The first publication demonstrating the effective use of continuous positive airway pressure (CPAP) therapy in patients with OSA was in 1981 by Colin Sullivan and colleagues [2]. In their initial study, Sullivan et al. developed both a nasal interface, which was made from soft plastic tubes fit into each nostril and sealed by silicone rubber, and a “vacuum cleaner blower motor” connected by a wide-bore tube that could control the level of pressure delivered via mechanical narrowing of the free end of the tube. Five patients with generally severe OSA characterized by a mean apnea index of 62 events per hour during non-rapid eye movement (NREM) sleep and 64 events per hour during rapid eye movement (REM) sleep were fitted with a nasal interface and underwent polysomnography (PSG) with CPAP applied. The patients demonstrated complete cessation of obstructive events and marked improvement in sleep continuity, with CPAP pressures ranging from 4.5 to 10 cm H₂O. Each patient remained awake and alert for the entirety of the following day, in stark contrast to their daytime hypersomnolence prior to using CPAP.

Despite early contradictory reports of the effectiveness of CPAP therapy [3, 4], it has become the standard of care for the management of OSA. CPAP consistently shows improvements in the apnea–hypopnea index, Epworth sleepiness scale (ESS), multiple sleep latency test (MSLT), maintenance of wakefulness test (MWT), and quality of life parameters as compared with placebo or conservative management in a recent systematic review [5]. CPAP is the standard of care therapy for moderate to severe OSA and should be considered in patients with mild OSA, particularly in those with daytime symptoms and/or other significant comorbidities [6]. The generally accepted mechanism by which CPAP resolves obstructive events is by acting as a pneumatic splint, maintaining patency of the upper airway in the setting of decreased intraluminal pressure during inspiration [7].

Nonadherence to CPAP therapy is a major challenge in the management of OSA. Adherence rates have been estimated to range from 46 to 83 % when adherence is defined as CPAP use of at least 4 h per night [8]. Current recommendations to improve compliance with PAP include close follow-up, the utilization of heated humidification, and CPAP educational programs [6]. Device manufacturers and the medical community have approached the problem of CPAP nonadherence by developing alternate modes of delivering PAP (bi-level PAP, autotitrating PAP) and pressure waveform modification (C-Flex, Bi-Flex, expiratory pressure relief (PR)) . The remainder of this chapter discusses these differing modalities in an evidence-based manner in order to provide the practitioner a working knowledge of the current technology of PAP therapy.

The first description of bi-level PAP (BPAP) in 1990 [9] investigated the potential of providing differential pressures during inspiration and expiration with the hope of lowering positive pressure during the expiratory phase in order to improve comfort while still maintaining airway patency. Thirteen patients underwent a full night diagnostic PSG, a CPAP titration PSG, and BPAP titration PSG. BPAP was titrated based on the respiratory events: inspiratory PAP (IPAP) was increased for desaturation events while expiratory PAP (EPAP) was increased for apneic events. CPAP and BPAP were equally effective in resolving OSA. Final mean pressures were 14 cm H₂O for CPAP and 14/8.9 cm H₂O (IPAP/EPAP) for BPAP, confirming that despite a lower mean airway pressure, obstructive events were prevented. Some individuals required oxygen supplementation during the CPAP study, while the same individuals did not with BPAP study.

In randomized controlled trials evaluating CPAP versus BPAP in newly diagnosed patients with OSA, BPAP was not found to improve compliance with therapy. Reeves-Hoche and colleagues randomized patients with OSA (apnea-hypopnea index (AHI) ≥ 10 events/h) to 12 months of CPAP or BPAP therapy following in-laboratory PAP titration [10]. BPAP titration differed from the study by Sanders et al. in that CPAP was initially increased to resolve apneas followed by an independent increase in IPAP to eliminate hypopneas and snoring. There was no difference in the mean machine timer hours over the 12-month period. Of note, dropout was higher in the group randomized to CPAP. In a second study, 27 PAP naïve patients were randomized to CPAP or BPAP with Bi-Flex (Respironics) after separate titration nights with each device and followed for 1 month [11]. There were no observed differences in compliance, and the improvement in the ESS and the functional outcomes of sleep questionnaire (FOSQ) were similar between the groups. Because of these findings, BPAP is not recommended as a first-line PAP therapy for OSA, although it could be considered as an optional therapy in patients with CPAP intolerance due to high pressures [6]. The use of BPAP in patients with coexisting hypoventilation syndromes is not included in this review or in these recommendations.

