“Pressure-relief” CPAP and bilevel PAP have recently been introduced [34, 35]. These provide expiratory pressure relief that is proportional to expiratory airflow, while pressure increases rising to the prescribed EPAP level at end-expiration. In expiratory pressure relief, the pressure-relief bilevel PAP device provides relief of pressure at end-inspiration [34]. Plausibly, this is facilitated by the high lung volume at end-inspiration that contributes to upper airway patency as well as the fact that, as inspiration reaches its end, airflow diminishes with consequent reduction in collapse-promoting negative intrapharyngeal pressure. Pressure-relief CPAP was as effective in reducing sleep-disordered breathing events and improving sleep continuity as conventional CPAP [35]. Gay and coworkers [34] demonstrated comparability between pressure-relief and convention bilevel PAP in alleviating OSA; however, there are too few studies of these modalities on which to base conclusions.

Although CPAP and bilevel PAP therapy may be associated with improvement in Cheyne–Stokes breathing (CSB) and central sleep apnea, sometimes there is no improvement or worsening [36, 37]. Adaptive servo-ventilation (ASV) has introduced into clinical practice primarily to address CSB in the context of heart failure as well as idiopathic and complex central sleep apnea or treatment-emergent central sleep apnea (central sleep apnea that becomes clinically problematic during application of conventional CPAP or bilevel PAP during treatment of OSA). In general, these devices provide a variable degree of inspiratory pressure support that stabilizes the patient’s ventilation over time. Thus, if there is a reduction in the patient’s spontaneous ventilation, the degree of inspiratory positive pressure support will increase proportionately to minimize a decrement in ventilation [38]. A backup rate of IPAP delivery may also be set by the clinician.

Teschler et al. [38] evaluated the effects of supplemental oxygen, CPAP, bilevel PAP, and ASV in patients with CSB (with primarily central sleep apnea) due to heart failure and reported that ASV provided the greatest reduction in AHI, slightly but statistically significantly lower than on bilevel PAP. Sleep continuity, sleep efficiency, percent slow-wave sleep, and percent rapid eye movement (REM) sleep were comparable on ASV and bilevel PAP. The ASV settings in this investigation included a backup rate of 15 breaths/min, and the backup rate on the bilevel PAP was set at the patient’s awake spontaneous breathing rate, less 2 breaths/min (for the study group, the range of backup rate was 13–18 breaths/min). It is not possible to determine how much improvement in CSB was specifically due to the ASV or bilevel PAP algorithm as opposed to the presence of a backup rate on both modes in this study. A subsequent case series by Banno et al. [36] reported improved CSB on ASV that had failed to improve on CPAP. These authors indicated that they employed the same default settings as Teschler et al. [38], and therefore a backup rate may also have been used.

A small case series of heart failure patients with CSB who failed CPAP and bilevel PAP (with a backup rate) reported a beneficial effect of ASV [39]. The ASV employed in this study included a default backup mode. A randomized 4-week crossover trial of ASV prescribed at therapeutic settings (including a backup rate) versus ASV prescribed at subtherapeutic settings (with a backup rate of 15 breaths/min) was performed on patients with heart failure and CSB [40]. The results demonstrated significantly greater improvement of the AHI during the intervention with ASV at therapeutic settings. Although the active intervention resulted in a substantial reduction in objectively assessed sleepiness compared with subtherapeutic ASV, there were no statistically significant changes in the secondary outcome measures of the study, including subjective sleepiness, questionnaire assessment of health status, and performance using a driving simulator with either therapeutic ASV or subtherapeutic ASV. However, plasma brain natriuretic peptide and urinary metanephrine excretion fell on the active intervention, suggesting a favorable physiologic impact of ASV. These results suggest that ASV results in physiologic improvements, but the absence of subjective improvement in sleepiness is disappointing.

Despite initial enthusiasm for ASV therapy in heart failure patients, the use of this mode has been questioned recently on the basis of a randomized examining the role of ASV in central sleep apnea with systolic heart failure, the SERVE-HF trial. In a study with 1325 subjects with a left ventricular ejection fraction of 45 % or less and clinical heart failure, there was no difference in AHI between the ASV group and the treatment-as-usual control group at 12 months after starting ASV therapy. Furthermore, there was a statistically significant increase in mortality in the ASV therapy group compared to the control group (hazard ratio for ASV was 1.28 (90 % CI 1.06–1.55). This study result has resulted in a questioning of the role of this mode of therapy in systolic heart failure. Further studies being conducted presently may clarify the results of SERVE-HF.

