INTRODUCTION — It is important to detect excessive sleepiness – defined as sleepiness that occurs in a situation when an individual would usually be expected to be awake and alert – because it is associated with morbidity and increased mortality, and can be treated in many cases . However, self-reported sleepiness is generally subjective and imprecise, which has led to the development of tests to quantify an individual's sleepiness [2-5].
The multiple sleep latency test (MSLT) and the maintenance of wakefulness test (MWT) use objective measures to quantify sleepiness. The Oxford SLEep Resistance (OSLER) test, which is a simplified variation of the MWT, is another objective that can be used to indirectly quantify sleepiness. In contrast, the Epworth Sleepiness Scale (ESS) and Stanford Sleepiness Scale (SSS) subjectively quantify sleepiness.
Here, methods for quantifying sleepiness are described, and each of the tests used to quantify sleepiness is discussed. A general approach to the clinical evaluation of excessive daytime sleepiness is discussed separately. (See "Approach to the patient with excessive daytime sleepiness".)
DEFINITIONS — The following definitions are important for understanding the discussion that follows:
●Epoch – An epoch is a standard 30-second interval of a polysomnogram (PSG) to which a sleep stage is assigned. In special situations, an epoch can be longer or shorter.
●Sleep onset – Sleep onset for the multiple sleep latency test (MSLT) is typically defined as the beginning of the first epoch with greater than 15 seconds of cumulative sleep of any stage. The recommended definition for sleep onset for the maintenance of wakefulness test (MWT) is three consecutive epochs of stage 1 sleep or one epoch of any other stage of sleep. (See "Stages and architecture of normal sleep", section on 'Sleep staging'.)
●Sleep latency – Sleep latency is the duration from lights out to the onset of sleep.
●Mean sleep latency – The mean sleep latency is the average of the sleep latencies determined during a test.
GENERAL APPROACH — A combination of subjective tests, objective tests, and clinical history are generally used to determine the degree of sleepiness. It is unknown which tests are best for quantifying sleepiness, and the roles of each are still being defined.
In our clinical practice, we generally use the Epworth Sleepiness Scale (ESS) for initial assessment of sleepiness. We proceed to the multiple sleep latency test (MSLT) if:
●The ESS or the patient's history identifies subjective sleepiness that seems out of proportion to objective findings from a polysomnogram (PSG), or
●The diagnosis of narcolepsy or idiopathic hypersomnia is being considered
The maintenance of wakefulness test (MWT) is primarily used in a research setting to assess an intervention's ability to improve alertness. Some commercial driving companies utilize the MWT to assess a driver's ability to operate a vehicle safely, although the utility of the MWT in clinical practice is limited by the test's inability to accurately predict safety in real world settings. Similarly, the Stanford Sleepiness Scale (SSS) and Oxford SLEep Resistance (OSLER) test are primarily research tools.
MULTIPLE SLEEP LATENCY TEST (MSLT) — The MSLT objectively measures an individual's tendency to fall asleep. It is based on the premise that individuals with a greater degree of sleepiness fall asleep faster than individuals with less sleepiness. The MSLT is considered the standard measurement of sleepiness and has proven to be a sensitive and reproducible test for quantifying sleepiness, regardless of the type of sleep deprivation (partial or complete, acute or chronic) or the underlying pathologic condition [1,6-13].
Indications — The MSLT is part of the routine evaluation of patients who are suspected of having narcolepsy and may also be helpful in the evaluation of patients with suspected idiopathic hypersomnia. While the MSLT may support a diagnosis of narcolepsy or idiopathic hypersomnia, alone it is insufficient to confirm a diagnosis of either. (See "Clinical features and diagnosis of narcolepsy in adults", section on 'Diagnostic evaluation' and "Idiopathic hypersomnia", section on 'Diagnostic evaluation'.)
A repeat MSLT may be warranted if the initial test was affected by inappropriate study conditions, the results are ambiguous or uninterpretable, or the test failed to support a diagnosis of narcolepsy despite strong clinical suspicion.
