The use of sleep tests for suspected sleep disordered breathing.

Introduction

Sleep testing evolved from early work on dream research and the identification of rapid eye movement (REM) sleep. For many years, testing involved the use of a modified electroencephalogram (EEG) to detect, filter, amplify, and record electrical data obtained from various sites in the brain during sleep. Early recordings were printed and stored on paper and required hand scoring of respiratory, EEG, and other variables. Today, most testing is performed with computer technology, which can record and even analyze data in newer and more sensitive ways to measure airflow, respiratory effort, and sleep.

As awareness of sleep disordered breathing increases and the demand for testing rises, there has been a move to improve access to testing and to reduce its cost. One approach has been to conduct sleep evaluations in the home. This so-called ambulatory testing can be as complex as a laboratory evaluation and measure all of the same parameters, or it can be limited to assessing just a few parameters. The physician who orders unattended ambulatory testing for sleep disordered breathing must be knowledgeable about other sleep disorders as well, because a reduction in the number of physiologic parameters that are measured could have an impact on the ability to make a correct diagnosis.

The American Sleep Disorders Association has defined four types (levels) of sleep testing: standard polysomnography, which requires the presence of a trained technician, and three variations of unattended studies (table). [1,2] The decision to use one over another depends on many variables, including the suspected diagnosis, the perceived need for technician intervention, patient reliability, and the test's availability, accuracy, and cost.

The aims of testing a patient with suspected sleep disordered breathing are 1) to confirm the diagnosis, 2) to determine the frequency and severity of respiratory events, and 3) to evaluate the physiologic consequences during sleep that occur secondary to the respiratory events. The physiologic variables that are evaluated in a standard polysomnogram include changes in oxygen saturation, cardiac arrhythmias, arousals caused by respiratory events, and sleep fragmentation caused by frequent arousals. The daytime physiologic consequences (daytime sleepiness, memory loss, irritability, psychomotor effects, etc.) are not evaluated by a standard polysomnogram.

Choosing the appropriate test

In order to determine which study to use on a patient, it is important to understand the types of recording devices used in sleep testing. Because sleep laboratories and portable recording devices often use different technology to measure physiologic variables, interpretation of testing results can vary.

Sleep and sleep staging, To monitor sleep, the standard approach is to record several EEG leads (usually at least one central and one occipital lead) that are referenced to a standard ground. The EEG records brain waves (alpha, theta, delta, and saw-tooth waves), sleep spindles, and K complexes. These measurements help define the various stages: wakefulness and stage 1, stage 2, stage 3, stage 4, and REM sleep.

Submental electromyography (EMG) is used to record static and phasic muscle tone. EMG activity is highest when the patient is awake and lowest during REM sleep; it increases during arousals. Therefore, EMG activity helps to differentiate the various sleep stages. Electrooculography (EOG) records eye movement. Identification of blinks, rapid eye movements, and slow movements helps to differentiate wakefulness, REM sleep, and non-REM (NREM) sleep. Sleep stages are scored in 30-second epochs, according to rules established by Rechtschaffen and Kales, [3] by analyzing a combination of EEG, EOG, and chin EMG activity.

One of the other approaches to differentiating sleep from wakefulness is actigraphy. Actigraphy is the recording of an intermittent sampling of wrist motions. Because a typical actigraph can record for several days, this technology has been used mostly in research and for analyzing circadian rhythm disorders. Another approach to differentiating REM from NREM sleep is a computer analysis of EOG rapid eye movements, although the occurrence of frequent rapid eye movements can be mistaken for wakefulness. And, of course, the onset of sleep can also be assessed visually by an attending technician.

Respiratory effort and airflow. An abdomen-chest belt that contains a strain gauge is the standard approach to recording abdominal and chest wall motion. During normal respiration, the abdominal and chest channels are synchronous. During obstructed breathing, there can be a paradoxical motion of the two channels.

Nasal and oral airflow are typically measured indirectly with a thermistor, a thermometer that can measure extremely small changes in temperature. By measuring increases in temperature on exhalation and decreases in temperature on inspiration, thermistor readings determine if respiration is normal, reduced, or absent.