The general features of various PAP modalities are summarized in Table 34.1. It is important to note that algorithms and modes of operation vary across manufacturers and models. This table is not intended to represent definitive features of specific brands and models but rather to describe general features.

There is abundant evidence that, when applied in sufficient pressure, PAP effectively eliminates OSA events as well as respiratory effort–related arousals. Following initiation of CPAP therapy during sleep, most OSA patients with daytime sleepiness at baseline report increased subjective alertness [41–43]. There is, however, variability across studies. In a meta-analysis of randomized trials, Marshall and coworkers [44] reported that, after controlling for placebo effects, the Epworth Sleepiness Scale score [45] increased significantly, but only by 1.2 points in patients with mild to moderate OSA. The effect of CPAP on objective metrics of sleep propensity during the day (e.g., Multiple Sleep Latency Test [MSLT] or Maintenance of Wakefulness Test) is less clear, with only some studies showing an effect and a less compelling impact in patients with mild OSA [42–47]. The meta-analysis by Marshall et al. examined the ability to remain awake under soporific conditions, assessed by the Maintenance of Wakefulness Test and observed that, over the three randomized trials in which this assessment was performed, sleep latency increased significantly but only by 2.1 min, and there was no significant change in the sleep latency during the MSLT in the four trials in which it was assessed [44]. The investigators called into question the clinical significance of their changes. Another meta-analysis [48] reported a significant reduction in the Epworth Sleepiness Scale score by an average of approximately 3 points, with relatively greater reductions in patients with severe OSA. These investigators also observed only a marginal improvement in the MSLT after introduction of CPAP therapy.

In the context of vigilance and alertness, a critically important functional outcome to examine is the effect of OSA and subsequent therapy on motor vehicle crashes [49]. In this regard, a number of studies employing driving simulators have demonstrated improved performance following initiation of CPAP therapy [50–53]. Similarly, a comparison of the number of accidents per driver per year over the 3 years before and following CPAP therapy in OSA patients demonstrated a notable reduction, reaching levels that were comparable to those in individuals without OSA [54] (Fig. 34.2).

Turkington et al. [52] reported that benefits on driving performance, assessed using a simulator, may be evident within 7 days of initiating therapy, while, more recently, Orth et al. observed improvement after only 2 days [53]. This is consistent with earlier data indicating that there may be a reduction in subjective daytime sleepiness after just 1 night of nasal CPAP therapy [55], and that progressive reduction in objective daytime alertness (assessed by the MSLT) may occur over 2 weeks following initiation of therapy [56]. The progressive nature of the improvement in symptoms and the apparent variability in response, as reflected in the previous discussion of the meta-analyses of CPAP effectiveness, highlights the importance of recognizing that patients may not be sufficiently alert to resume full activities (especially those that require vigilance, such as operating vehicles or potentially dangerous tasks) within the first several days of treatment. In addition, although studies have concluded that driving simulators may provide insight into on-the-road driving performance, differences do exist [53, 57]. It is important for clinicians to recognize that in general, studies examining treatment effect on motor vehicle crashes have been uncontrolled. To better assess the improvement in symptoms of daytime sleepiness, it is essential that the healthcare team make follow-up contact with the patient soon after the initiation of therapy. Close follow-up will facilitate evaluation of the patient’s ability to perform activities that require alertness as well as to assess therapeutic adherence (see discussion below).

PAP often has a beneficial impact on other symptoms of OSA. Studies have reported relief of tiredness, reduced snoring, decreased nocturnal awakenings with perceived choking or gasping, and a reduction in nocturia [46, 58–64]. It is appropriate to note at this point, however, that nocturia—at least in conjunction with benign prostatic hypertrophy—has been reported to negatively influence adherence to PAP therapy [65], probably due to the inconvenience of removing and replacing the interface. It may be possible to extrapolate this finding to OSA patients with nocturia that is unassociated with prostatic hypertrophy. In light of the potentially beneficial effect of PAP use on nocturia, patients should be counseled to stick with the PAP therapy to achieve a favorable outcome, and the etiology of residual nocturia should be investigated and treated.