The MSLT is not indicated as part of the routine evaluation of patients who are suspected of having excessive sleepiness due to obstructive sleep apnea (OSA, a disorder characterized by obstructive apneas and hypopneas), insomnia, circadian rhythm disorders, periodic limb movement disorder, medical disorders, or neurologic disorders . In addition, it should not be performed to assess the effectiveness of therapy.
The MSLT should not be used alone to evaluate the risk for driving, work, or home-related accidents because its validity for this purpose has not been proven . Instead, a combination of clinical history and the MSLT should guide clinical recommendations. (See "Drowsy driving: Risks, evaluation, and management".)
Protocol — It is important that the MSLT be performed as part of a standard protocol to minimize the many variables that can affect sleep latency, the test's primary measure [6,8,15]. A task force from the American Academy of Sleep Medicine (AASM) endorsed the following protocol :
●The patient should be instructed to maintain a regular sleep-wake schedule and adequate sleep times prior to undergoing the MSLT. Adequate sleep should be documented by sleep diary and, when available, actigraphy during the two weeks before the test. If the MSLT is being performed to evaluate persistent sleepiness in a patient with OSA treated with positive airway pressure therapy, clinicians should ensure efficacy and adherence based on review of downloaded data. (See "Downloading data from positive airway pressure devices in adults".)
●Stimulants, central nervous system depressants, and rapid eye movement (REM) sleep-suppressing medications (eg, antidepressants) (table 1) should be discontinued prior to testing because they can impact the ability to fall asleep as well as affect the tendency of an individual to attain REM sleep during testing [16,17]. In most cases, a two-week washout period should be sufficient. For medications with longer half-lives, up to six weeks may be necessary. Medication discontinuation should be discussed in advance and planned for a time when it is least disruptive to patient safety, responsibilities, and productivity. If safe discontinuation is not possible, all medications taken should be listed on the report to aid in interpretation of results.
Other medications, over-the-counter agents, herbal remedies, or other substances with stimulating or sedating side effects should also be minimized. Of note, recent marijuana discontinuation can cause REM rebound, and use should cease at least two weeks before the test.
●Patients should avoid vigorous physical activity, caffeinated beverages, and exposure to bright sunlight on the day of testing. When the patient arrives at the test center, urinary drug testing may be indicated to confirm that sleepiness is not drug-induced. In one study, 16 percent of patients referred for MSLT or a maintenance of wakefulness test (MWT) had a positive urinary drug screen, and none of these patients reported relevant medication or substance use prior to testing .
●A polysomnogram (PSG) is performed prior to the MSLT, preferably during the patient's usual sleep period because the results may be affected by the phase of the individual's circadian clock [12,19,20]. The PSG should allow for a minimum of seven hours in bed with at least six hours of sleep. For patients with obstructive sleep apnea, the overnight PSG and MSLT should be performed with the patient using their continuous positive airway pressure (CPAP) device or oral appliance at its therapeutic setting. Smoking and other stimulating activities should stop at least 30 minutes prior to the PSG.
The purpose of the PSG prior to the MSLT is to objectively characterize the preceding night's sleep and to look for common causes of poor-quality or insufficient sleep. The PSG must confirm at least six hours of sleep for the MSLT to be valid, to minimize the confounding effects of sleep deprivation on the MSLT.
●The MSLT begins 1.5 to three hours after the PSG. The patient is placed in a sleep-inducing environment and instructed to try to sleep. A sleep-inducing environment refers to a dark and quiet room with the room temperature based on the patient's comfort level. Monitoring includes audiovisual recordings, electroencephalography (EEG) with at least three recording leads (frontal, central, and occipital), electrooculography, mental or submental electromyography, and electrocardiography.
●Each nap session continues for 15 minutes after sleep onset to detect any occurrence of REM sleep. The sleep latency is documented for each nap session. Sleep onset is defined as the first 30-second epoch in which more than 15 seconds of the epoch is consistent with any stage of sleep. Sleep latency is defined as the time from lights out of the nap to the first epoch of sleep. If the patient does not fall asleep, the nap session is terminated after 20 minutes and the sleep latency is documented as being 20 minutes.
