Sleep Medicine Pearls 2nd

Answers: Apnea is present. The CO2 tracing fluctuates due to persistent exhalation (small expiratory puffs) during obstructive (inspiratory) apnea. Discussion: Monitoring of airflow during sleep studies can be performed by several methods, the most common of which features temperaturesensitive devices placed near the nose and mouth. Airflow causes a change in the temperature of the devices, which results in a change in voltage (thermocouple) or resistance (thermistor) of the transducer. The alteration in the signal originating from these transducers when appropriately amplified causes a deflection in the airflow tracing. Although convenient, temperature-sensitive devices do not always accurately reflect the magnitude of change in airflow and sometimes can be misleading. During apnea, the tracing may continue to fluctuate secondary to changes in temperature from contact with the patient's body or room air. Pneumotachographs accurately measure flow and can provide information about the shape of the airflow profile. However, they require masks covering the nose and mouth and therefore are less comfortable. Measurement of the pressure difference across a known resistance (the pneumotachograph wire screen) when calibrated reflects the flow rate (flow = pressure / resistance). Pneumotachographs also can reveal detail not seen in thermocouple monitoring, such as vibration in the flow (snoring), a flat profile (airflow limitation), or expiratory puffs of air during inspiratory apnea. Another method for monitoring airflow is to measure exhaled CO2, As expired air is rich in CO2 while ambient air has essentially none, respiration is detected by fluctuations in measured CO2, Air is sampled via a small nasal or nasal-oral cannula by continuous suction and is measured downstream at an analyzer. The increases in CO2 values (reflecting exhalation) are time-delayed because there is a finite time for the sampled air to reach the analyzer. The plateau value of measured CO2 (end tidal PC02) can provide an estimate of arterial PC02 and is increased during periods of hypoventilation. Apnea is detected by an absence of deflection in the CO2 tracing. This method of monitoring airflow is widely used in pediatric sleep monitoring because children with "sleep apnea" frequently have long periods of hypoventilation (increased end tidal PC02) rather than discrete apneas or hypopneas. However, the method can be misleading. During apnea (absence of inspiratory airflow), small expiratory puffs may continue. These small exhalations are rich in CO2 and can cause significant deflections in the CO2 signal, giving a false impression about the nature of airflow (C02 not airflow, is measured). . Recently, measurement of nasal pressure has been used to detect airflow. Pressure just inside the nasal inlet is measured by connecting nasal cannulas (oxygen- or CO2-monitoring) directly to sensitive pressure transducers. The pressure deflections are reasonable, semiquantitative estimates of the magnitude and shape of airflow. This method works similarly to a pneumotachograph by measuring the pressure drop across the resistance of the nasal inlet. The major difficulty with this method is that oral airflow is not detected (see Fundamentals 10). Simultaneous use of an oral thermocouple can solve this problem. Of note, nasal cannula are available that allow simultaneous measurement of end tidal CO2 and nasal pressure. One prong samples nasal pressure (connected to pressure transducer), and one prong samples exhaled CO2, The RIPsum signal (also known as Vsum) of respiratory inductance plethysmography can provide a semiquantitative estimate of tidal volume (see figure below). In this method, changes in the inductance of coils in bands around the rib cage (RC) and abdomen (AB) during respiratory movement are translated into voltage signals. The sum of the two signals (RIPsum = [a X RC] + [b X AB]) can be calibrated by choosing appropriate constants: a and b. During obstructive apnea, the two signals cancel (a * RC = -b * AB), and RIPsum is close to zero. inspiration RIPsum RC (rib cage) AS (abdomen) Respiratory inductance plethysmography can also be used to detect hypopneas. Central hyponeas are characterized by a reduction in all three signals (RC, AB, RIPsum) without evidence of paradox. Obstructive hypopneas are characterized by a reduction in the RIPsum and usually one or both of the RC and AB signals. In addition, the RC and AB signals often show paradox (move in opposite directions) during the hypopnea. In the present case, simultaneous monitoring with a pneumotachograph (see following figure) showed that the fluctuations in CO2 during apnea were secondary to small expiratory puffs (rich in CO2) during inspiratory apnea. While you could argue that this is not absolute apnea (no airflow), most would agree that very low flow (or no inspiratory flow) and apnea are equivalent physiologically. In fact, many clinicians characterize flow reduced to less than 10-25% of baseline as an apnea. 89

i Increasing Exhaled PC02 Pneumotach Thermocouple Clinical Pearls I. Temperature or CO) measuring devices for monitoring air flow can be inaccurate because they do not reflect the magnitude of airflow. 2. Pneumotachography accurately measures flow and provides a detailed airflow profile that may include information not seen in thermistor/thermocouple monitoring. However, a mask over the nose and mouth must be used. 3. More practical options for monitoring airflow include nasal cannulas connected to pressure transducers (nasal pressure) and respiratory inductance plethysmography. REFERENCES I. Chada TS. Watson H. Birch S. et al: Validation of respiratory inductance plethysmography using different calibration procedures. Am Rev Respir Dis 1982; 125:644-649. 2. Tobin M. Cohn MA. Sackner MA: Breathing abnormalities during sleep. Arch Intern Med 1983; 143: 1221-1228. 3. Monserrat JP, Farre R, Ballester E, et al: Evaluation of nasal prongs for estimating nasal flow. Am J Respir Crit Care Med 1997; 155:211-215. 4. Norman RG. Ahmed MM. Walsleben JA, et al: Detection of respiratory events during NPSG: Nasal cannula/pressure sensor versus thermistor. Sleep 1997; 20:1175-1184. 5. Kryger MH: Monitoring respiratory and cardiac function. In Kryger MH. Roth T, Dement WC (eds): Principles and Practice of Sleep Medicine, 3rd ed. Philadelphia. WB Saunders, 2000, pp 1217-1230. 90

PATIENT 27 A 50-year-old man with possible central apnea A 50-year-old, obese man was evaluated at another hospital for complaints of excessive daytime sleepiness. He underwent a sleep study in which many apneas were noted. Because of minimal movement in the chest and abdominal bands, the apnea was labeled as central by the technician scoring the study. A second sleep study was performed on presentation to this sleep laboratory because the patient's history suggested obstructive apnea. Figure: A sample tracing of one of the recorded events is shown below. Question: What type of apnea is present? Airflow Chest Abdomen 1 sec J-I 91

Diagnosis: Obstructive apnea. There is minimal chest and abdominal movement due to obesity. The most sensitive method of detecting respiratory effort is to measure the esophageal pressure. Changes in esophageal pressure are estimates of pleural pressure changes. In the past, this method required esophageal balloons, which were stiff and uncomfortable. Recently, esophageal pressure has been measured using small, soft, fluid-filled catheters (pediatric feeding tubes) connected to pressure transducers, such as the disposable transducers commonly used in intensive care units. This technique is usually well tolerated. New esophageal balloon catheters are also now available with a removable stylet. These are more flexible and comfortable for the patient than the older balloon catheters. In one study using both RIP and esophageal pressure monitoring, apneas were found to have been correctly classified by RIP alone in 91% of patients. Thus, monitoring chest and abdominal movement is satisfactory for most patients. In a few patients, obstructive apneas may be incorrectly labeled as central if esophageal pressure monitoring is not performed. Of note, RIP monitoring of chest and abodomen is probably more sensitive than piezo-electric belts at detecting respiratory effort. With piezo-electric belts the signal is dependent on the tension on the transducer. This mayor may not be proportionate to changes in the area surrounded by the belt. In addition, if the belts are not sufficently tight or move during the night, the signal can be very low. In all types ofsurface monitoring, correct placement ofthe belts is essential to obtain a good signal. In the present case, definite chest and abdominal movements are evident, although of a small magnitude. The simultaneous esophageal pressure trace (see figure below) reveals that inspiratory effort clearly is present and increasing during the event. Therefore, the event is an obstructive apnea. Esophageal Pressure Abdomen 1 sec .... Chest Airflow Some of the many different methods of detecting movement include piezo-electric transducers in bands, mercury strain gauges, and respiratory impedance plethysmography (RIP). RIP converts changes in the impedance of a coil in a band around the body secondary to changes in the enclosed area during chest/abdominal excursions into a voltage signal. The rib cage (RC) and abdominal (AB) signals are then added (RIPsum = [a X RC] + [b X AB)). If the coefficients a and b are selected by calibration, RIPsum is a reasonable estimate oftidal volume (see Patient 26). During obstructive apnea, the rib cage and abdominal contributions to RIPsum cancel (a X RC = - b X AB), and RIPsum is close to zero (apnea). In all of the above methods, a change in body position may alter the ability to detect chest/abdominal movement. This may require adjusting band placement or amplifier sensitivity. In addition, very obese patients may show little chest/abdominal wall movement despite considerable inspiratory effort. Thus, be cautious about making the diagnosis of central apnea solely on the basis of surface detection of inspiration effort. Discussion: The diagnosis of obstructive apnea depends on demonstration of apnea despite continued inspiratory effort. This usually is accomplished by detecting chest and abdominal movement. During obstructive apnea/hypopnea, the chest and abdominal tracings may show paradoxical motion (one in, the other out). Sometimes changes in phase between chest and abdomen are more subtle. In the following tracings from a patient with obstructive apnea, paradox is not present in the initial part (a) but is obvious in the last part of the tracing (b). Paradoxical movement tends to be most pronounced in REM sleep, when there is hypotonia of the chest wall muscles. 92

Clinical Pearls I. Chest and abdominal movements during obstructive apnea may be small in magnitude and difficult to detect in some obese individuals. 2. The most sensitive method of detecting respiratory effort is to monitor the esophageal pressure. 3. During obstructive apnea and hypopnea, the chest and abdomen movements may be out-of-phase, demonstrating a subtle difference rather than an obvious paradox. 4. If apneas are classified solely by chest and abdominal bands, some events may be incorrectly classified as central. REFERENCES 1. Chervin RD and Aldrich MS. Effects of esophageal pressure monitoring on sleep architecture. Am J Respir Crit Care Med 1997;56:881-885. 2. Staats BA. Bonekat HW. Harris CD. et al: Chest wall motion in sleep apnea. Am Rev Respir Dis 1984; 130:59-63. 3. Flemale A. Gillard C, Dierckx JP: Comparison of central venous. esophageal and mouth occlusion pressure with water-filled catheters for estimating pleural pressure changes in healthy adults. Eur Resp J 1988; 1:51-57. 4. Kryger MH: Monitoring respiratory and cardiac function. In Kryger MH. Roth T. Dement WC (eds): Principles and Practice of Sleep Medicine, 3rd ed. Philadelphia, WB Saunders. 2000. pp 1217-1230. 5. Berry RB: Nasal and esophageal pressure monitoring. In Lee-Chiong TL. Sateia MJ, Caraskadon MA (eds): Sleep Medicine. Philadelphia, Hanley and Belfus, 2002. pp 661-672. 93

PATIENT 28 A 30-year-old man with heavy snoring A 30-year-old man was referred for evaluation of heavy snoring. He denied excessive daytime sleepiness. The patient's wife reported that he stopped breathing during the night. Physical Examination: Blood pressure 150/85, pulse 88/min. HEENT: edematous palate. Neck: 18-inch circumference. Chest: clear. Cardiac: normal. Extremities: no edema. Sleep Study: Total sleep time: 350 minutes (normal 400--443). Total apneas: 200 (5% central, 30% mixed, 65% obstructive). Total hypopneas: 50. Figure: A typical apnea is illustrated below. Chest and abdomen tracings show movements of these areas (respiratory effort). Questions: Which type of apnea is illustrated? What is the diagnosis? Airflow Chest Abdomen Arterial Oxygen Saturation 94

Answers: The illustrated event is a mixed apnea. The diagnosis is obstructive sleep apnea. Discussion: An obstructive apnea is one in which ventilatory effort is present. A mixed apnea is composed of an initial central apnea (no inspiratory effort) followed by an obstructive portion. Most patients with obstructive sleep apnea have predominantly mixed or obstructive apneas. However, the presence of a small fraction of central events is not unusual. Both mixed and obstructive apneas have the same clinical significance and usually the same etiology (upper airway obstruction). The initial, central portion of a mixed apnea is believed to be due to the hyperventilation following the preceding apnea. As the patient returns to sleep, the PCOz is below the apneic threshold (the level of COz that triggers ventilation). The result is an absence of inspiratory effort (central apnea). During the apnea, the PCOz rises until inspiratory effort returns. However, apnea persists despite the return of inspiratory effort secondary to an obstructed upper airway. Adequate treatment of upper airway obstruction prevents apnea and postapnea hyperventilation. Thus, both the central and obstructive components of mixed apneas are eliminated by effective treatments such as tracheostomy or nasal continuous positive airway pressure (CPAP). Some central apneas may persist, but usually resolve with time. In the current case, the illustrated event is a mixed event with an initial central portion (A) followed by an obstructive portion. Paradoxical motion is seen in the chest and abdomen tracings during the obstructive portion (B). The patient was treated with nasal CPAP which virtually eliminated mixed and obstructive apnea (AHI < 5/hr). A few central apneas and hypopneas persisted in REM sleep. Clinical Pearls I. Many patients with the obstructive sleep apnea syndrome have both mixed and obstructive apneas. The underlying pathogenesis of both is upper airway obstruction. 2. In most cases of obstructive sleep apnea, adequate treatment of upper airway obstruction eliminates both obstructive and mixed apnea. REFERENCES l. Sanders MH: Nasal CPAP effect on patterns of sleep apnea. Chest 1984; 86:839-844. 2. Dempsey lA, Skatrud 18: A sleep-induced apneic threshold and its consequences. Am Rev Respir Dis 1986; 133:1163-1170. 3. Iber C, Davies SF, Champan RC, Mahowald MW: A possible mechanism for mixed apnea in obstructive sleep apnea. Chest 1986; 89:800-805. 95