The impact of PAP on quality-of-life measures and neurocognition has also been assessed. Engleman and co-investigators reported the results of a placebo-controlled trial of CPAP in patients with mild OSA; use of CPAP for >2.5 h/night was associated with improved visual-motor skill, social function, and vitality (using the Medical Outcomes Short Form-36 [SF-36]). In addition, the Hospital Anxiety and Depression Scale Depression Score was reduced. In a non-placebo-controlled study comparing “conservative” management and CPAP for moderate to severe OSA, McFayden et al. [66] examined the results of the disease-specific Functional Outcomes of Sleep Questionnaire (FOSQ) as well as the SF-36 and found that CPAP was associated with improved psychosocial function and the patient’s (but not the spouse’s) marital satisfaction. Conversely, Barnes et al. found no effect on neurobehavioral function or quality-of-life metrics on the SF-36 or the FOSQ [46]. These investigators speculated that differences in the study population, particularly related to gender, may explain the disparate results. Studies of patients with moderate to severe OSA (e.g., the most symptomatic individuals prior to treatment) indicate the largest effect sizes are in the contexts of sleepiness and vitality. Studies have suggested that patients with moderately impaired pretreatment cognitive performance experience modest improvements [42, 43, 67]. Moreover, data suggest that daily use of PAP for at least 6 h over 3 months is associated with clinically relevant improvement in verbal memory in some, but not all, OSA patients [68]. The most definitive data to date concerning neurocognition in sleep apnea comes from the Apnea Positive Pressure Long-Term Efficacy Study (APPLES) in which 1204 adults with OSA were randomly assigned to effective CPAP vs. sham CPAP and had neurocognitive testing performed. After adjusting for confounders such as baseline education level, gender and ethnicity, the investigators found no significant relationship between OSA (as quantified by AHI) and neurocognitive performance. The small association between OSA and neurocognition that was detected was largely associated with degree of hypoxemia [69].

Like most treatment interventions, PAP is often associated with a variety of generally minor, but troublesome side effects (Table 34.2) [16, 19, 25, 45–50, 58, 61, 70–75]. Side effects may be attributable to either the patient-device interface or the sensation of high airflow or pressure. Some patients simply perceive the lifestyle and other challenges associated with nasal PAP to be unacceptable [3, 46, 48, 61, 73, 76–78]. Such individuals are often, but not invariably, younger patients who are unable to envision indefinite nasal PAP therapy. Although some studies have concluded that side effects do not impact on adherence to therapy [61, 74], others have concluded the converse, citing side effects as a reason for nonadherence [79].

The most significant disadvantage to PAP therapy is that patients must actively participate in their own treatment both in terms of the number of nights used as well as the number of hours used per night. Although the optimal duration of nightly PAP use—or for that matter, the minimal amount of use that confers benefit—are unknown, recent studies suggest that use of at least 6 h/night confers greater cardiovascular mortality risk reduction compared with fewer hours of use per night [118] (Fig. 34.3). In addition, PAP use for >6 h/night over 3 months has been associated with a greater likelihood of normalized memory performance [68]. Viewed from the opposite perspective, sleeping for as little as 1 night without CPAP is associated with increased sleepiness [55, 119]. These data highlight the need for clinicians to facilitate maximal PAP use by patients.

Before adherence to a therapeutic PAP prescription can be considered, the patient must accept the opportunity to receive this therapy. The acceptance rate of CPAP varies across a number of studies, ranging from 62 to 92 % [71, 72, 120–125]. Among the reasons for nonacceptance are difficulty falling asleep, frequent nocturnal awakenings, and mask discomfort. In addition, those who accept CPAP therapy are generally more likely to complain of greater tiredness as well as episodes of falling asleep at undesirable times [121].