●This is repeated at two-hour intervals until the patient has had five opportunities to nap. Between nap sessions, the patient should be out of bed and encouraged to stay awake. Smoking and other stimulating activities should stop at least 30 minutes prior to each nap session. The five nap protocol is the recommended protocol. A shorter, four nap test may be performed, but this is unreliable for the diagnosis of narcolepsy unless two or more sleep onset REM periods (SOREMPs) have occurred. (See "Clinical features and diagnosis of narcolepsy in adults", section on 'Diagnostic evaluation'.)
Interpretation — The primary measures from the MSLT are the mean sleep latency and presence or absence of SOREMPs . These measures should be considered along with other clinical features (ie, subjective symptoms) when determining whether the results are clinically significant. Consider the two major indications for the MSLT, suspected narcolepsy or idiopathic hypersomnia:
●A mean sleep latency of eight minutes or less, plus two or more SOREMPs support a diagnosis of narcolepsy in the appropriate clinical circumstances [1,21]. A SOREMP within the first 15 minutes of sleep onset on the preceding overnight polysomnogram can replace one of the SOREMPs on the MSLT . (See "Clinical features and diagnosis of narcolepsy in adults", section on 'Diagnostic evaluation'.)
●A mean sleep latency less than eight minutes, plus fewer than two SOREMPs support a diagnosis of idiopathic hypersomnia in the appropriate clinical setting . (See "Idiopathic hypersomnia", section on 'Diagnostic evaluation'.)
There is no threshold mean sleep latency that, taken alone, can discriminate healthy individuals with normal (ie, expected) sleepiness from individuals with excessive sleepiness due to a sleep disorder. Previously, a mean sleep latency greater than ten minutes was considered normal and a mean sleep latency less than five minutes was considered abnormal [1,6,8,22]. These thresholds are no longer used because they discriminate poorly. Among the reasons that the MSLT discriminates poorly :
●Normal sleep latency has not been established. There are no large systematically collected data sets that include the mean sleep latency of healthy individuals. In addition, this information cannot be pooled from existing studies because variables that affect mean sleep latency differ among the studies, including the definition of sleep onset, the number of nap sessions, and the age of the patient population.
●There is a wide range of sleep latency among healthy individuals. As a result, there is significant overlap in the sleep latencies of normal individuals and individuals with sleep disorders. In a population-based study that included over 1000 patients who underwent MSLT, 22 percent of individuals had a mean sleep latency ≤8 minutes; the most important predictors of an abnormal MSLT were shift work and chronic sleep deprivation .
●Demographic variance is also seen in individuals referred for evaluation of excessive sleepiness. In a study that included review of nearly 2500 MSLTs performed for hypersomnia, age, gender, and use of REM-suppressing medications all had significant effects on MSLT outcomes . Age had a U-shaped effect, with children (≤12 years) and older adults (≥60 years) having longer sleep latencies and fewer SOREMPs compared with adolescents and younger adults. Sleep and REM also became increasingly less likely with successive naps, especially in young children. As expected, the use of REM-suppressing medication (antidepressants, antipsychotics) was independently associated with decreased SOREMPs.
When interpreting the MSLT, the clinician must be attentive to whether there was adherence to the protocol. Strict adherence to the protocol is required for valid and reliable results because arousing signals (eg, noise, caffeine, smoking) can interfere with the occurrence of sleep [1,8,24,25]. One study evaluated healthy individuals who underwent the MSLT after either a five-minute walk or a 15-minute rest in bed while watching television . Sleep latencies were significantly increased in the group that walked prior to the MSLT, regardless of whether the individuals were sleep deprived.
Limitations — The MSLT may not reliably measure objective sleepiness when sleepiness is due to chronic insomnia. Several studies have demonstrated that patients with chronic insomnia have a mean sleep latency that is significantly longer than that of matched healthy controls, despite patients with insomnia having chronically reduced total sleep times [26-28]. This reflects the difficulty that patients with insomnia have initiating sleep, rather than being an accurate measure of their sleepiness . (See "Risk factors, comorbidities, and consequences of insomnia in adults".)