PATIENT 29 A 33-year-old man complaining of daytime sleepiness A 33-year-old man was evaluated for daytime sleepiness of 3-year duration. His wife reported that he snored heavily and stopped breathing when sleeping on his back. Physical Examination: Normal except for mild obesity (140% of ideal body weight) and a long, edematous palate and uvula. Sleep Study: No apneas were noted; however, 400 events such as the one illustrated below were recorded. Airflow was monitored with a nasal-oral thermister and nasal pressure sensor. Respiratory effort was detected by recording piezo-electric belts around the chest and abdomen. Questions: What is the diagnosis? What is the illustrated event? 30.6 sec II 93% 97% ---:-:---------------------- $a02 nasal pressure airflow [therm) chest abdomen 96

Diagnosis: Obstructive sleep apnea (hypopnea) syndrome. The event is an obstructive hypopnea. Discussion: While the definition of apnea is well standardized, the definition of hypopnea varies among clinicians. All agree that hypopnea is a reduction in airflow (or tidal volume) from the preceding baseline for 10 seconds or longer. As discussed in Fundamentals of Sleep Medicine 10, some definitions require an associated arousal or desaturation (drop in the SaO? of 2-4%). The choice of definitions and the type of sensor used to monitor airflow can dramatically change the number of hypopneas that are detected. Nasal pressure monitors usually detect more hypopneas than devices measuring changes in temperature. Obstructive hypopneas occur because of upper airway narrowing and have the same consequences as obstructive apneas; they disturb sleep (arousals) and often result in variable drops in the SaO? The separation of severe hypopneas from apneas -is not precise. Whether an event appears as an apnea or hypopnea may depend on the amplifier gain and the method used to detect flow. Some sleep labs consider a reduction in flow to below 25% of baseline as an apnea. The use of nasal pressure to monitor flow tends to identify more events as apneas. This could be because oral flow (inspiratory and expiratory mouth breathing or oral expiratory puffs) may fail to cause a deflection in the nasal pressure signal, or because nasal pressure tends to underestimate flow when flow rates are low (see Fundamentals 10). The figure below shows an event that appears as an apnea in the nasal pressure signal, although deflections in the thermistor tracing are noted. airflow- ~JlnA!1n~ _~~~~nAn!1nJ\ thermistor VVV V v - - v VVVVV V' chest nasal I~ pressure iI The nasal pressure signal should be amplified or recorded as either a DC signal or an AC signal with a long time constant or a very small low filter setting (~ 0.0 I). If this is not done, the characteristic airflow flattening will be less apparent. An AC amplifier with a short time constant sees no change in flow (flow plateau) as zero flow. The figure below shows the same signal recorded with a short time constant (1.6 sec), a long time constant (5.0 sec), and as a DC signal. u § s; o ?: 0 LL - U o ?: 0 LL U - 0 ?: 0 LL 0 5 10 15 20 Time (sec) 25 30 Dotted line = zero flow 97

When using respiratory inductance plethysmography, hypopneas are characterized by a drop in RIPsum (an estimate of tidal volume) and usually a drop in either the chest or abdominal deflections (or both). Typically during obstructive hypopnea there may be chest-abdomen paradox (one moving in while the other moves out). However, with piezo-electric sensor bands, the changes in chest and abdominal movements are more variable. Paradox may not be detected. If a snore detector is employed, you may also see evidence of snoring during obstructive hypopneas. In the present patient, the nasal pressure signal (see figure on previous page) shows a clear decrease in magnitude and a flattened shape during the event. After event termination, the shape of the nasal pressure becomes rounded again. In contrast, there is little change in airflow by thermistor or chest/abdominal deflections. There is chest and abdominal paradox, but this is present both during and after event resolution. The obstructive hypopnea is followed by a 4% desaturation. Clinical Pearls I. The separation of events into apneas and hypopneas is imprecise and may depend on the sensitivity/calibration of the airflow recording system. Thus the apnea-hypopnea index is a more accurate (and inclusive) estimate of the severity of sleep-disordered breathing than the apnea index. 2. Paradoxical motion of the chest or abdomen during hypopnea suggests an obstructive hypopnea. However, this finding may be absent (especially if piezo-electric belts are used for monitoring chest/abdominal movement). 3. Nasal pressure monitoring detects more hypopneas than thermistors. The flattened shape of the nasal pressure signal also helps identify hypopneas as obstructive. It is important to record nasal pressure as a DC signal-or use the approprite filter settings if recorded as an AC signal. 4. Some events may appear as apneas in the nasal pressure signal and hypopneas in the thermistor signal. 5. Identification of hypopneas is important because these events may result in arterial oxygen desaturation and arousal from sleep. REFERENCES I. Block Al, Boysen PG. Wynne Wl. et al: Sleep apnea. hypopnea. and oxygen desaturation in normal subjects: A strong male predominance. N Engl 1 Med 1979; 300:513-517. 2. Gould GA, KF Whyte. GB Rhind, et al: The sleep hypopnea syndrome. Am Rev Respir Dis 1988; 137:895-898. 3. Berg S, Haight lSl, Yap V. Hollstein V. Cole P: Comparison of direct and indirect measurements of respiratory airflow: Implications for hypopneas. Sleep 1997;20:60-64. 4. Norman RG. Ahmed MM. Walsleben lA. Rapoport OM: Detection of respiratory events during NPSG: Nasal cannula/pressure sensor versus thermistor. Sleep 1997;20: 1175-1184. 98

FUNDAMENTALS OF SLEEP MEDICINE 11 Excessive Daytime Sleepiness Excessive daytime sleepiness (EDS) is the most common complaint evaluated by sleep disorder specialists. Sleep apnea is the most common cause. However, do not automatically assume that every patient with EDS and snoring has sleep apnea. All of the causes of daytime sleepiness listed below must be carefully considered. Note that it is common for more than one of these disorders to be present in a given individual. Become acquainted with the patient's medical history, and explore a relevant review of symptoms. For example, congestive heart failure is commonly associated with a type of central sleep apnea (Patient 69), and patients with renal failure often have periodic limb movements in sleep (PLMS). Hypothyroidism and acromegaly are predisposing conditions for sleep apnea. In addition, some medications can cause daytime sleepiness or fatigue. Patients with some of the disorders listed below may present with complaints of insomnia (difficulty initiating and maintaining sleep) rather than EDS. In fact, patients with PLMS more commonly present with insomnia complaints than EDS. Problems with insomnia also may predominate in those with depression. Even patients with sleep apnea (for which EDS is a cardinal manifestation) may seek medical evaluation primarily because of frequent nocturnal awakenings rather than daytime sleepiness. Evaluating Causes ofExcessive Daytime Sleepiness DISORDERS Ev ALUATION Sleep apnea syndromes Upper airway resistance syndrome Narcolepsy Depression Periodic leg (limb) movements in sleep Idiopathic hypersomnia Withdrawal from stimulants Insufficient sleep syndrome Drug dependence/abuse Medication side effects Post-traumatic hypersomnia Brain tumors All cases History Self-rating scales of sleepiness Sleep-wake diary Polysomnography Selected cases MSLT (narcolepsy) Drug screen A good history is essential in evaluating patients with EDS. Differentiating complaints of fatigue and daytime sleepiness can be difficult. Quantifying the degree of daytime sleepiness is challenging because patients tend to underestimate. Questionnaires such as the Epworth Sleepiness Scale or Stanford Sleepiness Scale are attempts to standardize the evaluation of self-rated symptoms of sleepiness. The Stanford Scale measures subjective feelings of sleepiness ("fogginess, beginning to lose interest in staying awake"). In contrast, the Epworth Scale measures average sleep propensity (chance of dozing) over eight common situations that almost everyone encounters. The test was developed by Johns at the Epworth Hospital in Melbourne, Australia. It has gained popularity because it is simple and short. The propensity to fall asleep is rated as 0 (never), 1,2, or 3 (high chance of dozing; see table). The maximum score is 24, and normal is assumed to be 10 or less. The Epworth Sleepiness Scale correlates roughly with the severity of ob99

structive sleep apnea and improves after CPAP treatment. The Scale has a low but significant correlation with the MSLT (an objective measure of sleepiness). However, changes in the Epworth Scale after treatment may not always correlate with changes in the MSLT. Epworth Sleepiness Scale SITUATION - "USUAL WAY OF LIFE IN RECENT TIMES" Sitting and reading Watching TV Sitting, inactive in a public place (e.g., a theater or a meeting) As a passenger in a car for an hour without a break Lying down to rest in the afternoon when circumstances permit Sitting talking to someone Sitting quietly after a lunch without alcohol In a car, while stopped for a few minutes in traffic Total: CHANCE OF DOZING (SCORED 0, I, 2, 3 ) 0-3 0-3 0-3 0-3 0-3 0-3 0-3 0-3 0-24 (0-10 normal) o = would NEVER doze. I = SLIGHT chance of dozing, 2 = MODERATE chance of dozing, 3 = HIGH chance of dozing Question the patient about which activities are compromised by decreased alertness (e.g., driving, work, social situations). Record the normal bedtime, waketime, and average hours of sleep. Surprisingly, the simple answer for some patients is that they are trying to exist on an inadequate amount of sleep. Many sleep disorders centers have patients fill out a sleep-wake diary for 2 weeks before being evaluated. This diary documents patterns of sleep and daytime sleepiness. Questioning bed partners is absolutely essential in evaluating patients with EDS. A history of loud snoring and observed gasping or apnea is suggestive of obstructive sleep apnea syndrome, whereas a history of leg jerks or kicking suggests periodic limb movements in sleep. The age at onset of symptoms provides a clue to the disorder. While sleep apnea can start at any age, it typically presents in middle-aged or older individuals. In contrast, narcolepsy usually starts in late adolescence or the 20s. The patient should be questioned about cataplexy (loss of muscle tone during moments of increased emotion such as laughter), which is characteristic of narcolepsy. Sleep paralysis (inability to move while still awake, at sleep onset, or after awakening) and hypnagogic hallucinations (vivid sensory imagery, usually visual, occurring at sleep onset while still awake) also are common in narcolepsy, but can occur in normal individuals as well. Symptoms of depression suggest that this common disorder is the cause of daytime sleepiness. Physical examination should pay special attention to the blood pressure, upper airway (nose, mouth, and throat), neck circumference, and signs of right or left heart failure or hypothyroidism. History in Excessive Daytime Sleepiness Age of onset Duration of symptoms Daily activities impaired (e.g., driving, work, social situations) Medications, ethanol, sleeping pill use Sleep habits: bedtime, duration of sleep Bed partner observations: snoring, gasping, apnea, leg kicks Symptoms of narcolepsy: cataplexy, sleep paralysis, hypnagogic hallucinations Symptoms of depression In addition to the history and physical examination, a nocturnal sleep study (polysomnography) is required for most patients presenting with EDS. The severity of the disorder frequently is underestimated by the patient. Moreover, several disorders may be present in the same individual (e.g., narcolepsy, obstructive sleep apnea, and periodic limb movements in sleep). If a diagnosis of narcolepsy is suspected, an MSLT following the nocturnal sleep study can be useful. The MSLT provides objective evidence of the tendency to fall asleep during the day and can help make the diagnosis of narcolepsy (sleep-onset REM). 100

REFERENCES I. Johns MW: Sleepiness in different situations measured by the Epworth Sleepiness Scale. Sleep 1994; 17:703-710. 2. Johns MW. Daytime sleepiness, snoring, and obstructive sleep apnea. The Epworth Sleepiness Scale. Chest 1993: 103:30-36. 3. American Sleep Disorders Association: International Classification of Sleep Disorders: Diagnostic and Coding Manual. Rochester, Minnesota, ASDA, 1997. 4. Mitler MA, Carskadon MA, Hirshkowitz M: Evaluating sleepiness. In Kryger MH, Roth T, Dement WC (eds): Principles and Practice of Sleep Medicine, 3rd ed. Philadelphia, WB Saunders, 2000, pp 1251-1257. 101

PATIENT 30 A 45-year-old man with a snoring problem A 45-year-old man sought treatment of his snoring, which had been present for many years. His wife slept in another bedroom because the snoring "shook the walls." The patient denied symptoms of excessive daytime sleepiness (Epworth Sleepiness Scale score 6/24 [normal]), morning headache, or problems at work (he was a successful accountant). He did admit to drinking more than five cups of coffee daily. There was no history of recent weight gain or alcohol use. Physical Examination: Height 5 feet 10 inches, weight 190 pounds, blood pressure 150/90. Neck: 18-inch circumference, no jugular venous distention. HEENT: edematous uvula, dependent palate (see figure). Chest: clear. Cardiac: normal. Extremities: no edema. Figure: Compare the typical adult with the individual presenting in this case. Question: What missing bit of historical information is essential to determining whether or not a sleep study should be performed to rule out obstructive sleep apnea? 102 normal patient Dependent palate Long edematous uvula High tongue base