The combination of airflow and abdominal and chest wall motion is usually sufficient to infer the presence or absence of respiratory effort and to differentiate between obstructive, mixed, and central apnea. The difference between an obstructive and central hypopnea, however, is more difficult to determine. Esophageal manometry measures intraesophageal pressure (and therefore intrathoracic pressure) and is the most direct measure of respiratory effort. Because negative intrathoracic pressure is a direct result of respiratory effort, its measurement can infer the degree of upper airway resistance. Patients with obstructive apnea, obstructive hypopnea, or episodes of upper airway resistance demonstrate increasing respiratory effort during these events. Currently, esophageal pressure recording devices are used in clinical studies by only a small number of sleep laboratories in the United States.

There are many other approaches that can measure airflow and respiratory effort. They include the monitoring of snoring, breathing, or tracheal airflow with a stethoscope or microphone. An analysis of breathing sounds can be just as sensitive as a thermistor measurement, but it does not differentiate between inspiration and expiration. A pneumotachometer measures airflow through a mask that is placed over the patient's mouth and nose. The mask is more sensitive than a thermistor, although it is not as comfortable for the patient.

Inductance plethysmography, which measures the amount of airflow and respiratory effort with an abdomen-chest belt, is used in several ambulatory recording devices. Also, piesoelectric recorders (similar to electrocardiographic [EGG] electrodes) can be attached to the skin of the chest to record chest wall motion and respiratory effort.

Airflow and respiratory effort can also be measured by a pressure sensor that is placed in front of the nose or mouth and is attached to either a nasal oxygen cannula or to an oral-nasal facemask (Pressure Transducer Airflow Sensor, Pro-Tech Services, Inc., Woodinville, Wash.; AutoSet Recorder, ResMed Corp., San Diego). The recorder output shows a flattening of the inspiratory pressure curve during episodes of inspiratory airflow limitation. This type of recording is often more sensitive than standard thermistor measurements because it determines upper airway resistance to airflow. Some ambulatory recorders (e.g., the AutoSet Recorder) use a computer analysis of the pressure curve (flattening index) to identify respiratory events and upper airway resistance syndrome.

Respiratory effort and upper airway resistance can be measured indirectly with a piesoelectric sensor that is attached to the skin overlying the suprasternal notch (Suprastemal Esophageal Pressure Sensor, EPM Systems, Midlothian, Va.). Retraction of the soft tissues of the suprastemal notch stretches the sensor and correlates with intraesophageal and intrathoracic pressure. Intercostal EMG has also been shown to correlate with intrathoracic pressure and respiratory effort.

Oximetry. The oximeter probe is usually attached to the patient's finger. The recorded desaturation is delayed following a respiratory event because of the circulation time from the heart to the finger. The normal delay is 10 to 12 seconds. Patients with low cardiac output often have a longer delay. Other factors can affect reported oxygen saturation. For example, oximeters sample data at different rates, so a low sampling frequency can hinder the device's sensitivity. Also, inaccurate calibration can skew measurements.

ECG and heart rate. Heart rate can be recorded by two EGG leads or by the finger oximeter. EGG also enables assessment of cardiac arrhythmias.

Limb movements. EMG leads can be connected to one or both legs at the tibialis anterior to record leg movements. Limb movements during sleep often occur as part of the arousal response to respiratory events. Repetitive limb movements can also represent a separate disorder called periodic limb movements of sleep (PLMS). Limb movements might cause arousals and sleep fragmentation and lead to excessive daytime sleepiness.

Sleep position. Sleep position can be monitored by an attending technician or with a position sensor. Recording of sleep position is important in the management of sleep apnea because apnea that occurs primarily while the patient is supine can be treated with positional therapy.