Other disadvantages of the MSLT are that it is time consuming (6.5 to 8.5 hours), labor intensive, and expensive to perform. Reducing the MSLT protocol to two or three naps is not an acceptable alternative. When this is done, the correlation coefficient for reproducibility of the mean sleep latency (over a four to fourteen month interval) is reduced from 0.97 to 0.85 for a three nap protocol and 0.65 for a two nap protocol .
MAINTENANCE OF WAKEFULNESS TEST (MWT) — The MWT objectively measures the ability of an individual to remain awake for a defined period of time. It is based on the premise that individuals with a greater degree of sleepiness are less likely to remain awake than individuals with less sleepiness.
The MWT should not be considered a substitute for the multiple sleep latency test (MSLT), as the tests measure different processes: the tendency to fall asleep (ie, the MSLT) and the ability to stay awake (ie, the MWT). Because the tests measure different tendencies, the tests can give conflicting results, even when the same individual is given both tests on the same day .
Different results are in part related to the instructions given to patients prior to each nap trial. During the MSLT, patients are instructed to "lie quietly, assume a comfortable position, keep your eyes closed, and try to fall asleep." During the MWT, patients are instructed to "sit still and try to remain awake for as long as possible. Look directly ahead of you and do not look directly at the light." In addition, during the MWT, patients are instructed to avoid extraordinary measures to stay awake (eg, slapping the face, singing). The different postures maintained during each test (ie, lying flat in bed for the MSLT versus sitting in a recumbent position for the MWT) may also affect the results [25,31].
Indications — The MWT may be used to assess an individual's response to therapy. It is the direction of change, not the degree of change, that is meaningful in this situation because the degree of change that is clinically significant has not been established.
Protocol — The MWT should be performed following a standard protocol. Using a protocol minimizes the variables that can impact sleep latency, the test's primary measure. Several acceptable protocols exist including the following, which was endorsed by a task force from the AASM :
●In preparation, clinicians and patients should define goals for adequate sleep. Similar to the MSLT, adequate sleep should be documented by sleep diary and actigraphy, whenever possible, for two weeks prior to the study. (See 'Protocol' above.)
●Patients should maintain their normal routine prior to the test. Upon arrival, they should be questioned to determine whether their sleep prior to the test was adequate in quality and quantity, and whether they feel alert. The MWT should be delayed if the patient reports suboptimal sleep or not feeling alert. A polysomnogram (PSG) on the prior night is not necessary. Urine drug testing may be indicated to ensure that the result is not influenced by substances other than prescribed medications and is usually performed on the morning of the MWT or as directed by the sleep clinician. Stimulating activities such as consuming nicotine and use of electronic devices should end at least 30 minutes prior to each wake trial.
●The MWT begins 1.5 to three hours after the patient's usual wake-up time. The patient is placed in a room with little or no external light. The only light source should be dim, slightly behind the patient's head, and just out of the patient's field of vision. The room temperature is based on the patient's comfort level. The patient sits upright in bed or reclining chair, with their back and head supported, and is instructed to try to stay awake as long as possible. Monitoring includes audiovisual recordings, electroencephalography (EEG) with at least three recording leads (frontal, central, and occipital), electrooculography, mental or submental electromyography, and electrocardiography.
●A session is ended after unequivocal sleep, or after 40 minutes if sleep does not occur. Sleep is considered unequivocal after three consecutive epochs of stage 1 sleep or one epoch of any other stage of sleep. For each session, the sleep latency is recorded. It is documented as being 40 minutes if the patient does not fall asleep.
●This is repeated every two hours, until the patient has completed four sessions.
Interpretation — The primary measure from the MWT is the mean sleep latency . There are few data regarding what constitutes a normal mean sleep latency, as measured by the MWT. Among healthy individuals who complete the four session, 40 minute protocol described above, the mean sleep latency is approximately 30 minutes, with greater than 97 percent of individuals having a mean sleep latency of eight minutes or greater . As a result, a mean sleep latency less than eight minutes is generally considered abnormal. Staying awake for at least 40 minutes during all four sessions is strong objective evidence that an individual can stay awake. A mean sleep latency between 8 and 40 minutes has uncertain significance.
The relationship between MWT findings and driving risk are discussed separately. (See "Drowsy driving: Risks, evaluation, and management", section on 'Evaluation of drowsy drivers'.)