Answer: Question the spouse about observed apnea/gasping during sleep. Discussion: Obstructive sleep apnea (OSA) is a common disorder occurring in about 4% of men and 2% of women. During sleep, closure of the upper airway results in cessation of airflow despite continued respiratory effort. The termination of apnea is associated with a brief awakening. The resulting sleep fragmentation reduces the amount of slow wave and REM sleep and causes varying degrees of daytime sleepiness. The presence and severity of OSA can be precisely defined by sleep monitoring. However, because attended in-lab monitoring is expensive and of limited availability, appropriate screening techniques are important. In a study examining the predictive value of a number of historical and physical findings-neck circumference, hypertension, habitual snoring, and bed partner reports of gasping/ choking - respirations were found to be the best predictors. Body weight, recent weight gain, and older age also were significant factors. Interestingly, the classic daytime symptoms said to be present in sleep apnea (daytime sleepiness, morning headaches, and cognitive impairment) were not predictive of the disorder. Unfortunately, many patients who deny symptoms of daytime sleepiness do have significant sleep apnea. Structured questionnaires such as the Berlin Questionnaire have also proved useful in identifying patients likely to have OSA. Once a patient has been selected, the physician decides on the appropriate diagnostic study (see table). Levell-attended, full polysomnographyis the gold standard and the only study recommended by the American Academy of Sleep Medicine under usual circumstances. Level 2 - full, unattended (ambulatory) polysomnography-is now available and has been used successfully. However, significant amounts of data are lost (i.e., leads fall off) in 10-15% of these ambulatory studies. Level 3 (cardiopulmonary studies) is recommended when there is a high pre-test probability of OSA, and traditional polysomnography is not available locally or the delay in obtaining a study in not acceptable. Level 3 might also be useful in patients who cannot be moved to the sleep laboratory or for follow-up after treatment. Level 4 studies consist of monitoring of one or two bioparameters (usually oximetry with our without snoring or airflow), but these have an appreciable false negative rate. The other factor to consider is that if a positive-pressure titration is ultimately needed, the combination of a Level 3-4 study plus a traditional in-lab CPAP titration is no cheaper than a single partial-night sleep study (initial part is diagnostic, second part is positive-pressure titration). An additional issue is reimbursement. Currently, unattended studies are poorly reimbursed in the U.S. How the data from portable monitoring is analyzed is as important as which data is recorded. Full disclosure (a second-by-second view of the raw data rather than a simple overnight summary) is the optimal way to determine if the recording was technically adequate and to arrive at a correct diagnosis. False negative studies may occur for a number of reasons: inadequate sleep, no REM sleep recorded, no sleep in the supine position, respiratory events with minimal desaturation (oximetry screening), sleep disturbance for reasons other than apnea (periodic leg movements), or respiratory-related arousals with minimal apnea/hypopnea. Regardless of whether full polysomnography or an unattended screening study is performed, it is critical that a physician knowledgeable in sleep medicine evaluate the technical adequacy of the study and correlate the findings with the patient's symptoms. When the clinical suspicion of a sleep disorder is high (e.g., sleep apnea), a negative screening study should prompt more comprehensive sleep testing. In the present case, the patient's wife reported hearing her husband stop breathing and then abruptly "snort and gasp for air." This history along with the presence of hypertension, a large neck, and habitual snoring suggested that a sleep study should be performed. The patient had an AHI of 60/hour. After nasal CPAP therapy his energy level improved, and he was more productive at work. TYPE OF STUDY Levels ofSleep Studies BIOPARAMETERS MEASURED Levell Level 2 Level 3 Level 4 Attended polysomnography Unattended full polysomnography Cardiopulmonary study-four or more bioparameters (usually unattended) Single or dual bioparameter(s) EEG, EOG, chin EMG, airflow, effort, EKG, Sa02, ± leg EMG, body position Same as Level I, but technologist not present Airflow, effort, Sa02, EKG (or heart rate), ± body position Oximetry, oximetry + airflow or snoring 103

Clinical Pearls 1. The presence or absence of excessive daytime sleepiness-the cardinal manifestation of OSA - is not a good predictor of OSA. 2. A large neck circumference, hypertension, habitual snoring, and witnessed choking! gasping during sleep are good predictors of the presence of sleep apnea. 3. Screening studies (depending on the data recorded) may have a high number of false negatives if mild-to-moderate cases of OSA are studied. 4. Obtain an attended, full polysomnography study when there is a high index of suspicion for OSA but a negative screening study. 5. Raw data and tracings of screening studies must be visualized-summary of data can be misleading. 6. Screening studies probably do not save money if a subsequent attended CPAP titration is required in most cases. Attended polysomnography of the partial-night type (initial diagnostic portion, then CPAP titration) is probably more cost-effective if moderate to severe OSA is likely. REFERENCES 1. Young T, Palta M, Leder R, et al: The occurrence of sleep-disordered breathing among middle-aged adults. N Engl J Med 1993; 328: 1230-1235. 2. Flemons WW, Whitelaw WA, Brant R, et al: Likelihood ratios for a sleep apnea clinical prediction rule. Am J Respir Crit Care Med 1994; 150: 1279-1285. 3. Standards of Practice Committee, American Sleep Disorders Association: Portable recording in the assessment of obstructive sleep apnea. Sleep 1994; 17:378-392. 4. Netzer NC, Stoohs RA, Netzer CM, et al: Using the Berlin Questionnaire to identify patients at risk for the sleep apnea syndrome. Ann Intern Med 1999; 132:485--491. 104

PATIENT 31 A 40-year-old woman with "mild" sleep apnea A 40-year-old woman was referred by her internist for evaluation of daytime sleepiness. About a year ago he had ordered a screening sleep oximetry study, which showed minimal desaturation. The patient was labeled as having mild disease and told to lose weight. However, her daytime sleepiness persisted despite 10 pounds of weight loss. Her husband reported that she snored softly and frequently had a "pause" in breathing. Her score on the Epworth Sleepiness Scale was 20/24 (severe sleepiness). Physical Examination: Mildly obese woman - weight 120 pounds, height 5 feet 2 inches. HEENT: edematous palate. Neck: 15 ~inch circumference. Chest: clear. Cardiac: normal. Extremities: no edema. Figure: The initial portion of the oximetry study is shown below. Question: Do you agree that this patient has mild sleep apnea? 105

Diagnosis: Severe sleep apnea with minimal arterial oxygen desaturation. Discussion: In determining the severity of obstructive sleep apnea (OSA), several factors must be considered. Some patients with minimal desaturation have frequent events, a high arousal index, and severe daytime sleepiness. Thus, if an oximetry study is used to screen patients for OS A, the tracing should be scrutinized for evidence of the sawtooth pattern of repeated changes in SaO,. It is not sufficient to simply observe summary information. In fact, some patients have respiratory arousals without any changes in SaO,. Alternatively, other patients with an AHI in the -moderate range (15- 30/hr) have impressive arterial oxygen desaturations and a low sleeping baseline SaO,. Studies of breath holding in normal subjects suggest that the rate of Sa02 fall is inversely proportional to the baseline Sa02 and to the lung volume (oxygen stores) at the start of breath hold. The rate of fall is disproportionately higher at low lung volumes secondary to increases in ventilation/perfusion mismatch. A study of OSA patients found that the severity of nocturnal arterial oxygen desaturation was related to several factors, including: the awake supine Pa02, the percentage of sleeptime spent in apnea, and the expiratory reserve volume. Patients with a baseline Pa02 of 55-60 mmHg are on the steep part of the oxyhemoglobin saturation curve. A small fall in Pa02 results in significant desaturation. While apnea duration is an obvious factor in the severity of desaturation, the length of the ventilatory period between events also is important. Some patients do not completely resaturate between events as they quickly return to sleep, and the airway closes again. Long event duration and short periods between apneas mean the percentage of sleeptime spent in apnea is high. The clinical significance of a low expiratory reserve volume (ERV) may be less obvious. The ERV is the difference between the functional residual capacity (FRC; end expiratory lung volume) and the residual volume (RV; volume at maximal exhalation). The FRC is reduced in obesity secondary to low compliance of the chest wall/abdomen. The residual volume is increased if patients have any degree of obstructive airway disease (airtrapping). 106 Thus a low ERV means that the patient has low oxygen stores at the start of apnea (a low FRC) and significant ventilation/perfusion mismatch at low lung volumes (identified by a high RV). Clinically, the groups of OSA patients with severe desaturation include patients with a low POJ for any reason (severe obesity, daytime hypoventilation, and chronic obstructive pulmonary disease [COPD]). In fact, some patients can have significant desaturation after events as short as 10-15 seconds. The severity of desaturation also depends on sleep stage. In most OSA patients, the longest apneas and most severe desaturations occur in REM sleep. Some studies also have suggested that at equivalent apnea length, the severity of desaturation is worse in obstructive than central apnea. In evaluating oximetry, remember that patients with a high carboxyhemoglobin (heavy smoking) will have a falsely high SaO, because the devices do not differentiate carboxyhemoglobin and oxyhemoglobin. In addition, oximeters vary widely in their ability to record desaturations. This is especially true if the averaging time is long (see Fundamentals 10). Factors Determining the Severity of Desaturation Event duration Length of ventilatory period between events Baseline sleeping Sa02 (P02) Oxygen stores (FRC) Tendency for V/Q mismatch-presence of lung disease, ERV = FRC - RV In the present case, the oximetry tracing revealed a subtle "sawtooth" pattern consistent with frequent small changes in Sa02. A complete sleep study revealed an AHI of 60/hr. The events were short (mean duration 15 seconds), and the baseline sleeping Sa02 was 95%. The patient was only mildly obese and had no evidence of COPD. Thus, it is not surprising that the arterial oxygen desaturation was mild. However, the arousal index was 55/hr-consistent with severe sleep fragmentation. The patient was treated with nasal CPAP, and she experienced a rapid improvement.

Clinical Pearls 1. Some patients with minimal arterial oxygen desaturation have severe OSA, indicated by a high AHI and severe sleep fragmentation. 2. The severity of arterial oxygen desaturation in patients with OSA depends on the baseline supine P02. the percentage of apnea time (apnea length), and the degrees of obesity (decreased FRC) and obstructive lung disease (increased RV). 3. Arterial oxygen desaturation usually is more severe in REM than NREM sleep and in obstructive rather than central apnea (equivalent length). 4. Screening for OSA with oximetry can result in false negatives in patients with mild arterial oxygen desaturation during apnea/hypopnea. 5. When evaluating oximetry results always examine the tracings as well as the summary data. REFERENCES I. Findley U, Ries AL, Tisi OM: Hypoxemia during apnea in normal subjects: Mechanisms and impact of lung volume. J Appl Physiol 1983; 55: 1777-1783. 2. Bradley TD, Martinez D, Rutherford R, et al: Physiological determinants of nocturnal arterial oxygenation in patients with obstructive sleep apnea. J Appl Physiol 1985; 59: 1364-1368. 3. Series F, Cormier Y, La Forge J: Influence of apnea type and sleep stage on nocturnal postapneic desaturation. Am Rev Respir Dis 1990; 141:1522-1526. 4. Netzer N, Eliasson AH, Netzer C, et al. Overnight pulse oximetry for sleep-disordered breathing in adults: A review. Chest 200 I; 120:625-633. 107

FUNDAMENTALS OF SLEEP MEDICINE 12 Respiratory Arousals Respiratory arousals are associated with respiratory events. They include arousals associated with the termination of apnea, hypopnea, and respiratory effort-related arousals (RERAs). RERAs are also called upper airway resistance syndrome events. RERAs are defined as arousals associated with episodes of high inspiratory effort that do not meet criteria for obstructive apnea or hypopnea. The gold standard measurement of inspiratory effort is esophageal (or supraglottic) pressure deflections. Usually arousal follows a crescendo pattern of progressive effort over several breaths, but arousal from a sustained increased effort also qualifies. The rationale for the definition of RERAs is based on studies of the mechanisms by which respiratory stimuli induce arousal from sleep. These studies suggest that the arousal stimulus during upper airway narrowing or closure is related to the level of inspiratory effort (airway suction pressure, esophageal pressure deflection), at least during NREM sleep. While either hypoxia or hypercapnia may drive inspiratory effort, it is the magnitude of the inspiratory effort and the threshold for arousal that determine if arousal occurs. During experimental mask occlusion in normal subjects, arousal usually occurs when suction pressure reaches 20-30 em H20 . However, snorers and patients with OSA may not arouse until esophageal pressures reach 40-80 cm H20 ; these results imply a decrease in arousability in these groups. During obstructive apnea/hypopnea, the arousal response and apnea termination are associated with a preferential increase in upper airway muscle activity and restoration of airway patency. Thus, while frequent arousals are believed to cause excessive daytime sleepiness in patients with OSA, the arousal response is believed essential for termination of the apnea. The fact that only 60-80% of event terminations are associated with clear-cut cortical EEG changes may represent a lack of sensitivity of routine monitoring methods (limited montage) or arousal on a subcortical level. It has been demonstrated that subcortical or "autonomic arousals" (increase in heart rate or blood pressure) also result in daytime sleepiness. Does counting and tabulating arousals add anything to the managment of patients with OSA? Some have argued that the respiratory arousal index (RAI), which is the number of respiratory arousals per hour of sleep, adds little to the AHI with respect to patient management. However, neither the AHI nor the RAJ correlate very well with objective or subjective measures of sleepiness. This probably is explained by variability in the sleep need of individuals. The RAJ might still be useful if a level could be identified that would allow the determination that a patient's sleepiness is the result of sleep-disordered breathing. The original description of the upper airway resistance syndrome identified a group with an AHJ < 5/hr who had daytime sleepiness that improved after treatment of upper airway obstruction. The group had an RAI > lO/hr as a criterion. However, there are no large studies of the RAJ in normal subjects. One small study found a median RAJ of 10/hr in a group of normals and 21/hr in a group with upper airway resistance syndrome (Epworth score> 10 and AHJ < l5/hr). Until better information is available, it is reasonable to assume that an RAJ> lO/hr can explain sypmptoms of daytime sleepiness as long as no other cause for sleepiness is available. Alternatively, there may be some individuals with an RAJ < 10 /hr on a given study who do have upper airway-associated daytime sleepiness. Monitoring of esophageal pressure is the gold standard for detecting increased respiratory effort. As such, measurement of esophageal pressure is considered essential to detect RERAs. However, esophageal pressure is rarely measured in most clinical sleep laboratories. It has been recognized that flattening of the nasal pressure signal identifies periods of high respiratory effort with reasonable accuracy. A flow limitation arousal (FLA) may be defined as an arousal preceded by significant flattening which temporarily resolves after the arousal. One study (see figure) suggested that FLAs and RERAs using esophageal pressure correlated fairly well. Of note, in this study there were patients with daytime sleepiness and an 108