Snoring. Recording snoring sounds is helpful in identifying apneas and in diagnosing upper airway resistance syndrome. Computer analysis of audiotaped snoring can accurately identify an apnea, [4] but it is less accurate in distinguishing obstructive, mixed, and central events, and it cannot accurately identify a hypopnea. [5]

Data acquisition, storage, retrieval, and analysis. Recorded data must be sampled at adequate rates in order to detect ongoing changes in physiologic parameters. For example, EEG, EMG, and ECG data must be sampled at a much higher frequency (200 Hz) than oxygen saturation, airflow, and abdominal movement (25 Hz).

A digital recording system must have an adequate storage capacity. A typical device might require 50 megabytes of storage for an 8-hour recording. Data are usually downloaded onto a computer system separate from the recording device.

Raw data should be collected and stored in a fashion that allows for accurate reproduction and visual interpretation (similar to analog data from a standard polysomnogram). Data can be stored in a digital format as long as the sampling rates provide a degree of accuracy similar to analog data. Some ambulatory recorders automatically transfer and store data in a formatted file; others allow for the review of raw data. Algorithms for scoring the presence, type, duration, and consequences of respiratory events vary according to the type of recording device used. Algorithms for the analysis of sleep stages attempt to simulate visual scoring, but they vary significantly in accuracy. Standard visual analysis of analog data of sleep stages and respiratory events is performed according to the rules set down by Rechtschaffen and Kales. [3]

Power requirements. Recorders connected to wall outlets must be protected from electrical power artifact (60 Hz) as well as power surges that could damage equipment or data. Battery-powered devices must have adequate power for full acquisition of data.

Artifact and data loss. Artifact occurs during both attended and ambulatory studies. It is important to know how much of a study cannot be evaluated and scored because of artifact. It is also important to delete artifact so that incorrect data are not included in the final report.

Types of testing

Level 1. The level 1 (attended) polysomnogram is the standard of care for the diagnosis of sleep disordered breathing. Its major advantage is the presence of a trained technician, who can replace any leads that become disconnected or nonfunctional. In addition, in the event of severe apnea, serious oxygen desaturation, or a life-threatening cardiac arrhythmia, the technician can intervene and administer continuous positive airway pressure (CPAP) or supplemental oxygen. Attended monitoring also allows visual confirmation of many of the sleep variables and observation of other sleep disorders (e.g., seizure activity, parasomnia).

Measurement of respiratory variables and oxygen saturation allows for the identification of the frequency, type, length, and severity of respiratory events and desaturation. Recording EEG, EGG, and chin EMG permits identification of respiratory, PLMS, and spontaneous arousals that can lead to sleep fragmentation. Patients with sleep apnea only during REM sleep can also be identified. Assessment of the EEG can identify alpha intrusion, which is associated with nonrestorative sleep and excessive daytime sleepiness. EEG can also identify seizure disorders and parasomnia.

A visual identification or measurement of sleep position allows for a diagnosis of positional apnea. The identification of cardiac arrhythmias or apnea-related heart rate variability is made possible by the EGG. The use of the leg EMG is required for the diagnosis of periodic limb movements and restless legs syndrome. The measurement of sleep latency, REM latency, and sleep efficiency helps identify other sleep disorders, including insomnia, chronic insufficient sleep, and narcolepsy.

Level 2. The level 2 polysomnogram measures the same variables as a level 1 study, but it is performed in the patient's home. Most investigators report a higher incidence of data loss during an unattended study than during an attended study.2 However, one recent study demonstrated that with the proper selection of patients and recording equipment and with careful instruction, the failure rate of an ambulatory level 2 test was similar to that of an attended study.6

Although the quality can be similar, it has not yet been determined if a level 2 study is less expensive than a level 1 study. In theory, cost savings should be realized because there is no need for a nighttime technician or sleep room. On the other hand, overall costs can increase because of a need for more equipment, damage to or loss of equipment, additional technician time for patient instruction and return of equipment, and repetition of failed studies. In addition, it has not yet been determined if CPAP titration can be effectively performed in the home with a level 2 test.