In recent years, there has been increasing interest in “split-night” polysomnography in which the initial portion of the night is spent in performing a diagnostic evaluation for OSA and the remainder of the night is devoted to establishing a PAP prescription [104, 124–131]. If the diagnosis of OSA is established during the initial portion of the night, a therapeutic titration of CPAP is undertaken. The impact of this paradigm has been explored and, in general, split-night studies provide an acceptable strategy for laboratory evaluation and initial PAP prescription without negatively impacting acceptance and adherence in patients with severe OSA [124, 125, 130–132]. Some data suggest, however, that long-term adherence may not be as high in patients with mild to moderate OSA as in patients with more severe OSA, thereby mandating particularly close follow-up in the former group [130]. Current recommendations of the American Academy of Sleep Medicine include that a CPAP titration may be conducted after observing an AHI of 40 (or AHIs of 20–40, based on clinical judgment) over at least 2 h of diagnostic polysomnography, and a CPAP prescription may be established based on a titration over at least 3 h of sleep that documents that CPAP eliminates or nearly eliminates the respiratory events during non-REM and REM sleep (including REM sleep in the supine position) [132]. Although a number of investigations have included bilevel PAP devices in assessing acceptance as well as adherence to therapy following split-night studies, neither these devices nor APAP modalities have been the specific subject of these studies.

Once a patient has “accepted” CPAP therapy, he or she must be adherent to it. In the last several years, investigators and clinicians have been able to objectively monitor daily use and patterns of use over time with meters and software that have been incorporated into the PAP devices. Such objective metrics are particularly important in assisting clinicians with management since subjective patient reports overestimate time of use [61, 71, 73, 123]. Objective adherence monitoring is now the standard of care, with its utility highlighted by the observation that suboptimal patterns of usage are established shortly after initiation of therapy, so that the clinician needs to recognize and address the contributory factors early on [76–78, 80, 96]. Moreover, information regarding the degree to which a patient is adherent to PAP is essential for assessment of a suboptimal clinical response. If a patient’s symptoms are inadequately resolved after the initiation of PAP treatment, possible reasons other than poor adherence include delivery of insufficient pressure to maintain upper airway patency during sleep (perhaps due to an incorrect prescription or because of technical issues such as air leaks through the mouth or skin-mask interface that impair delivery of the prescribed pressure), misdiagnosis of the etiology of the individual’s symptoms, the contribution of comorbid elements to the patient’s symptoms, or failure to use the device for a sufficient duration on a regular basis. Currently available software within PAP devices provides information regarding delivered pressure and the magnitude of leaks, as well as an estimate of the AHI on PAP therapy. This constellation of information can provide the clinician with important management insights.

In general, utilization of PAP ranges from 4 to 6 h/day, with considerable interindividual variability and a measurable proportion of patients with <2 h/night or complete nonadherence. As discussed earlier, >6 h of use per night is associated with reduced cardiovascular mortality and increased likelihood of normalizing memory function [68, 118]. The relative value of <6 h of use per night is unclear, as if there is a threshold of nightly use above which no further benefit is obtained in this regard as well as with respect to cardiovascular and cerebrovascular morbidities. Existing data suggests that optimal relief of daytime sleepiness requires nightly use of PAP. As also noted earlier, sleeping for as little as one night without PAP results in increased sleepiness [55, 119]—but is it necessary to use PAP during all sleep time? Hers et al. observed persistent benefit in oxyhemoglobin saturation and sleep continuity for the remainder of the night after CPAP was discontinued following 4 h of use [133]. The investigators postulated that the persistent improvement is related to greater sleep continuity while on CPAP, with increased upper airway stability during the latter portion of the night after CPAP was removed. This hypothesis is based on earlier data demonstrating increased upper airway collapsibility following a period of sleep fragmentation [134]. These investigators as well as others [135] speculated that duration of nightly use of CPAP by at least some OSA patients is determined by their perception of the amount of use required to obtain a satisfactory degree of symptomatic benefit.