o 10 20 30 Flow limitation events with arousal by nasal cannula (events/hr) o 20 . I 0 asymptomati~ • EDSlsnorinz I 30 zQ5 E o c o E ~~OlE O~ .cc 0.(1) &> 10 (1)- w a:: RAI < lO/hr. In addition, some arousals were preceded by flow limitation but no increase in esophageal pressure, and others were preceded by an increase in esophageal pressure without flow limitation. Flow limitation events can be seen without an associated arousal. These appear to be common even in normal individuals. The significance of these events remains to be determined. In sleep laboratories not using nasal pressure, "snore arousals" (defined as those following heavy snoring) are included in the RAI. In patients with milder aSA, the definition of hypopnea really determines how many RERAs are present. If you accept any drop in the flow (nasal pressure) for lO or more seconds + an arousal as a hypopnea, there will be few RERAs as most events will be termed hypopneas. If you require a 4% desaturation and a 30% drop in flow, then there will be fewer hypopneas and more RERAs. Key Points 1. Respiratory effort arousals (RERAs) are arousals following periods of high inspiratory effort that do not meet criteria for apnea or hypopnea. 2. Flow limitation arousals (nasal pressure) appear to provide a reasonable estimate of the RERA frequency in most patients. Esophageal monitoring is still the gold standard for detecting RERAs. 3. The respiratory arousal index is the number of arousals per hour of sleep that are associated with apnea, hypopnea, or RERAs. 4. An RAI > lO/hr + symptoms of daytimes sleepiness + no other reason to explain sleepiness can be used to define the upper airway resistance syndrome or mild aSA. 5. Although the RAI is an index of the amount of sleep fragmentation secondary to respiratory events, it does not correlate better with measures of sleepiness than the AHI in unselected patients (both have a significant but low correlation in most studies). 6. The relative number of hypopneas and RERAs will depend on the definition of hypopnea. REFERENCES I. Guillemenault C, Stoohs R, Clerk A, et al: A cause of excessive daytime sleepiness: The upper airway resistance syndrome. Chest 1993; 104:781-787. 2. Berry RB, Gleeson K: Respiratory arousal from sleep: Mechanisms and significance. Sleep 1997; 20:654-675. 3. Martin SE, Wraith PK, Deary IJ and Douglas NJ: The effect of nonvisible sleep fragmentation on daytime function Am J Respir CritCareMed 1997; 155:1596-1601. 4. Douglas NJ: Upper airway resistance syndrome is not a distinct syndrome. Am J Resp Crit Care Med 2000;161: 1410-1415. 5. Ayappa I, Norman RG. Krieger AC, et al: Noninvasive detection of respiratory effort-related arousals (RERAs) by a nasal cannula/pressure transducer system. Sleep 2000;23:763-771. 6. Rees K, Kingshott RN, Wraith PK, Douglas NJ: Frequency and signficance of increased upper airway resistance during sleep. Am J Resp Crit Care Med 2000;162: 1210-1214. 109

PATIENT 32 A 30-year-old woman with severe fatigue A 30-year-old woman of normal body weight complained of severe fatigue and daytime sleepiness of 3-year duration. Her husband reported that she snored, especially during periods of nasal congestion. There was no history of symptoms characteristic of narcolepsy (cataplexy, sleep paralysis, or hypnagogic hallucinations). The patient reported getting at least 8 hours of sleep each night and denied feeling depressed. There was no history of alcohol or sedative use. An extensive medical examination found no cause for the patient's fatigue. A polysomnogram and a multiple sleep latency test (MSLT) were performed at another hospital. The Epworth Sleepiness score was 18/24 (moderate to severe sleepiness). Physical Examination: General: normal body weight. HEENT: high arched hard palate, dependent soft palate, long uvula. Prior Sleep Study* Total sleep time Sleep period time (SPT) Sleep latency REM latency AHI AHI, NREM AHI, REM Desaturations** 406.55 min (394-457) 432.5 min (414-453) 2.5 min (0-19) 70 min (69-88) 3 events/hour « 5/hr) o 12.3 o Sleep Stages Stage Wake Stage I Stage 2 Stages 3 and 4 Stage REM O/OSPT 6 (0-6) 8.8 (3-6) 56.1 (46-62) 13.8 (7-21) 15.3 (21-31) Total arousal index Respiratory arousal index 25/hr 2/hr Spontaneous arousal index PLM arousal index 23/hr O/hr *Airflow detected by thermistor **Drops in SaO) of 4% or more Hypopnea = 30% reduction in airflow for'" 10 seconds + 2% drop in SaO) or an arousal ( ) = normal values for age. AHI = apnea + hypopnea index. PLM = periodic limb (leg) movement MSLT: Mean sleep latency 2 minutes, no REM periods in five naps. Figure: The results above prompted another sleep study, with nasal pressure monitoring. Events similar to the one illustrated below were common. Questions: What is the cause of the patient's severe sleepiness? How can you explain the high number of spontaneous arousals? 94% 5a02 abdomen C4-Al 02-Al ROC-A1 LOC-A2 chin EKG ~~. H --\ I j. to M I I I H I I r l .. , I I • I snore nasal pressure chest airflow (therm) 110

Diagnosis: Upper airway resistance syndrome or mild OSA. Discussion: When symptoms or findings of daytime sleepiness are more severe than expected from the AHI on a sleep study, several possibilities must be considered. The sleep study may have underestimated the severity of illness (no supine monitoring, low amount of REM sleep). Additionally, a variety of disorders can cause daytime sleepiness, including insufficient sleep, narcolepsy, depression, periodic leg movements in sleep, idiopathic hypersomnolence, drug abuse, and the upper airway resistance syndrome (UARS). UARS is manifested by little or no discrete apnea or hypopnea but repeated arousal during periods of high upper airway resistance (increased inspiratory effort). Fatigue rather than daytime sleepiness can be the major complaint. While snoring is common in UARS, not all patients with this syndrome snore. This diagnosis may be missed with routine sleep monitoring unless the large number of unexplained arousals is noticed. On close examination, subtle changes in airflow or inspiratory effort precede the arousals. Monitoring of esophageal pressure in these patients reveals high esophageal pressure deflections preceding arousal (see figure, next page). A progressive increase in respiratory effort-the crescendo pattern-may be seen prior to arousal. Other patients have a stable but high level of inspiratory effort associated with arousal. The arousals characterizing UARS are called respiratory effort-related arousals (RERAs). An RERA is defined as an arousal following a period of high or crescendo respiratory effort (esophageal pressure deflections) that does not meet criteria for obstructive apnea or hypopnea. A given event might be classified as either an RERA or a hypopnea depending on the definition for hypopnea. For example, a reduction of airflow for more than 10 seconds associated with increased inspiratory effort and followed by an arousal but not a 4% desaturation would be classified as an RERA if one requires a 4% desturation for an event to be classified as a hypopnea. The same event would be called an obstructive hypopnea using definitions not requiring a desaturation. While esophageal pressure is not available in many sleep labs, nasal pressure monitoring is gaining popularity. As discussed in Fundamentals 10, nasal pressure monitoring is noninvasive and requires only a sensitive pressure transducer and a monitoring system with either DC capability or AC amplifier with a very low frequency filter capability. Nasal pressure monitoring in UARS reveals repetitive episodes of airflow limitation (flattening) followed by an arousal and temporary reversal of the flattening (return to a rounded pattern). One study has suggested that nasal pressure will detect the majority of the RERAs detected by esophageal pressure monitoring (see Fundamentals 12). However, esophageal pressure monitoring remains the gold standard. It is believed that the symptoms of UARS are secondary to arousals and sleep fragmentation. However, other factors could also be important. Interestingly, a group of sleepy women with high esophageal pressure deflections during sleep but minimal arousals improved after treatment of presumed UARS. An expert panel of sleep physicians recently suggested that UARS is not a separate syndrome, but simply part of the spectrum between simple snoring and obstructive apnea. It was also suggested that RERAs be separately tabulated as part of polysomnography. An alternative is to simply include RERAs as a component of the events comprising respiratory arousals (i.e., in addition to arousals following apneas and hyopneas). In patients with a high RERA index (RERAs per hour of sleep), the AHI may not accurately classify the severity. Some have suggested adding the RERA index and AHI to identifity a true "respiratory disturbance index." In the current case, the previous polysomnogram showed a low overall AHI, but the arousal index was elevated. Airflow was monitored with a thermistor, and the definition of hypopnea was at least a 30% reduction in flow for 2: 10 seconds associated with a 2: 2% desaturation or an arousal. The AHI was < 5/hr, and the MSLT confirmed daytime sleepiness without evidence of REM onsets. The patient was given the diagnosis of idiopathic hypersomnia. However, repeat testing using the same hypopnea definition but nasal pressure to monitor flow revealed an AHI of 8/hr. If a 4% desaturation was required for an event to be classified as a hypopnea, the AHI was < 5/hr. The patient might then be labeled as having UARS. However, with the hypopnea definition used in the study the patient was considered to have mild OSA. In the tracing on page 110, note the flattening in the nasal pressure signal but minimal changes in the airflow (thermistor) signal. There was no desaturation. The event was scored as an RERA, as the flow was not reduced by 30% for 2: 10 seconds but the flattened nasal pressure signal suggested increased inspiratory effort. (Using more liberal definitions of hypopnea, you might consider the event an obstructive hypopnea.) The patient did not wish to try CPAP or an oral appliance. She underwent a uvulopalatopharyngoplasty and attempted to lose weight. She experienced a gradual resolution of her symptoms over the next 3 months. 111

While one may argue about definitions of RERAs and hypopneas, respiratory arousals were causing this patient's symptoms. The spontaneous arousals on the first study were correctly identified as respiratory arousals with the use of nasal pressure. The normal range forthe respiratory arousal index (RAl) has not been well defined. However, many consider an RAJ > lO/hr in a symptomatic patient without other explanations for sleepiness to be an indication for treatment. ROC-A 1 LOC - A 2 chin EMG nasal pressure (airflow) 5 sec 50 uv I --=---- Esoph Pressure Clinical Pearls I. The UARS (or RERAs) always should be considered in patients with unexplained, excessive daytime sleepiness. 2. Using standard monitoring, the only clue that UARS may be present is repetitive episodes of subtle changes in respiration followed by arousals (or transient EEG changes). 3. Monitoring of esophageal pressure can help diagnose UARS by documenting high levels of inspiratory effort (upper airway resistance) preceding arousal. 4. Nasal pressure monitoring may allow recognition of RERAs in most cases. However, esophageal pressure remains the gold standard. 5. Whether an event is labeled an RERA or an obstructive hypopnea depends on the definition of hypopnea. REFERENCES I. Guilleminault C. Stoohs R. Clerk A. et al: A cause of excessive daytime sleepiness: The upper airway resistance syndrome. Chest 1993; 104:781-787. 2. Guilleminault C. Stoohs R. Clerk A, et al: Excessive daytime somnolence in women with abnormal respiratory effort during sleep, Sleep 1993; 16:S137-138, 3, Guilleminault C. Stoohs R. Kim U. et al: Upper airway-sleep-disordered breathing in women. Ann Intern Med 1995; 122:493-501. 4. American Academy of Sleep Medicine Task Force: Sleep-related breathing disorders in adults: Recommendation for syndrome definition and measurement techniques in clinical research. Sleep 1999;22:667-689. 5. Loube DI. Gay PC. Strohl KP. et al: Indications for positive airway pressure treatment of adult obstructive sleep apnea patients. A consensus statement. Chest 1999; 115:863-866. 6. Ayappa I. Norman RG. Krieger AC, et al: Noninvasive detection of respiratory effort-related arousals (RERAs) by a nasal cannula/pressure transducer system. Sleep 2000;23:763-771. 112

PATIENT 33 A 30-year-old man with heavy snoring and daytime sleepiness A 30-year-old man was evaluated for complaints of heavy snoring, moderate daytime sleepiness (Epworth Sleepiness Scale score 18/24), and apnea (witnessed by his wife) of at least 5-year duration. The patient had gained about 30 pounds over this period. He admitted to drinking several cocktails nightly. There was no history of cataplexy or sleep paralysis. Physical Examination: Blood pressure 160/88, pulse 88. General: obese - weight 210 pounds, height 5 feet 10 inches. HEENT: dependent, edematous uvula. Neck: 18-inch circumference. Chest: clear. Cardiac: normal. Extremities: no edema. Sleep Study AHI AHI, NREM sleep AHI, REM sleep PLM index Arousal index Minimum Sa02 Body position AHI = apnea + hypopnea index, PLM = periodic limb movement 12/hr 10/hr (80% of total sleep time) 20/hr (20% of total sleep time) 5/hr 20/hr 85% 25% lateral decubitus, 75% supine Question: In view of the severe symptoms, why does the sleep study document only mild apnea? 113