Level 3. In selected patients, a limited ambulatory study can be a simple and effective method for the diagnosis of sleep disordered breathing. A level 3 study has the same limitations as a level 2 study in terms of displaced leads and data loss. Moreover, because a level 3 study tests only respiratory variables, it is not possible to determine actual sleep time, detect arousals, assess sleep fragmentation, identify cardiac arrhythmias, or diagnose other sleep disorders. One consequence of these drawbacks is that there is the potential to underestimate the apnea-hypopnea index, because the number of respiratory events would be divided by the recording time rather than the total sleep time.

Despite these limitations, some studies that compared level 3 devices with standard polysomnography indicated that when they are applied to patients with presumed sleep apnea, many of these devices appear to be highly sensitive and specific for the diagnosis of sleep disordered breathing. [7-9] The level 3 study might be most appropriate for postoperative followup or for an initial diagnosis in patients suspected of having moderate or severe sleep apnea.

Level 4. A level 4 study is an unattended ambulatory recording of only one or two physiologic parameters. It is not without its disadvantages. For example, recording only tracheal breath sounds has been shown to overestimate the frequency of apneic episodes and underestimate total apnea time; it also has difficulty detecting hypopnea. [10] Other methods of recording snore sounds have been marketed for the diagnosis of sleep apnea, but validation studies on them have yet to be published.

Oximetry has been compared with standard polysomnography in numerous studies. There are significant variations in its sensitivity (40-100%) and specificity (36-100%), depending on the respiratory disturbance index (RDI) level used to define sleep apnea, the degree of desaturation considered to be abnormal, the addition of visual analysis of the oximeter tracing, and the response characteristics of individual oximeters. [2,11] general, oximetry is often useful for those patients in whom there is a strong clinical suspicion of moderate or severe sleep disordered breathing. The disadvantages of oximetry are that healthy young patients without desaturation can be misdiagnosed as being normal, while those with chronic obstructive pulmonary disease can be misdiagnosed as abnormal.

In patients with sleep disordered breathing, variations in cyclical heart rate occur in response to the arousal stimulus at the termination of respiratory events. Isolated recordings, however, are neither sensitive nor specific enough to be useful in the diagnosis of sleep apnea.

To date, oximetry is the only level 4 test that has undergone sufficient study to determine its usefulness. Oximetry in a patient with a pretest clinical suspicion of sleep disordered breathing can be useful in selecting patients for further assessment with a level 1 or 2 study.

Split-night studies. A split-night study is a level 1 diagnostic polysomnogram that is performed on the same night as a CPAP titration study. The diagnostic portion of the study can last 1 to 3 hours, which is sufficient time to confirm the presence and severity of sleep apnea. CPAP titration is then initiated and continued until optimal pressure is identified. The major advantage of a split-night study is a presumed decrease in the overall cost of diagnosis, because the two tests are administered in one night rather than two.

Among the disadvantages of split-night studies are that they are not as accurate in patients with mild sleep apnea. [12] Also, one study has shown that patients are less willing to accept CPAP therapy during a split-night study than during a full-night evaluation (62.5 vs. 72-92%). [13] Furthermore, compliance rates 6 weeks after titration are lower among those patients who undergo a split-night study rather than a full night of CPAP titration.

In conclusion, a variety of testing options is available for the diagnosis of sleep disordered breathing. To date, standard attended polysomnography is still the gold standard. Many ambulatory level 2 and 3 testing devices have been validated for use in the home, and they are suitable for selected patients. An understanding of the methods of data acquisition and analysis is important for understanding the limitations of a particular device. Only with a full knowledge of the advantages and limitations of a recording technique can an appropriate type of test be ordered.

Abstract

Rapid advancements in technology have expanded the range of available methods for testing patients with sleep disordered breathing. This article reviews the components of a sleep test and the currently available technologies that are used in standard attended, laboratory (level 1) studies. This paper also reviews the advantages and limitations of unattended at-home evaluations (level 2, 3, and 4 studies) for patients with suspected sleep disordered breathing.

From the Atlanta Ear, Nose, and Throat Associates.