It is intuitively evident that greater insight regarding the determinants of adherence would facilitate treatment modifications that would promote more universal and optimal utilization by patients. Unfortunately, our understanding remains incomplete. Some reports suggest that adherence improves as the patient’s perception of sleep propensity increases [71, 79, 136, 137]. Notably, the patient’s perception of daytime sleepiness, assessed using specific questionnaires such as the “hypersomnia score” [71] or the Epworth Sleepiness Scale [45], predicts adherence with CPAP more reliably than the MSLT, which is an objective measure of sleepiness [58, 73, 138]. Some studies have noted that, after adjusting for confounding factors, lower AHI is an independent risk factor for nonadherence [79, 136]. Conversely, several investigators have observed that adherent patients cannot consistently be differentiated from nonadherent patients by the frequency or variety of side effects of CPAP therapy, initial AHI, gender, weight, or the prescribed level of CPAP [58, 73, 75, 123, 138–140]. More recently, the contribution of “self-efficacy,” including the patient’s perception of the consequences of untreated OSA and expectations of treatment outcome [76], as well as psychologic factors such as coping strategies and willingness to modify behavior [77, 78], have been examined in the context of adherence. Stepnowski et al. [77] observed that adherence to CPAP was uninfluenced by baseline depression, stress, or anxiety but was significantly related to the patient’s score on a Ways of Coping Questionnaire (Fig. 34.4). Additional important contributions to adherence include the response to therapy with regard to sleepiness, performance and mood, home/family environment/support, encumbrance on lifestyle (e.g., ease of travel, intimacy), and interface comfort.

As discussed earlier, patients commonly experience side effects in conjunction with PAP therapy. Although evidence regarding the impact of side effects on acceptance and adherence varies, it is reasonable and prudent for the clinician to make every effort to minimize if not eliminate them (see Table 34.2). It is likely that the perception of a given side effect varies across individuals and therefore may have a different degree of impact. Thus, a simple comparison of prevalence in compliant and noncompliant patients may be misleading and obscure the impact of a particular side effect on the compliance of individual patients. Several practices may enhance patient acceptance and compliance with CPAP therapy. These are outlined in Table 34.2 as well as later in this chapter. Appropriate patient selection for chronic PAP therapy is an important factor, and educating the patient, utilizing discussion and educational literature addressing the nature of OSA, the consequences of untreated OSA, and a detailed discussion of therapeutic options and implications, is essential [141]. It is intuitively obvious that patients who require “arm-twisting” to take a PAP unit home, even after efforts have been made to explain the need for the device and the manner in which it operates, are unlikely to use it conscientiously on a long-term basis. Therefore, PAP should be provided only to patients who are reasonably receptive to using it or who are sufficiently open minded to give it a reasonable home trial.

Traditionally, patients have undergone an attended (by a technologist), monitored (by polysomnography) trial of PAP to establish therapeutic levels of pressure prior to initiating long-term therapy. Because the requisite level of PAP may vary according to body position and sleep stage, clinicians should be certain that the delivered pressure is effective in maintaining adequate upper airway patency and oxygenation during sleep in all positions, including and especially the supine position (and when possible during supine REM sleep). It also provides an opportunity for the patient to examine the PAP unit and various interfaces before home use. The most comfortable and leak-free interface with the device (i.e., nasal mask, prongs, or oral-nasal mask) can be selected. Then, while the patient is still awake, he or she may be provided with an opportunity to experience PAP across a wide range of pressures, to permit familiarization with the associated sensations. Another advantage of attended evaluation of the patient on PAP is the immediate availability of knowledgeable and caring healthcare professionals who can respond to questions and allay concerns. When wearing a PAP device for the first time, patients have been anecdotally reported to awaken in the middle of the night disoriented and perhaps frightened by the apparatus. Under these circumstances, albeit rare, reassurance is readily supplied by the laboratory personnel conducting the trial.

Several other benefits have also been attributed to polysomnographically monitored trials of CPAP therapy. Fry et al. [151] observed an increased frequency of periodic leg movements in sleep (PLMS), with and without accompanying arousals during nasal CPAP therapy. These investigators hypothesized that the improved sleep quality and architecture associated with relief of OSA by nasal CPAP “unmasks” PLMS. However, several subsequent studies observed that the Periodic Limb Movement Index was not appreciably different during a diagnostic polysomnogram and during CPAP therapy [152, 153], although arousals may diminish in conjunction with PLMS during CPAP therapy, at least acutely [152]. Nonetheless, periodic limb movement disorder may coexist with OSA and may persist after alleviation of OSA on CPAP, and a patient may not obtain symptomatic abatement of daytime sleepiness or fatigue. Thus, a monitored initial trial of CPAP addresses many issues and concerns that, if not considered, may lead to dismissal of this form of therapy as a viable therapeutic option. Attention to these factors at the outset of therapy will maximize the opportunity for a successful outcome.