Answer: The patient was abstinent from alcohol at the time of the study. Discussion: The clinician often is faced with the problem of interpreting a sleep study that at first glance appears to show milder obstructive sleep apnea (GSA) than suspected on the basis of the history (significant daytime sleepiness). Rough guidelines for interpreting the apnea + hypopnea index (AHI) are: AHI < 15 mild, 15-30 moderate, > 30 severe. However, the AHI is only one means of assessing the severity of GSA. In fact, the correlation between AHI and symptomatic or MSLT measures of daytime sleepiness is low (but statistically significant). Two patients with the same AHI may have quite different amounts of daytime sleepiness. The arousal index also must be considered. Some patients with an AHI indicating mild apnea have considerably more respiratory arousals secondary to high respiratory effort (respiratory effort-related arousals). Another explanation for a sleep study showing milder than expected GSA is the normal night-tonight variation in the AHI, which is due to different body positions, variations in nasal resistance, variable amounts of REM sleep, and the effects of medications and beverages. Some patients have apnea only while supine or in stage REM sleep. Therefore, in these patients the AHI depends on the amount of time spent sleeping supine or in REM sleep. Nasal congestion is a factor because increases in nasal resistance increase the amount of apnea. Always consider the patient's use of ethanol when evaluating GSA. This drug has a powerful, preferential, inhibitory effect on upper airway muscle activity and increases snoring and apnea. In addition to increasing the AHI, ethanol impairs the arousal response to airway occlusion; thus, apneas tend to be longer and associated with more severe desaturations. Ethanol suppresses REM sleep. This usually results in an increase in the REM latency and a shift of REM toward the morning (as the ethanol level drops). Therefore, the regular use of ethanol worsens sleep apnea considerably. Conversely, abstinence from ethanol could reduce the AHI and apnea duration/degree of arterial oxygen desaturation recorded on a sleep study. Some have hypothesized that ethanol might reduce the effectiveness of an optimal level of nasal CPAP. However, at least two studies have shown that this is not the case. The use of alcohol still should be discouraged in patients undergoing this treatment, because they may fall asleep without putting the nasal CPAP on, or they may remove the CPAP during the night. Many sleep labs find pre- and post-study questionnaires very helpful in documenting the typical and actual prestudy intake of alcohol or hypnotics. Patients should be asked about drugs they took prior to the sleep study as well as about any usual medications or ethanol they did not take. This will help the physicians reading the study. In the present patient, the arousal index was only slightly higher than the AHI, and both REM sleep and the supine position were evaluated. However, the patient did not ingest his favorite alcoholic beverages before this sleep study. Repeat testing after the patient's usual intake of ethanol showed an AHI of 40 and an increase in mean apnea duration by 5 seconds. Clinical Pearls I. When a sleep study reveals milder apnea than suspected on the basis of clinical symptoms, consider the effects of body position, the amount of REM sleep, and the possible presence of the upper airway resistance syndrome (arousal index much higher than the AHI). 2. In milder GSA there may be more night to night variability in the AHI 3. The effects of ethanol use (or abstinence) on the severity ofsleep apnea always should be considered. 4. Ethanol use increases the amount of apnea and the duration of obstructive events, as well as the severity of desaturation. 5. Ethanol increases the REM latency and decreases the amount of REM sleep. 114

REFERENCES I. Taasan V, Wynne JW, Cassisi N, Block AJ: The effect of nasal packing on sleep-disordered breathing and nocturnal oxygen desaturation. Laryngoscope 1981; 91(7): 1163-1172. 2. Issa FG, Sullivan CE: Alcohol, snoring, and sleep apnea. J Neurol Neurosurg Psychiatry 1982; 45:353-359. 3. Remmers JE: Obstructive sleep apnea- A common disorder exacerbated by alcohol. (Editorial) Am Rev Respir Dis 1984; 130:153-155. 4. Berry RB, Desa MM, Light RW: Effect of ethanol on the efficacy of nasal continuous positive airway pressure as a treatment for obstructive sleep apnea. Chest 1991; 99:339-343. 5. Berry RB, Bonnet MH. Light RW: The effect of ethanol on the arousal response to airway occlusion during sleep in normal subjects. Am Rev Respir Dis 1992; 145:445-452. 115

PATIENT 34 A 45-year-old man with a distinct pattern of desaturation A 45-year-old man was evaluated for complaints of heavy snoring for many years and daytime sleepiness of about 4-year duration. There was no history of cataplexy or sleep paralysis. The patient did not drink alcohol. Physical Examination: Height 5 feet 10 inches, weight 180 pounds. HEENT: edematous uvula. Neck: 15-inch circumference. Otherwise normal exam. Figure: A tracing of oximetry for this patient is show below. Sleep Study Time in bed Total sleep time Sleep period time (SPT) Sleep latency REM latency Arousal index (/hr) AHI AHI, NREM sleep AHI, REM sleep 480 min (390-468) 350 min (343-436) 425 min (378-452) 10 min (2-18) 90 min (55-78) 20/hr 12/hr «5/hr) 5/hr 50/hr Sleep Stages Stage Wake Stage I Stage 2 Stages 3 and 4 Stage REM PLM index O/OSPT 18 (1-12) 13(5-11) 49 (44-66) 5 (2-15) 15 (19-27) O/hr ( ) = normal values for age, PLM = periodic leg movement. Questions: Does the overall AHI of 12/hr mean this patient has mild OSA? Should he be treated? 2 hrs 4hrs 6 hrs 8 hrs Body Back Side Back Position 100 8a0 2 80 t REM periods 116

Answer: REM-related obstructive sleep apnea with severe arterial oxygen desaturations. Discussion: Some patients have episodes of obstructive sleep apnea (OSA) primarily during REM sleep. Thus, the overall AHI may be low even if the AHI during REM sleep is fairly high. Other OSA patients have a much higher AHI or more severe arterial oxygen desaturation during REM sleep. In the absence of other disorders to explain the excessive daytime sleepiness, many clinicians have empirically treated patients who have REMspecific sleep apnea with usual treatments such as nasal continuous positive airway pressure (CPAP). Some patients with REM-specific sleep apnea have heavy snoring during NREM or exhibit repetitive respiratory effort-related arousals secondary to high inspiratory effort during NREM sleep without overt apnea or hypopnea. If nasal pressure is used to detect airtlow, these events may be more obvious. One study of a group of patients with a low overall AHI « IO/hr), but varying amounts of apnea-hypopnea during REM sleep, found that 80% with a REM-specific AHI > 15/hr had a short sleep latency during daytime naps (evidence of excessive sleepiness). Thus, frequent apneahypopnea during REM sleep alone may justify treatment. Although the finding of a higher AHI during REM sleep is common, the reason is still not clear. It has been assumed that since REM sleep is associated with muscle hypotonia, it results in a more collapsible upper airway. However, studies of the critical closing pressure have not confirmed that the airway is more collapsible during REM sleep. It is possible that this study method does not accurately represent the physiology of REM sleep. REM sleep is not homogeneous, and upper airway muscle tone is often lowest during bursts of phasic eye movements. This can be seen with intramuscular electrodes in the genioglossus but not on the chin surface EMG (see figure below). In addition, ventilatory drive (esophageal pressure deflections) may also decrease during bursts of phasic eye movements. In the tracing below in a normal individual, note the three smaller breaths and lower genioglossus muscle activity (tongue muscle) illustrating the non-homogenous nature of REM sleep. In addition, the frequency of eye movements increases in the later REM periods of the night. Periods of decreased ventilation also tend to be more common later in the night even in normal individuals (see Patient 10). During REM sleep, end expiratory lung volume also falls, and this may also have an effect on upper airway patency as upper airway volume decreases with decreases in lung volume. The desaturations also tend to be more severe and apneic events longer during REM than NREM sleep. In the present case, the patient had a very high AHI in REM sleep and almost no apnea-hypopnea during NREM sleep (AHI = 5/hr). The REM-associated apneas were quite long, and the desaturation impressive (see first figure). The patient underwent a nasal CPAP trial and was treated with 10 cm H2 0 , with resolution of his symptoms. During the trial, he had a large rebound in the amount of REM sleep (30% of SPT). EOG«===~= = EEG< 'r: ~'--~ airflow f"L'7 integrated genioglossus eMG J 117

Clinical Pearls I. Some patients have obstructive sleep apneas and hypopneas primarily during REM sleep, resulting in a low overall AHI despite frequent events and associated arousals and desaturations during REM sleep. 2. In patients with excessive daytime sleepiness and significant, REM-specific sleep apnea, treatment is indicated. Other possible causes of daytime sleepiness should be excluded. 3. The reasons for the higher AHI during REM sleep are complex. REM sleep is not homogenous, and episodes of decreased upper airway muscle activity or ventilatory drive may be the cause of hypopneas or apneas during REM sleep. REFERENCES I. Gould GA. Gugger M. Molloy J. et al: Breathing pattern and eye movement density during REM sleep in humans. Am Rev Respir Dis 1988; 138:874-877. 2. Wiegand L, Zwillich CWoWiegand D, White DP: Changes in upper airway muscle activation and ventilation during phasic REM sleep in normal men. J Appl PhysioI1991;71:488-497. 3. Kass JE, Akers SM, Bartter TC, et al: REM-specific sleep-disordered breathing: A possible cause of excessive daytime sleepiness. Am J Respir Crit Care Med 1996; 154: 167-169. 4. Boudewyns A, Punjabi N, Van de Heyning PH, et al: Abbreviated method for assessing upper airway function in obstructive sleep apnea. Chest 2000; 118:1031-1041. 5. Penzel T, Moeller M, Becker HF: Effects of sleep position and sleep stage on collapsibility of the upper airways in sleep apnea. Sleep 2001;24:90-95. 118

FUNDAMENTALS OF SLEEP MEDICINE 13 Treatment of Obstructive Sleep Apnea The first, essential determination in selecting appropriate treatment for obstructive sleep apnea (OSA) is the severity of illness. The AHI provides an overall index of the frequency of respiratory disturbance. Arbitrary but useful guidelines are: mild 5-14, moderate 15-30, severe> 30 events/hr. Factors in Assessing GSA Severity Apnea + hypopnea index Amount of supine sleep and REM sleep in the diagnostic study Respiratory arousal index (sleep fragmentation) Severity of daytime sleepiness/requirement for alertness Severity of arterial oxygen desaturation Comorbid illness (e.g., congestive heart failure, ischemic heart disease) Sleep-associated arrhythmias Objective measurements of daytime sleepiness It also is essential to consider factors that may have prevented a night of sleep recording from accurately estimating the usual severity of illness (e.g., minimal supine sleep, lack of REM sleep). This is especially important for partial-night studies. The amount of sleep fragmentation (frequency of arousals) may correlate better with symptoms of daytime sleepiness than the AHI in individual patients with minimal desaturation. The severity of arterial oxygen desaturation can have important consequences for the development of pulmonary hypertension and right heart failure. Some patients have minimal desaturation, but are very sleepy. Other patients have mild symptoms of daytime sleepiness, but severe signs of cor pulmonale. When assessing the severity of daytime sleepiness, question family members about their estimates of the patient's sleepiness. Additionally, consider the requirements for alertness; for example, a professional truck driver has a more critical need for alertness than a retired 80-year-old patient. A consenus conference recently recommended that all patients with an RDI of > 30/hr be treated whether or not symptoms were present. In addition, symptomatic patients with an RDI > 5/hr were also felt to require treatment. Here RDI was defined as AHI + RERA index. The clinical practice parameters of the American Academy of Sleep Medicine suggest that treatment is also indicated in sleepy patients with a respiratory arousal index of 10/hr or more. Generally, each category of severity in OSA is best managed by specific treatments (see table). However, a given treatment might work in other categories of disease as well. For example, a patient with severe OSA can have dramatic improvement with an oral appliance, or a patient with upper airway resistance syndrome may tolerate treatment with nasal continuous positive airway pressure. All of these treatments are discussed in the following cases. 119

UARS/MILD OSA Weight loss Position therapy Treat nasal congestion Oral appliances UPPP for OSA, UARS LAUP for snoring (UARS?) Positive airway pressure General Treatment Guidelines MODERATEOSA Positive airway pressure Oral appliances UPPP,GAHM Weight loss-adjunctive Position therapy-adjunctive SEVERE OSA Positive airway pressure Tracheostomy GAHM,MMO Oral appliances Weight loss-adjunctive UARS = upper airway resistance syndrome, UPPP = uvulopalatopharyngoplasty, LAUP = laser-assisted uvulopalatoplasty, GAHM = genioglossus advancement hyoid myotomy, MMO = mandibulo-maxillary osteotomy, positive airway pressure = CPAP or bilevel pressure REFERENCES I. Thorpy M, et al: ASDA Standards of Practice Committee. Practice parameters for the treatment of snoring and obstructive sleep apnea with oral appliances. Sleep 1995; 18:511-513. 2. Thorpy M, et al: Practice parameters for the treatment of obstructive sleep apnea in adults: The efficacy of surgical modifications of the upper airway. Report of the American Sleep Disorders Association. Sleep 1996; 19:152-155. 3. Chesson AL, et al: ASDA Standards of Practice Committee. Practice parameters for the indications for polysomnography and related procedures. Sleep 1997;20:406-422. 4. Loube Dr, Gay PC, Strohl KP,et al: Indications for positive airway pressure treatment of adult obstructive sleep apnea patients. A consensus statement. Chest 1999; 115:863-866. 120