Limitations — Disadvantages of the MWT include its long duration and high cost. In addition, it is labor intensive. Although the MWT has been used to evaluate the risk for driving, work, or home-related accidents, its validity for this purpose has not been proven .
The MWT has demonstrated little correlation with obstructive sleep apnea (OSA) severity as defined by the apnea hypopnea index (AHI) and is currently not recommended as a measure to assess safety in commercial truck drivers with OSA. In addition, while the Federal Aviation Administration (FAA) recommends that an MWT be considered in pilots with OSA who are not compliant with therapy, the FAA does not recommend a specific MWT protocol or threshold value below which flying would be deemed unsafe.
EPWORTH SLEEPINESS SCALE (ESS) — The ESS subjectively measures sleepiness as it occurs in ordinary life situations . It can be used to screen for excessive sleepiness or to follow an individual's subjective response to an intervention.
Eight situations are described on a questionnaire:
●Sitting and reading
●Sitting inactively in a public place
●Riding as a passenger in a car for one hour without a break
●Lying down to rest in the afternoon when circumstances permit
●Sitting and talking with someone
●Sitting quietly after lunch without alcohol
●Sitting in a car as the driver, while stopped for a few minutes in traffic
Each situation receives a score of zero to three, which is related to the likelihood that sleep will be induced:
●0 = would never doze
●1 = slight chance of dozing
●2 = moderate chance of dozing
●3 = high chance of dozing
Thus, the total ESS score can range from zero to 24, with higher scores correlating with increasing degrees of sleepiness (calculator 1).
A score greater than ten is consistent with excessive sleepiness. This threshold between normal and abnormal subjective sleepiness is derived from an observational study of 180 adults . The mean ESS score was approximately six among healthy adults, compared to 16 or greater among patients with narcolepsy, idiopathic hypersomnia, or moderate to severe obstructive sleep apnea (OSA, defined as a respiratory disturbance index [RDI] greater than 15 events per hour).
The ESS can be performed in the examination room, waiting room, or at home prior to or in between visits via electronic medical record-based questionnaire. It is relatively simple and generally takes only a few minutes to complete. It should be repeated at subsequent visits to assess for change.
Correlation — Subjective sleepiness measured by the ESS appears to correlate moderately with objective sleep tendency. A systematic review of 35 studies examining the psychometric properties of the ESS in adults concluded the following :
●Although the ESS is widely used in research and clinical practice, there have been few high-quality studies on its test properties. The internal consistency of the ESS is relatively high, suggesting that it is appropriate to use for group-level comparisons.
●The ESS correlates moderately well with objective measures of sleepiness such as the multiple sleep latency test (MSLT) and maintenance of wakefulness test (MWT; pooled correlation coefficients -0.4 and -0.3, respectively). The ESS correlates weakly with measures of obstructive sleep apnea severity such as the apnea hypopnea index or oxygen saturation.
●The test-retest reliability of the ESS has not been well studied.
Advantages — The ESS is widely available, reliable for assessing subjective sleepiness, and can be performed easily and quickly.
Disadvantages — The ESS cannot be used on multiple occasions throughout the day and is not a good measure of short-term acute sleepiness. Although the ESS score tends to increase with the severity of OSA and many patients with moderate to severe OSA have an ESS score greater than 10, the ESS should not be used to screen for OSA because it is neither sensitive nor specific for this purpose [34,35]. High test-retest variability has also been identified as a potential limitation in patients being referred for polysomnography .
STANFORD SLEEPINESS SCALE (SSS) — The SSS is the best validated subjective measure of sleepiness [3,37]. It is typically used as a research tool to measure the impact of short-term acute sleep loss on subjective sleepiness. During the SSS, one of seven statements is chosen that best describes an individual's level of sleepiness:
●1 = feeling active, vital, alert, wide awake
●2 = functioning at a high level, not at peak, able to concentrate
●3 = relaxed, awake, not at full alertness, responsive
●4 = a little foggy, not at peak, let down
●5 = fogginess, losing interest in remaining awake, slowed
●6 = sleepiness, prefer to be lying down, fighting sleep, woozy
●7 = almost in reverie, sleep onset soon, losing struggle to remain awake
Individuals who choose the fourth, fifth, sixth, or seventh statement at a time when they should be feeling alert, may have excessive sleepiness.