References

(1.) Practice parameters for the use of portable recording in the assessment of obstructive sleep apnea: Standards of Practice Committee of the American Sleep Disorders Association. Sleep 1994;17:372-7.

(2.) Ferber R, Millman R, Coppola M, et at. Portable recording in the assessment of obstructive sleep apnea. ASDA standards of practice. Sleep 1994;17:378-92.

(3.) Rechtschaffen A, Kales A, eds. A manual of standard terminology techniques and scoring systems for steep stages of human subjects. Bethesda, Md,: National Institutes of Health, 1968.

(4.) Van Brunt DL, Lichstein KL, Noe SL, et al. Intensity pattern of snoring sounds as a predictor for steep-disordered breathing. Sleep 1997;20:1151-6.

(5.) Rosenthal L, Mickelson SA, Day R, et at. The diagnosis of OSA based on the pattern of snoring and nocturnal oximetry. Presentation at the 11th annual meeting of the Associated Professional Steep Societies, San Francisco, June 11, 1997.

(6.) Fry JM, DePhillipo MA, Curran K, et al. Full polysomnography in the home. Steep 1998;21:635-42.

(7.) Stoohs R, Guilleminault C. MESAM 4: An ambulatory device for the detection of patients at risk for obstructive steep apnea syndrome (OSAS). Chest 1992;101:1221-7.

(8.) Redline 5, Tosteson T, Boucher MA, Millman RP. Measurement of sleep-related breathing disturbances in epidemiologic studies: Assessment of the validity and reproducibility of a portable monitoring device. Chest 1991;100:1281-6.

(9.) Gyulay 5, Gould D, Sawyer B, et at. Evaluation of a microprocessor-based portable home monitoring system to measure breathing during steep. Steep 1987;10:130-42.

(10.) Cummiskey J, Williams TC, Krumpe PE, Guilleminault C. The detection and quantification of steep apnea by tracheal sound recordings. Am Rev Respir Dis 1982;126:221-4.

(11.) Gyulay 5, Olson LG, Hensley MJ, et al. A comparison of clinical assessment and home oximetry in the diagnosis of obstructive sleep apnea. Am Rev Respir Dis 1993;147:50-3.

(12.) Yamashiro Y, Kryger MH. CPAP titration for sleep apnea using a split-night protocol. Chest 1995;107:62-6.

(13.) Strollo PJ Jr., Sanders MH, Costantino JP, et al. Split-night studies for the diagnosis and treatment of steep-disordered breathing. Sleep 1996;19:S255-9.

Types of studies for evaluating sleep apnea (minimum of 6 hours of
overnight recording)

                     Level 1                     Level 2
No. parameters       [greater than or equal to]7 [greater than or equal to]7
 measured
EEG [*]              C4-A1                       C4-A1
                     or C3-A2                    or C3-A2
EOG [+]              Required                    Required
Chin EMG [++]        Required                    Required
ECG, [ss] heart rate ECG required                ECG or HR
Airflow              Required                    Required


Respiratory          Required                    Required
 effort

[O.sub.2] saturation Required                    Required
Body position        Visual or                   Measured
                      measured                    (optional)
Leg movement         Optional                    Optional
Personnel            Present                     Not present
Intervention         Possible                    Not possible


                     Level 3                     Level 4
No. parameters       [greater than or equal to]4 [greater than or equal to]1
 measured
EEG [*]              Optional                    Not measured

EOG [+]              Optional                    Not measured
Chin EMG [++]        Optional                    Not measured
ECG, [ss] heart rate ECG or HR                   Optional
Airflow              Either two channels         Optional
                     of effort or one
                     channel plus airflow
Respiratory          Either two channels         Optional
 effort              of effort or one
                     channel plus airflow
[O.sub.2] saturation Required                    Optional
Body position        Measured                    Not measured
                      (optional)
Leg movement         Optional                    Not measured
Personnel            Not present                 Not present
Intervention         Not possible                Not possible


(*.)Electroencephalography; (+.)Electro-oculography;
(++.)Electromyography; (ss.)Electrocardiography.