PATIENT 35 A 30-year-old man with weight loss and sleep apnea A 30-year-old man-height 5 feet 10 inches, weight 230 pounds-was diagnosed as having severe obstructive sleep apnea (AHI 60/hr). He underwent a nasal CPAP titration, and CPAP at a level of 12 cm H20 was found to reduce his AHI to < 5/hr. The patient underwent treatment and noted a rapid resolution in his daytime sleepiness. However, he found CPAP to interfere with his social life. After 1 year of treatment he began a high-fiber, low-fat diet to reduce elevated cholesterol. Over the next 5 months he lost 30 pounds, and his cholesterol improved. He stopped using CPAP and did not feel that his daytime sleepiness returned. Another sleep study was performed. Based on the results of the study, the patient declined to resume CPAP. Surgical and oral appliance options were discussed. but he did not wish to pursue these treatments. Physical Examination: Neck circumference: 15 inches (was 18 inches at 230 pounds). Sleep Studies Weight (Ibs.) 230 230 200 190 CPAP None 12 cmH20 None None AHI 60/hr 5/hr 5/hr 10/hr AHI supine 70 5/hr 30 10/hr AHI non-supine 50 n/a 5 O/hr Arousal index 45/hr 5/hr 15/hr 10/hr Question: What treatment option do you recommend? 121

Answer: Side sleep position and continued weight loss. Discussion: Obesity is a major risk factor for the development of OSA. In some studies, approximately 66% of patients with OSA were obese (body weight> 120% of predicted). Neck circumference is a better predictor of OSA than body weight, indicating that nuchal obesity is more important than total body obesity. Studies suggest that obesity has direct effects on upper airway function (changes in the nature of upper airway tissues) or indirect effects secondary to changes in lung volume. One MRI study of the upper airway in OSA suggested that fat deposition adjacent to the lateral pharyngeal walls reduced the upper airway size. Another study found thickened lateral pharyngeal walls, but these changes were not secondary to fat or edema. Patients with OSA tend to have small upper airways, although there is considerable overlap. Shape may be as important as size. The shape of the upper airway in OSA patients is different, with the the lateral dimension being the smallest (see figure). Weight loss increases the lateral dimension. The size of the upper airway varies with lung volume, probably secondary to the effects of tracheal displacement ("tracheal tug"). Obesity reduces supine lung volume and, therefore, upper airway size. Exactly how obesity and weight loss change the size and shape of the upper airway remains to be determined. Studies also have shown that the collapsibility of the upper airway is decreased by weight loss, implying that obesity increases the tendency of the upper airway to collapse. Many studies have documented that weight loss of modest proportions (5-10%) can produce significant improvement in sleep apnea. Even patients with mild obesity (110-115% of ideal body weight) can benefit from weight reduction. The results vary among patients; for example, a given amount of weight loss may have more effect on upper body obesity in one individual than in another. Nevertheless, weight loss decreases both the AHI and the collapsibility of the upper airway in patients with OSA. The level of nasal CPAP required to maintain upper airway patency also may decrease following weight reduction. Surgical, behavioral. and pharmacologic approaches to weight loss have all been successful in select groups of patients. The major problem to date has been maintenance of weight loss. Many patients with obstructive sleep apnea (OSA) have a significant worsening of apnea in the supine position. In this position, gravity tends to pull the tongue backward and narrow the airway. Thus, an overall apnea index may reflect moderateto-severe apnea when the patient is supine and minimal apnea in the lateral decubitus body position. In one study, approximately 55% of a large group of patients with sleep apnea had AHIs at least two times higher in the supine than in the lateral decubitus position. Unfortunately, many obese patients with severe sleep apnea still have severe apnea in the lateral decubitus position. In these patients, studies have suggested that elevation of the head of the bed (30 degrees) may be a more effective postural maneuver both to decrease the AHI and to decrease the required level of CPAP to maintain upper airway patency. Because the amount of supine sleep can dramatically affect the overall AHI, always note the amount of supine sleep when comparing two sleep studies. To this end, sleep study reports should detail the amount ofsupine sleep and the AHI both supine and non-supine. In the present patient, weight loss from 230 to 200 made him very position-dependent (much lower AHI in the non-supine position. As he had a low AHI in the lateral position, continued weight loss and the side-sleep position were recommended. He slept with a cylinder of firm foam sewed into a pocket on the back of at-shirt. He also continued to lose weight. A sleep study at 190 pounds showed him to have an AHI of only 5/hr in the lateral sleeping position. right and left p A GSA A NORMAL A = anterior, P = posterior, R ,L 122

Clinical Pearls l. Weight reduction can significantly improve the severity of aSA. The amount of reduction required varies widely, but can be as little as 5-10% of body weight. 2. Weight reduction (weight gain) may decrease (increase) the level of nasal CPAP required to maintain upper airway patency. 3. The major challenge of weight reduction as a treatment for aSA is maintaining the weight loss. 4. Body position has an important effect on the severity of aSA in more than 50% of patients. Always consider this effect when assessing the results of polysomnography. 5. Position therapy can be effective in selected patients. Long-term effectiveness and compliance have not been well documented. 6. Elevating the head of the bed may be a more effective postural maneuver in very obese patients. REFERENCES I. Smith PL, Gold AR, Meyers DA. et al: Weight loss in mildly to moderately obese patients with obstructive sleep apnea. Ann Intern Med 1985; 103:850-855. 2. Schwartz AR. Gold AR, Schubert N. et al: Effect of weight loss on upper airway collapsibility in obstructive sleep apnea. Am Rev Respir Dis 1991; 144:494-498. 3. Schwab RJ. Functional properties of the pharyngeal airway: Properties of tissues surrounding the upper airway. Sleep 1996; 19:5170-S174. 4. Strobel RJ, Rosen RC: Obesity and weight loss in obstructive sleep apnea: A critical review. Sleep 1996; 19:104-115. 5. Neill AM, Angus SM, Sajkov D, McEvoy RD: Effects of sleep posture on upper airway stability in patients with obstructive sleep apnea. Am J Respir Crit Care Med 1997; 155:199-204. 6. Oksenberg A, Silverberg DS, Arons E, Radwan: Positional vs nonpositional obstructive sleep apnea patients. Chest 1997; 112:629-639. 7. Jokic R, Klimaszewski A, Crossley M, et al: Positional treatment vs continuous positive airway pressure in patients with positional obstructive sleep apnea syndrome. Chest 1999; 115:771-781. 123

FUNDAMENTALS OF SLEEP MEDICINE 14 Positive Airway Pressure Titration Positive airway pressure (continuous [CPAP] or bilevel) titration can be performed using an entire night following a diagnostic sleep study, or as the second part of a split-night sleep study following an initial diagnostic portion. Studies have suggested that an adequate positive-pressure titration can be obtained in about 60-80% of patients with a split- or partial-night study. If an adequate titration is not obtained, a repeat positive-pressure titration will be needed. While you might assume that a lack of sufficient supine or REM sleep is the cause of insufficent titration, this was not the case in one study? comparing optimal CPAP pressure on a partial and subsequent full night of CP AP titration. The Standards of Practice Committee of the American Academy of Sleep Medicine has published guidelines for appropriate candidates for split studies. Basically, there must be sufficient time remaining after a diagnosis is firmly established to perform an adequate positive-pressure titration. Note, too, that Medicare guidelines for reimbursement for CPAP require that the AHI used to qualify the patient must be based on 2 hours of recorded sleep (not monitoring). Therefore, it seems prudent to record at least 2 hours ofsleep during the diagnostic portion (see table for essential elements of a good positive-pressure titration). It is important to remember that if a split study is a possiblity, patients should be educated about CPAP (several videos are available), fitted for a mask, and given an adapation period to positive pressure before the diagnostic portion begins. Education and mask fitting are rarely successful at 2 AM! The adaptation period on a low level of CPAP identifies patients with nasal congestion, those who are very pressure intolerant, and those with clasutrophobia. If nasal congestion is a problem, a full face mask (see Patient 44) may be fitted or a decision made to use heated humidity (see Patient 42) during the positive-pressure titration. Clastrophobic patients may find nasal pillows or other similar interfaces more tolerable. Pressure should be increased slowly in pressure intolerant patients. Bilevel pressure may be used. if high pressure levels are needed (see Patient 45). Criteria for Split Titration • 2 hours of recorded sleep in diagnostic portion (so patient will qualify for Medicare payment of CPAP) • AHI > 40/hr • AHI 20--40 Ihr with significant desaturation or other factors • At least 3 or more hours remaining for CPAP titration ESSENTIALS OF POSITIVE-PRESSURE TITRATION • Patient education about positive pressure • Mask fitting • CPAP training acclimitization period (identify problems early) • Titration to eliminate apnea, hypopnea, snoring, desaturation, and RERAs in all body positions and sleep stages (including supine REM) • Reduction in the AHI to < 51hr if possible; at least to < lO/hr RERA = respiratory effort-related arousals During titration the goal is to prevent apnea, hypopnea, arterial oxgyen desaturation, respiratory effortrelated arousals, and snoring in all sleep stages and body positions. In general, supine REM sleep requires 124

the highest pressure. Most diagnostic positive-pressure machines provide a flow output (CPAP flow or Vest [estimated flow]) that is obtained from a built-in pnemotachograph or other flow-measuring device. The CPAP flow signal is derived from the total flow delivered minus an estimate of leak determined by the machine. This flow signal provides information on both magnitude and shape of airflow. A flattened flow signal signifies some element of airway narrowing (airflow limitation). In general, a reasonable goal is to eliminate significant airflow limitation, especially if it precedes repetitive arousals (RERAs). Some degree of airflow limitation may be tolerated-especially in pressure-sensitive patients-if it is not associated with repetitive arousals. In Figure I, CPAP pressure is reduced (arrow), and the flattened shape of the flow signal correlates with an increase in supraglottic pressure and resistance. As pressure is increased, obstructive apneas, hypopneas, and snoringlRERAS are eliminated in that order. In Figure 2, note the snoring and slightly flattened airflow signal (arrow) on CPAP of 7 em H20. A I em H20 increase in pressure rounds the shape of the CPAP flow tracing and eliminates snoring. Central hypopneas are sometimes noted as the patient returns to sleep after arousals, or during sleep onset. They are characterized by a round airflow profile and a reduced magnitude of flow and effort (chest and abdominal movement). They should not be treated with increases in CPAP. ;:112 11 10 =;::=~===~==::::====::::===:::===: 10 15 TIme (seconds) FIGURE I 20 25 Positive-pressure devices also provide leak information. All mask interfaces have an intentional leak to wash out the CO2 (usually small holes or slots in the mask). The amount of leak increases with increasing pressure. However, sudden large increases in the leak signal suggest that either mask or mouth leak is occurring. In some labs, an oral thermistor is also used, to detect mouth leak. Of note, the flow signal from the positive-pressure device registers flow via the nasal mask, but not oral flow. If the patient breathes through both the nose and mouth while using a nasal mask, the CPAP flow signal will be reduced (simulating a hypopnea) but the leak signal will increase (Fig. 3). Recording of leak information allows the sleep physician to detect such events. ,. CPAP 7 em H20 CPAP Flow Snore CPAP 8 em H20 CPAP Flow -,!~/\r-J~t-~_/~~ Snore I 'If 18,,, ",,"'.' ••, 1 , $d' b • 4.'*4 FIGURE 2 12S

In Figure 3, CPAP flow and leak are shown during a period of simulated nasal and mouth breathing. The leak information does not identify the source of the leak (mask or mouth). However, fluctuating leak without body movement suggests mouth leak. The sleep technologist can zoom in with video monitoring to see if the mouth is open. Attention to the shape of the flow profile and leak information may prevent increases in pressure to address "hypopneas" caused by mouth leaks, central hypopneas, and ineffective pressure secondary to high mask leak. The technician can reseat the mask or employ measures-such as a chin strap, full face mask, addition of heated humidity, or a change to bilevel pressure - to reduce mouth leak. Simulated mouth leak - mimics a hypopnea CPAP flow (Vest) Leak FIGURE 3 REFERENCES l. Iber C, O'Brien C, Schluter 1, et al: Single-night studies in obstructive sleep apnea. Sleep 1991; 14:383-385. 2. Sanders MH. NB Kern, 1P Costantino, et al: Adequacy of prescribing positive airway pressure therapy by mask for sleep apnea on the basis of a partial-night trial. Am Rev Respir Dis 1993; 147:1169-1174. 3. American Thoracic Society: Indications and standards for use of nasal continuous positive airway pressure (CPAP) in sleep apnea syndromes. Am 1 Resp Crit Care Med 1994; 50: 1738-1745. 4. Condos R, Norman RG, Krishnasamy I. et al: Flow limitation as a noninvasive assessment of residual upper-airway resistance during continuous positive airway pressure therapy of obstructive sleep apnea. Am 1 Respir Crit Care Med 1994; 150:475--480. 5. Montserrat 1M, Ballester E, Olivi H: Time course of stepwise CPAP titration. Am 1 Respir Crit Care Med 1995: 152:1854-1859. 6. American Sleep Disorders Standard of Practice Committee, Chesson A. Chairman: Practice Parameters for the Indications for Polysomnography and Related Procedures. Sleep 1997;20:406--422. 126