Advantages of the SSS are that it can be administered multiple times throughout the day and night and it correlates with standard measures of performance. In addition, it correlates well with acute total sleep deprivation (but, less well with chronic partial sleep deprivation). This may be related to an individual's perception of their ability to get used to chronic sleep loss.
The major disadvantage of the SSS is its inability to differentiate sleep deprived normal subjects from individuals with sleep disorders. There also appears to be some discordance between gross behavioral indicators of sleep (eg, closed eyes) and the first and second choices . Finally, the SSS only conveys information about a patient's state of sleepiness at single points in time.
OSLER TEST — The OSLER (Oxford SLEep Resistance test) test is a simplified version of the maintenance of wakefulness test (MWT) . It has been used as a surrogate for the MWT because it is less labor intensive.
The primary difference between the OSLER test and the MWT is that the onset of sleep is detected behaviorally during the OSLER test, instead of by electroencephalography (EEG) criteria. Specifically, the individual sits in front of an LED screen that flashes light every three seconds and is told to touch a button every time the light flashes. Sleep is confirmed when seven consecutive flashes (21 seconds) occur without a response. This is the minimal duration that can define an epoch of sleep . The OSLER test has a sensitivity and specificity of 85 and 94 percent, respectively, for detecting sleep lasting more the three seconds .
The OSLER test is not used much in clinical practice. It has been used to measure sleep tendency in healthy individuals, sleep deprived individuals, and patients with sleep apnea . However, the studies were small and further testing is required to confirm its validity and reliability using larger samples with different diseases.
Advantages of the OSLER test include its simplicity, low cost, and automatic reading (ie, the computer program can determine when the test is complete based on when the subject misses the seven consecutive flashes). In addition, there are relatively few technical requirements that personnel must master .
SUMMARY AND RECOMMENDATIONS
●It is important to detect excessive sleepiness and its underlying cause because excessive sleepiness is associated with morbidity and increased mortality. (See 'Introduction' above.)
●A combination of subjective tests, objective tests, and clinical history are generally used to determine the degree of sleepiness. It is unknown which tests are best for quantifying sleepiness. (See 'General approach' above.)
●In our clinical practice, we generally use the Epworth Sleepiness Scale (ESS) for initial assessment of sleepiness. We proceed to the multiple sleep latency test (MSLT) if the patient's ESS or clinical history detects subjective sleepiness that seems out of proportion to objective findings from a polysomnogram (PSG) and the diagnoses of narcolepsy or idiopathic hypersomnia are being considered. (See 'General approach' above.)
●The MSLT objectively measures an individual's tendency to fall asleep. It is part of the routine evaluation of patients who are suspected of having narcolepsy because it can detect both excessive sleepiness and sleep-onset rapid eye movements (SOREMPs), essential features of narcolepsy. The MSLT may also be helpful in the evaluation of patients with suspected idiopathic hypersomnia. (See 'Multiple sleep latency test (MSLT)' above.)
●The maintenance of wakefulness test (MWT) objectively measures the ability of an individual to remain awake for a defined period of time. It is used in clinical trials to assess an individual's response to therapy. Its utility in clinical practice is unclear because its ability to accurately predict a sleepy individual's ability to work safely or to operate a motor vehicle safely has not been validated. (See 'Maintenance of wakefulness test (MWT)' above.)
●The ESS subjectively measures sleepiness as it occurs in ordinary life situations by scoring eight situations that are described on a questionnaire. The Stanford Sleepiness Scale (SSS) subjectively measures sleepiness by choosing one of seven statements that best describes an individual's level of sleepiness. (See 'Epworth Sleepiness Scale (ESS)' above and 'Stanford Sleepiness Scale (SSS)' above.)
●The OSLER test is a method of measuring sleepiness that is occasionally used in a research setting. The OSLER test is a simplified version of the MWT in which the onset of sleep is detected behaviorally instead of electroencephalographically. (See 'OSLER test' above.)