiness without hypoxia. Home PSG units will continue to become more user-friendly, and I feel that when their ease of use and cost begin to approach those of a recording pulse oximeter, then we all will be inclined to test every patient we suspect of OSA.

Abstract

Many experts believe that a polysomnogram to screen for obstructive sleep apnea should be performed on every patient who has a history of loud snoring and sleepiness. In contrast, the author believes that with a careful history and physical examination, there is no need to study all such patients, at least not until home polysomnography units become as convenient and economical as pulse oximetry.

From the Division of Otolaryngology-Head and Neck Surgery, Stanford (Calif.) University Medical Center, and the Palo Alto (Calif.) VA Healthcare System.

References

(1.) Tami TA, Duncan HJ, Pfleger M. Identification of obstructive sleep apnea in patients who snore. Laryngoscope 1998;108:508-13.

(2.) Johns MW. Daytime sleepiness, snoring, and obstructive sleep apnea: The Epworth Sleepiness Scale. Chest 1993;103:30-6.

(3.) Terris DJ, Clerk AA, Norbash AM, Troell RI. Characterization of postoperative edema following laser-assisted uvulopalatoplasty using MRI and polysomnography: Implications for the outpatient treatment of obstructive sleep apnea syndrome. Laryngoscope 1996;106:124-8.

(4.) Davidson TM, Do KL, Justus S. The use of ENT-prescribed home sleep studies for patients with suspected obstructive sleep apnea. Ear Nose Throat J 1999, in press.

(5.) Series F, Marc I, Cormier Y, La Forge J. Utility of nocturnal home oximetry for case finding in patients with suspected sleep apnea hypopnea syndrome. Ann Intern Med 1993;119:449-53.