PATIENT 36 A 55-year-old man with heavy nighttime snoring and daytime sleepiness A 55-year-old obese man was evaluated for complaints of excessive daytime sleepiness of 4-year duration. He snored heavily and was observed by his wife to stop breathing during sleep. Because of a strong suspicion for obstructive sleep apnea (OSA), the patient underwent a partial or split-night study. The first portion of the night was a diagnostic study and the second portion a nasal continuous positive airway pressure (CPAP) titration. The test was terminated at the patient's request because he no longer felt sleepy. He had some initial difficulty tolerating CPAP at 10-12.5 em Hp. Physical Examination: Obese-weight 240 pounds, height 5 feet 11 inches. HEENT: long, dependent palate; 18-inch neck circumference. Chest: clear. Cardiac: normal. Extremities: no edema. Sleep Study CPAP(cm HP) Monitoring time (min) NREM (min) REM (min) AHI (thr) Body position o(diagnostic) 240 180 10 75 Supine 5.0 60 40 10 60 Supine 7.5 60 50 10 50 Lateral 10 60 30 o 40 Lateral 12 40 20 o 10 Lateral Question: Why is this CPAP titration suboptimal? 127

Answer: The efficacy of nasal CPAP in the supine position has not been documented. Discussion: Nasal CPAP is probably the treatment of choice for most patients with moderate-tosevere OSA. The mechanism of action is to provide a pneumatic splint to preserve upper airway patency. The pressure level required to maintain airway patency is determined by CPAP titration: pressure is incrementally increased to find the point at which apnea, hypopnea, snoring, desaturation, and respiratory effort-related arousals are prevented (the "optimal pressure"). This pressure varies 5-20 em H20 among patients. The required pressure even varies in the same patient, depending on body posture (higher supine) and sleep stage (usually the highest pressure is needed in supine REM sleep). Thus, the optimal pressure for a patient is one that is effective in all body positions and stages of sleep. To avoid the high cost of performing two separate sleep studies and delay in treatment, many sleep centers perform partial or split-night studies. As in this case, the first portion of the night is diagnostic; the second portion is a CPAP titration. The length of time required for the diagnostic portion and the indications for initiating a CPAP titration vary between sleep centers. Practice parameters published by the American Academy of Sleep Medicine suggest that a split study should be performed when AHI > 40/hr, or AHI 20-40/hr if very significant desaturation is present. The decision to obtain a split study also is influenced by the wait time required for a repeat CPAP titration in a given locale. At least 3 or more hours of monitoring are needed during the CPAP titration. Approximately 60-80% of patients can have an adequate CPAP titration using the split-study protocol. If titration is inadequate, a repeat sleep study consisting of an entire night of positive-pressure titration is indicated. There is a certain amount of art as well as science in the CPAP titration. It also is a desensitization exercise in which the patient becomes acclimated to the mask and pressure. Sometimes the level of pressure must be reduced (at least temporarily) even if it is not the optimal pressure. Patient education before the sleep study begins can help tremendously in the adjustment process. (Education about nasal CPAP at 2 AM tends to be poorly received by most patients.) The initial acceptance of CPAP depends on an adequate mask fit. Mask and headgear fit also is best determined prior to beginning the study. The art of dealing with patient concerns and determining when to increase the pressure (and when to back offfor awhile) is a challenging part of the sleep technician's job. 128 The proper endpoint for CPAP titration is still the subject of research. When initially introduced, the usual goal was prevention of apnea, hypopnea, and desaturation. Most centers continued to increase the CPAP level until snoring was eliminated. With the recognition that high upper airway resistance (high inspiratory effort) can induce arousals even if snoring is absent (upper airway resistance syndrome), upward titration of CPAP typically is continued if repiratory effort-related arousals are occurring. A few laboratories place esophageal catheters as part of routine sleep studies/CPAP titrations. Those centers titrate until esophageal pressure swings reach 5-10 em H20 . Other centers have used the shape of inspiratory airflow (flow limitation) to identify high upper airway resistance. Flow-limited breaths are characterized by a constant or decreasing flow despite increased respiratory effort (esophageal pressure deflection). The tracing of inspiratory airflow is flat and wide rather than rounded. The pattern of flow limitation usually is noted during periods of high inspiratory effort. This method requires more accurate airflow monitoring than possible with thermocouples. Monitoring must be performed with pneumotachographs or nasal cannula attached to pressure transducers. Fortunately, most diagnostic CPAP units have an analog output derived from a flow signal measured with an internal pneumotachograph. CPAP is titrated upward until evidence of clinically significant flow limitation is eliminated. The best method of choosing the optimal CPAP pressure remains to be determined. Ultimately, the goal of CPAP titration is to produce high-quality, restorative sleep rather than endpoints related to upper airway function. In the present patient, the AHI was reduced to 10/hr-a fairly reasonable level-at a CPAP level of 12 ern H20 . However, efficacy was not documented in the supine position. Thus, the true optimal pressure for this patient has not been determined. In retrospect, less time should have been spent on the diagnostic part of the study (a case of obvious severe OS A) and more time on the CPAP titration. In this case, you could try empiric treatment with a slightly higher pressure of 13-14 ern H20 ; order another CPAP titration (the gold standard); or treat the patient with autoCPAP for a few weeks to establish a prescription pressure. A final alternative is to begin treatment with 12 em H20 and increase pressure if snoring is present in the supine position or symptoms do not resolve.

Clinical Pearls 1. The goal of the CPAP titration is to determine the optimal pressure-one that is effective in all body positions and stages of sleep. 2. Body position has a major impact on the pressure required to maintain upper airway patency. The highest pressure required is usually in supine REM sleep. 3. The optimal CPAP level should not only prevent sleep-disordered breathing, but allow for maintenance of high-quality, restorati ve sleep (reduced arousals, normal sleep architecture). 4. A split-night study is economical, but reduces the time both for diagnosis and CPAP titration. Appropriate education pre-study is essential. 5. CPAP titration is an art as well as a science. The importance of education and technicianpatient rapport cannot be overemphasized. REFERENCES 1. Pevernagie OA, Sheard JW 11": Relations between sleep stage, posture and effective nasal CPAP levels in OSA. Sleep 1992;15: 162-167. 2. Montserrat JM. Ballester E. Olivi H: Time course of stepwise CPAP titration. Am J Respir Crit Care Med 1995; 152:1854-1859. 3. Yamashiro Y, Kryger MH: CPAP titration for sleep apnea using a split night protocol. Chest 1995; 107:62-66. 4. Oksenberg A, Silverberg OS. Arons E, Radwan H: The sleep supine position has a major effect on optimal nasal continuous positive airway pressure. Chest 1999; 116: 1000-1006. 129

PATIENT 37 A 50-year-old man with problems during a CPAP titration A 50-year-old man complained of snoring and excessive daytime sleepiness. He underwent a CPAP titration after the initial diagnostic portion of his sleep study showed an AHI of 35/hr. The patient initially seemed to do well at 7 cm H20 . However, on higher pressures his AHI was higher. The next morning he complained of a dry mouth. Figure: A 90-second tracing on 9 em H20 is shown below. Of note, the leak on 7 cm H20 was 5-10 Llmin. Sleep Study (CPAP Treatment Table) DIAGNOSTIC PRESSURE 0 7 CM H2O 9 CM H2O II CMHp Monitoring time 180 min 60 40 40 TST (min) 140 min 50 30 20 REM (min) 0 30 0 0 Body position supine supine supine supine AHI (zhr) 35 4.8 24 30 Obstructive apneas 20 0 0 0 Mixed apneas 0 0 0 0 Central apneas 5 2 2 0 Hypopneas 57 2 10 10 Desaturations 65 0 2 1 Min. SaO, 85% 94% 91% 90% Resp arousal index Uhr) 40 5 30 30 TST = total sleep time; hypopneas = a reduction in flow associated with an arousal or 4% desaturation Question: Why are there more events at pressures higher than 7 cm H20 ? A B A,., fi Ji fI,.,~ 20 sec CPAP flow ,J VV v: v V~~.) chest abdomen arousal C ( ) Y~YDI"IY!n"""n~n"n~~~~~~u~~~~ A B C -----.-.'i ! ! ! , , , , , , , x , , , ,~_.~......- 4 5 3 16 14 12 14 20 19 20 21 21 31 31 12 32 31 31 12 28 19 10 l 3 13 5 3 4 90 sec Leak Sa02 FIGURE 1 130

Answer: Overtitration and increased mouth leak on pressures higher than 7 em H20 cause more events. Discussion: CPAP treatment tables are sometimes useful for choosing an appropriate pressure for treatment. However, they do not tell the entire story. It is not uncommon for the AHI to temporarily increase as CPAP pressure increases. This is usually secondary to the effects of changes in body position or sleep stage. Higher pressure is almost always needed in the supine position and may also be required during REM sleep. Another possibility is an increase in central events. Therefore, you must look not simply at the AHI, but at the composition of events. Two other reasons that the AHI might increase at higher pressures are patient intolerance (arousals) and mask or mouth leaks. Arousals can result in central apneas or hypopneas on return to sleep. High mask or mouth leak may compromise the ability of pressure to maintain airway pressure. The set pressure is not always equal to the mask pressure. Although CPAP blowers are designed to compensate for leak, the mask pressure may be lower than the set pressure if a high leak is present. Most positive-pressure devices used in the sleep laboratory provide several analog or digital outputs that can be recorded. CPAP flow is an estimate of flow and is derived from the total flow the machine delivers minus an estimate of the bias flow or leak (Fig. 2). The leak is also available to be recorded. All CPAP mask interfaces have a port for controlled leak that washes out CO 2 and prevents rebreathing. This intentional leak increases with the amount of pressure and varies between masks. The total leak is the intentional leak + unintentional leak. Unintentional leak is due to a "mask leak" (when using a full face mask) or a "mouth leak" (if using a nasal mask). Leak is seen by the machine as an inspiratory flow onto which the variation in patient flow is superimposed. A flow-sensing device, such as a pneumotachograph, in the flow generator estimates both magnitude and shape of flow. High upper airway resistance is manifested by a flattened profile; the normal profile is rounded. Some CP AP flow signals can also show snoring, although many do not have a sufficient high-frequency response. Nearly all units will provide a signal of the "machine pressure." This is the pressure at the machine and can differ slightly from the set pressure (value entered by technologist). The actual mask pressure may be somewhat lower during inhalation and sometimes slightly higher during exhalation. Some positive-pressure controllers provide a pressure transducer that can be connected to the mask by tubing. Actual mask pressure then can be recorded. Leak information is very useful for a number of reasons. First, a high leak may mean a poor mask seal. The mask pressure may be lower than the machine pressure. This can compromise the ability of pressure to maintain airway patency. High mask leaks can also cause arousal (noise or leaks into the eyes). A high leak can also mean that a mouth leak is present. The leak magnitude does not reveal the source of the leak. However, the pattern of leak can be very suggestive. A leak that fluctuates without any evidence of patient movement or increase in pressure suggests a mouth leak. Sometimes the technician can look for an open mouth with the monitoring lowlight camera. A thermistor over the mouth can also be used to detect leak. Unfortunately, a thermistor does not provide quantitative information about the size of the mouth leak. Mouth leak can be solely expiratory, or both inspiratory and expiratory. In Figure 3, the snore microphone detects vibration during exhalation. Also note that the expiratory portion of the flow signal is truncated (*) as exhalation is mostly via the mouth and not the nasal mask. In Figure 4, a thermistor placed outside the mask and over the mouth records flow mainly during exhalation in REM sleep. Although the value of the leak is high at 39-40 liters per minute (lpm), the patient slept undisturbed. (High leak> 0.4 liters per second or > 24 liters per minute.) Minutes before, in NREM sleep, the leak was 12 lpm. The CPAP flow detects flow via the nasal mask. If Flow 30lpm o Total flow CPAP flow v FIGURE 2 leak j.; 131

t inspiration Nasal CPAP = 10 em H20 CPAP flow Snore FIGURE 3 some portion of the ventilation is via the mouth, this will not register as flow. Periods of mouth breathing on CPAP can simulate a hypopnea. Usually there is no desaturation associated with such an event. Alternatively, mouth leak can actually compromise the ability to keep the airway open and result in a true hypopnea with an associated desaturation. Patients often arouse during mouth leak. Commonly, the leak dramatically decreases after arousal when the patient wakes up and closes the mouth. In the current patient, the CPAP was increased from 7 to 9 cm H20 because of some questionable events during a brief episode of REM sleep. On 7 cm H?O, the leak had been low and constant at 5-10 lpm, and the AHI was low even during supine REM sleep. On 9 cm H?O, the leak was higher and fluctuated. There were many hypopneas without desaturation (see Figure 1, B). Many of the hypopneas were associated with arousal. Again, the definition of hypopnea used here requires an arousal or a desaturation. Still looking at Figure 1, note the change in flow and leak from A to B. Flow decreases and leak increases. Leak falls back to a low value after the arousal (C). Note also that the shape of the flow is round, denoting a central hypopnea. In this case, the CPAP could have been reduced back to 7 ern H?O, a chin strap applied, or a full face mask used. The sleep technologist did apply a chin strap and start the patient on heated humidity. Heated humidity can help reduce the tendency for mouth leak in patients with nasal congestion. Unfortunately, the pressure was also increased to II cm H?O, which only worsened the situation. Although the patient had an uncomfortable night in the sleep laboratory ("they blasted my mouth open"), he was willing to try CPAP. Subsequent treatment with 7 cm Hp and heated humidity resulted in near complete resolution of symptoms. i· , CPAP flow chest abdomen Leg EMG ..... ' I .. "'I"flII111t1._"O~1 ~, i '4 , ~,.~ , ........ ," .........N , ,4 .4 t .... I: snore oral thermistor C4-Al C3-A2 02-Al ROC-A1 LOC-A2 EKG chin 5002 ·tB '19 38 "" 39 "" 'J!! 't! 'J!! '19 '19 '18 'IB 'I" OJ' ss 99 99 99 99 g! 98 9' 99 "9 98 "9 98 !!... Leak .2 .i! 19 U! ~9 ~9 13 ·12 13 ·~9 .!9 .!9 :~9 ~9 !9 ·19 "'0 '19 39 ~9 39 ·m '10 39 39 ~9 ~9 39 .!O FIGURE 4 132

