A Guide to Sleep Testing
How do I know if I have a sleep disorder? Should I have a sleep test if I drive professionally? Are there different types of sleep tests? How do I know if I need the kind of expensive sleep test that requires overnight observation? As a lung physician, I meet people daily who ask these very questions.
Written tests have been developed to predict the probability of sleep-related disorders, and in some cases they are required by insurance companies as a requisite to determine that more testing is needed. You may be wondering, just how accurate are these preliminary written sleep questionnaires? A study published in the Journal of Clinical Sleep Medicine assessed the accuracy of a test called the Berlin questionnaire.
The Berlin questionnaire is a predictive test for obstructive sleep apnea (OSA). It is a relatively simple test and it takes into account history of snoring and witnessed apneas by bed partner as well as daytime fatigue and sleepiness and the presence of as a co-morbid condition. A total score is given based on the person’s answers, and that gives a clinician the pre-clinical probability of sleep apnea.
The importance of this study is that it was done in the general population, not in the primary care setting when the doctor (already) has a suspicion that leads him to administer the test. This approach to the study was intentional because sleep apnea remains (highly) undiagnosed in the general population. The positive predictive value (the likelihood of being accurate when it is positive) of the questionnaire and the negative predictive value (the likelihood of being accurate when it is negative) were measured and compared with the results of a home sleep test.
The study had some interesting mixed results. The questionnaire was more accurate, and identified the possibility of a sleep disorder, when the patient had a more dramatic or severe form of the disease. Even more important was the finding that the questionnaire was accurate when the results were negative (meaning that sleep apnea was not present). Overall, a positive test was not as predictive. That shortcoming is likely due to the fact that the questions are highly subjective. People have a varied perception of snoring behaviors and their personal level of fatigue.
The other problem with this study was the comparison of the Berlin questionnaire specifically to a portable home study. These home tests, although convenient, are flawed in their accuracy. They measure breathing events, but they do not measure actual sleep (quality, duration). It would be too cumbersome to have a full EEG in the home for testing, which explains the limitation of testing done in the home.
The home test divides the number of abnormal events by the duration of the recording time (a simple formula) so it can also underestimate the severity of the disease. If someone, for example, actually slept much less than the recording time, then the number of abnormal breathing events per hour of sleep would be underestimated by the home study. So it’s not surprising that the home sleep test would likely be accurate only in the case of truly severe disease.
The study really showcases why there is no good substitution for face-to-face contact with a physician who will elicit a precise and extensive sleep history from the patient, and who then forms a clinical suspicion based on the information (including identifying other medical conditions) and a full physical exam.
Care must be taken when interpreting the results of a sleep questionnaire. How questions are asked can “color” the answers patients offer. Misinterpreting the questions is all the more possible when a piece of paper is put in front of the patient rather than when a health care professional does an in-person interview.
It’s important to note that not everyone is a good candidate for the home sleep test. Individuals with heart and lung disease, patients who may have central sleep apnea, and individuals with neurological symptoms such as periodic limb movements should undergo a sleep lab-based study.
It’s clear that the initial evaluation done by a clinical person is crucial to identifying who should have sleep testing. Unfortunately, the current trend with many insurers is that, before a consultation can even take place, an inexpensive home sleep test must be done to determine the need for further evaluation.
Other commonly used written tests:
The Epworth Sleepiness Scale — This test assesses the degree of “tendency to doze off” in eight different situations. It helps to identify a level of daytime sleepiness that is considered abnormal.
The Insomnia Severity Scale — This asks questions to elicit an individual’s perception of the severity of insomnia and how it affects the person’s daytime activities.
The Pittsburgh Sleep Quality Index — This questionnaire is a more involved test that has 32 questions. The information the person supplies is then correlated with sleep quality and associated with psychological consequences.
These tests have a role in diagnosis of sleep disorders like sleep apnea, but their value depends on the clinician who uses the information in the context of the individual. Important considerations are an assessment of the initial presenting complaint, whether it was the patient or the bed partner who initiated the complaint, and the other aspects of the patient’s general health.