Clinical Pearls l. When the AHI worsens with higher positive pressure, always consider the possibility that an increase in mouth or mask leak is causing the problem. 2. Recording leak information can be very useful for both the reviewer and the technician 3. A fluctuating pattern of leak without change in body position suggests a mouth leak. 4. Mouth dryness is an important clue that a mouth leak is OCCUlTing. 5. Mouth leak can produce a pattern of flow resembling central hypopneas. REFERENCE Berry RB: Medical therapy In Johnson JT, Gluckman JL, Sanders MH. (eds): Obstructive Sleep Apnea. London, Martin Dunniz, 2002, pp 89-1 18. 133

PATIENT 38 A patient refusing to accept treatment for obstructive sleep apnea A 40-year-old man was diagnosed with moderate obstructive sleep apnea. He came to medical attention because his wife could not stand his snoring. The patient denied daytime sleepiness (Epworth Sleepiness Scale score 8/24 [normal :s 10]). Medical history included coronary artery disease and hypertension. There was no history of nocturnal angina. Although the patient did well on nasal CPAP, he declined any treatment for OSA. He stated: "My wife is the only one with a problem." Split Sleep Study* TST (min) REM (min) Respiratory arousal index (lhr) AHI (lhr) AHINREM AHIREM Desaturations (number) Lowest Sa02 (0/0) DIAGNOSTIC PORTION 90 10 30 42.2 40 60 80 75 CPAP TITRATION 200 50 15 12.5 10 20 10 85 OPTIMAL CPAP (10 CM H2 0 ) 60 20 10 1.7 o 5 I 92 *Initial diagnostic portion. followed by CPAP titration TST = total sleep time. AHI = apnea/hypopnea index Question: What should you tell the patient? 134

Answer: Discuss the the increased risk of cardiovascular events associated with untreated OSA with patient and spouse. Discussion: There are three main reasons to treat patients with OSA (see also Patient 39). The first is to relieve symptoms of daytime sleepiness and improve the quality of life (mood and concentration). The second is to prevent the sleep disturbance endured by the bed partner and the resulting social discord. The third is to prevent long-term sequelae of untreated OSA. Nasal CPAP has been shown to improve both objective and subjective sleepiness compared to sham CPAP in a randomized controlled study.' Of note, sham CPAP did improve subjective sleepiness, but not as well as effective CPAP. Only effective CPAP improved objective measures of sleepiness. Nasal CPAP can also improve the sleep of bed partners by eliminating snoring and snorting. One study of patients with OSA and their bed partners estimated that effective nasal CPAP treatment improved the sleep efficiency of the bed partner by 13%, which is the equivalent of an extra hour of sleep nightly. The frequent arousals from sleep of the bed partner of patients with OSA has been termed the "spousal arousal syndrome." The third reason for treating OSA has been the most difficult to document with convincing evidence. Data showing morbidity of untreated OSA is confounded by the influences of obesity, hypertension, and increased age. A retrospective study by He et al suggested that patients with an untreated apnea index of > 20/hr had a decreased survival that was normalized with treatment using tracheostomy or CPAP, but not UPPP. Unfortunately, the cause of death of the patients was not documented. The Sleep Heart Health Study (SHHS) was initiated to determine the long-term impact of OSA on cardiovascular health. This study has accumulated a tremendous amount of information on a large cohort of patients using ambulatory polysomnography. Most of the cohort had mild OSA. However, even at modest AHI levels there was a moderate increase in the odds of having a stroke or congestive heart failure as the AHI increased. The evidence for an increased risk of coronary artery disease in OSA was less convincing. Follow-up cardiovascular outcomes on this cohort should provide important information. In another SHHS publication, evidence for an association of mild OSA with hypertension was presented. Patients with untreated OSA have increased sympathetic tone, and other evidence suggests OSA may contribute to atherosclerosis via a number of mechanisms. Because of the increased risk of cardiovascular complications, most clinicians recommend treatment of asymptomatic patients with moderate to severe forms of OSA. Such an approach seems especially important in patients with known cardiovascular disease. The present patient had severe OSA. The findings would likely have been even more severe if an entire diagnostic night been recorded. Even though he did well on nasal CPAP, the patient was not sleepy and did not want treatment. The long-term risks of untreated sleep apnea, including possible worsening of his coronary disease, were discussed with the patient. He was not convinced and declined nasal CPAP or consideration of other treatment options. His wife now sleeps in a separate bedroom. Clinical Pearls I. Even asymptomatic patients with moderate to severe OSA should be offered treatment. 2. Sleep disturbance of bed partners is a frequently forgotten morbidity of untreated OSA. 3. Patients with untreated OSA and even relatively low AHIs increase their risk of cardiovascular disease. REFERENCES I. He J, Kryger MH, Zorick FJ, et al: Mortality and apnea index in obstructive sleep apnea. Chest 1988; 94:9-14. 2. Beninati W, Harris CD, Herold DL, Shepard JW Jr: The effect of snoring and obstructive sleep apnea on sleep quality of bed partners. Mayo Clin Proc 1999;74:955-958. 3. Jenkinson C, Davies RJO, Mullins R, Stradling JR: Comparison of therapeutic and subtherapeutic nasal continuous positive airway pressure for obstructive sleep apnea: A randomised prospective parallel trial. Lancet 1999;353:2100-2105. 4. Nieto FJ, Young TB, Lind BK et al: Association of sleep-disordered breathing, sleep apnea, and hypertension in a large community-based study. Sleep Heart Health Study. JAMA 2000;283: 1829-1836. 5. Leung RST, Bradley TO: Sleep apnea and cardiovascular disease. State of the Art. Am J Resp Crit Care Med 2001: 164:2147- 2165. 6. Shahar E, Whitney CW, Redline S, et al, for the Sleep Heart Health Study Research Group: Sleep-disordered breathing and cardiovascular disease: Cross-sectional results of the Sleep Heart Health Study. Am J Resp Crit Care Med 200 I; 163:19-25. 13S

PATIENT 39 A 66-year-old man with mild OSA, but no symptoms A 66-year-old man was evaluated because his snoring disturbed his wife's sleep. He denied symptoms of daytime sleepiness (Epworth Sleepiness Scale score 8/24, normal). He had gained about ten pounds over the last 3 years, and his snoring and snorting had increased in severity. The patient did admit to drinking two cups of coffee to "get going in the morning." He had well-controlled hypertension, but there was no history of ischemic heart disease, heart failure, or cerebovascular disease. Physical Examination: BP 130170, HEENT: dependent palate, long edentulous uvula; neck 17- inch circumference. Sleep Study: Heavy snoring noted; total sleep time supine = 80%; respiratory arousal index = 15 fhr. The table below shows the AHI using different definitions of hypopnea. AHI Definition ofHypopnea 3fhr >30% reduction in airflow with at least a 4% arterial oxygen de saturation 8fhr >30% reduction in airflow with at least a 4% arterial oxygen desaturation or an arousal Question: The patient declined surgical and oral appliances options. Should he be treated with CPAP? 136

Answer: Nasal CPAP treatment of asymptomatic patients with mild OSA is controversial. Discussion: Treatment options for patients with mild OSA include weight loss, position restriction, a UPPP, an oral appliance, nasal CPAP, or simple observation. If surgical or oral appliance options are declined, you must decide between nasal CPAP and conservative treatment (weight loss, position restriction). A consensus conference recommended CPAP treatment for all patients with a respiratory disturbance index (RDI) 2: 30/hr and for symptomatic patients with an RDI of 5-30/hr. The RDI was defined as the number of apneas, hypopneas, and respiratory effort-related arousals (RERAs) per hour of sleep. "Symptomatic" was defined as excessive daytime sleepiness, impaired cognition, mood disorders, insomnia, or documented cardiovascular diseases to include hypertension, ischemic heart disease, or stroke. Hypopnea was defined as either a 50% decrease in airflow or less than a 50% decrease in airflow + 3% desaturation or arousal. New Medicare guidelines for CPAP reimbursement are listed in the table. Many insurance providers may adopt these guidelines. The AHI does not include RERAs, and a 4% desaturation is required for hypopneas. Note that both recommendations include symptoms other than daytime sleepiness in the "symptomatic" category. The question of CP AP treatment is difficult in patients with mild OSA who are truly asymptomatic. Most recommend CPAP treatment for patients with severe OSA (AHI > 30/hr) even if symptoms or cardiovascular disease are absent. However, a study by Barbe and coworkers found no short-term benefit (quality of life measures, blood pressure, cognitive function, objective sleepiness by MSLT) in a group of patients with an AHI > 30/hr and a mean Epworth Sleepiness Scale score of 7/24 (normal). The fact that these patients were not sleepy is probably explained by the wide variability in the susceptibility to sleep fragmentation. Moreover, some have suggested that a spouse should fill out the Epworth questionnaire, because some patients underestimate their sleepiness. As discussed in Patient 38, the other reasons for treatment of OSA besides daytime sleepiness are to alleviate socially unacceptable snoring and to prevent cardiovascular consequences ofOSA. In milder OSA one might argue that cardiovascular consequences may be less likely. However, even mild increases in AHI were associated with increased risk of cardiovascular disease in the Sleep Heart Health Study. What has not been demonstrated is that CPAP treatment prevents these consequences. For milder cases of OS A, factors such as disturbance of the bedrnate's sleep, coexistent cardiovascular disease, and patient attitudes toward the problem must be considered. Of note, patients in the Barbe study were normotensive, Nasal CPAP has been shown to be effective in symptomatic patients with mild OSA. If you try nasal CPAP in asymptomatic patients, there are some factors to consider. First, adherence (formerly called compliance) is typically lower. You might want to be conservative in the pressure prescription if pressure intolerance is a problem. Adherence might be increased if the patient understands the impact of snoring on the bedpartner or their other medical conditions. Second, some asymptomatic patients actually feel better on CPAP. However, remember that in some studies the score on the Epworth Sleepiness Scale improved on sham CPAP (placbo effect). In the present case, note that classification of the patient as a simple snorer or a patient with mild OSA (AHI 2: 5/hr) depends on the definition of hypopnea. The patient felt that having his wife Medicare Guidelinesfor Positive Pressure Reimbursement DEFINITIONS Apnea = absence of airflow for 10 seconds or longer Hypopnea = 30% or more reduction in flow from baseline with at least a 4% arterial oxygen desaturation The AHI must be based on a minimum oj2 hours ofsleep recorded by polysomnography (not a cardiopulmonary study) AHI 5-14 IHR Patient must be symptomatic (i.e., documented symptoms of excessive daytime sleepiness, impaired cognition, mood disorders, or insomnia, or documented hypertension, ischemic heart disease, or history of stroke) AHI2: 15 IHR Patient can be with or without symptoms 137

sleep with him was very important, so he agreed to try CPAP. He underwent a CPAP titration and was started on 7 em HoG and a weight-loss program. Although the patient could be considered symptomatic (documented hypertension), he did not quite meet the new Medicare guidelines of an AHI of at least 5/hr (using the Medicare definition of hypopnea). However, he was willing to pay for his own CPAP through a rental agreement. While he found CPAP difficult to use for the first 6 weeks, he eventually felt better. His Epworth Sleepiness Scale score decreased to 5, and he no longer had to drink two cups of coffee to get going in the morning. Clinical Pearls I. Nasal CPAP can be effective in symptomatic patients with mild GSA. 2. Treatment decisions in mild asymptomatic GSA patients are difficult. Consider the attitude of the patient, degree of bed partner sleep disturbance, and comorbid medical conditions. 3. Most recommend treatment of asymptomatic GSA patients with moderate to severe GSA (AHI > 30/hr [hypopneas not requiring desturation] or > 15/hr [hypopneas with 4 % desaturation)). 3. Remember that sleepiness may be underestimated (in addition to having the patient complete the Epworth questionnaire, also ask the significant other to complete one for the patient). 4. In new Medicare guidelines, "symptomatic" means more than daytime sleepiness. Coexistent cardiovascular disease qualifies the patient for being symptomatic. A lower AHI is required for CPAP reimbursement with patients who are symptomatic. REFERENCES I. Redline S, Adams N, Strauss ME, et al: Improvement of mild sleep-disordered breathing with CPAP compared to conservative therapy. Am J Respir Crit Care Med 1998; 157:858-865. 2. Engleman HM, Kingshott RN. Wraith PK. et al: Randomized placebo-controlled crossover trial of continuous positive airway pressure for mild sleep apnea/hypopnea syndrome. Am J Respir Crit Car Med 1999; 159:461-467. 3. Loube 01. Gay PG, Strohl KP. et al: Indications for positive airway pressure treatment of adult obstructive sleep apnea patients. Chest 1999; 115:863-866. 4. Barbe F. Mayorala LR. Duran J. et al: Treatment with continuous positive airway pressure is not effective in patients with sleep apnea but no daytime sleepiness. Ann Intern Med 200 1;134: 1015-1023. 5. Pack AHI, Maislin G: Who should get treated for sleep apnea? Ann Intern Med 200 I; 134: 1065-1066. 138