Although PAP naïve OSA patients do not demonstrate improved usage patterns with BPAP when compared with CPAP use, failure of CPAP therapy, either during the titration study night or on follow-up after PAP has been used in the home, may be an indication to consider BPAP therapy. In a two-phase intervention trial of OSA patients with poor CPAP compliance, Ballard and colleagues demonstrated that compliance may be improved with BPAP therapy [12]. Of 204 patients with OSA, 24 % became compliant (≥ 4 h of nightly CPAP use) after phase 1 of the trial, which consisted of educational measures, mask adjustment or replacement, therapy for nasal symptoms, and initiation of heated humidification. If these measures failed to increase CPAP adherence , patients underwent a “split-night” titration trial with 4 h of CPAP and 4 h of Bi-Flex (BPAP with pressure modification, to be discussed later). The patients were then randomized to home CPAP or Bi-Flex therapy for 3 months. The Bi-Flex group demonstrated greater compliance (49 % vs. 28 %; p = 0.03) and a higher increase in mean FOSQ scores from baseline compared with CPAP treatment.

Based on these findings, routine administration of BPAP after a diagnosis of OSA does not improve treatment adherence and hence, cannot be recommended. However, if a patient demonstrates treatment noncompliance with CPAP despite behavioral and medical optimization (education, humidification, mask fitting, nasal steroids for rhinitis, etc.), second-line therapy, of which BPAP is an option, should be considered. The choice of which second-line modality (BPAP, oral appliance, surgery) to pursue depends on the patient characteristics and preferences. If BPAP therapy is pursued, a formal in-laboratory titration should be performed to ensure pressure optimization and improved patient comfort. Of note, differences in BPAP titration techniques and pressure modifications (Bi-Flex vs. BPAP without pressure modification) may result in variable outcomes and different rates of improvement in adherence.

The purported utility of BPAP therapy for OSA was to decrease expiratory pressure, and therefore mean airway pressure, against which an individual must breathe, thereby increasing comfort with the device and subsequent improvement in compliance. Another strategy for improved compliance that was and is still being explored relies on varying the delivered pressure throughout sleep via device-specific algorithms, which fluctuates between higher and lower pressures to relieve flow limitation and obstruction, as changes in position and sleep stage may require. The characterization of flow limitation was a crucial first step in this process [13]. Condos and colleagues demonstrated that upper airways resistance is increased when the contour of the flow-time curve is flattened. Automated analysis of the flow-time curves could then be used to guide computer-generated adjustment of pressure in the setting of obstructive respiratory events. Current autotitrating positive airway pressure (APAP) technology relies on the determination of obstructive events characterized by flow limitation (flattening of the flow-time curve), apneas (no or markedly reduced flow over time), and snore (measured via vibration sensors). In addition, current devices use the forced oscillation technique (FOT) [14] to help determine whether reduction in flow is obstructive (i.e., non-patent airway) or nonobstructive. If events are obstructive, delivered pressure will be increased, while raising pressures during non-obstructive events (central apneas or hypopneas) can be avoided. The FOT is a measure of mechanical impedance of the respiratory system that is determined by applying high frequency, low amplitude pressure waves via the mask interface; patterns of FOT help determine if the airway is open (patent) or closed (obstructed) [15].

APAP and CPAP demonstrate equal efficacy in relieving obstructive respiratory events and improving daytime functional parameters. In the first randomized controlled trial of APAP versus CPAP, 16 patients with OSA were randomized to fixed CPAP with pressure determined from full PSG (P_eff) or APAP (minimum pressure—(P_eff − 2 cm H₂O), maximum pressure—(P_eff + 4 cm H₂O)) [16]. Sleep and respiratory parameters were similar with CPAP and APAP therapy (AHI, total sleep time, sleep architecture). Following 3 weeks of PAP therapy, patients in both groups demonstrated equal improvement in ESS, MWT, and the trail making test (a test of cognitive impairment) compared with baseline; no differences between the groups were detected. Subjects randomized to APAP spent 49 % of the time requiring pressure below P_eff, suggesting that APAP use at home may deliver pressures lower than that determined during a titration study for a significant portion of the night and thereby improve tolerance of PAP. Along these lines, another study of 15 patients with generally severe OSA demonstrated a positive correlation between the polysomnographically determined P_eff and the time spent below P_eff on APAP [17]; patients with more severe OSA may benefit to a greater extent with APAP due to spending a greater duration of the night at pressure levels below P_eff.

Although there is some evidence that APAP improves compliance compared with fixed pressure CPAP, the evidence depends on the study type and the improvement is of questionable clinical importance. A Cochrane Collaboration review demonstrated a statistical improvement of 0.21 h of machine usage time in crossover studies comparing APAP to fixed CPAP [18]. Although statistically significant, an overall increase of 13 min per night is not likely to result in improvements in clinically meaningful outcomes. Crossover trials generally show equivalent residual AHI between APAP and fixed CPAP [19–23], equal improvement in ESS scores from baseline [19–21, 24–26], and a reduction in delivered pressure with APAP treatment [20, 22, 23]. Although there was a general patient preference for APAP treatment [21, 22], this was not a universal finding with some patients preferring fixed CPAP [20]. Similarly, improvement in machine usage was not a universal finding; among parallel studies or the first arm of crossover studies, a statistically significant difference in machine usage was not seen (mean increase in machine usage with APAP—0.22 h (− 0.05, 0.49; 95 % CI)) [18].

It is clear from the above discussion that APAP compared with CPAP therapy in an unselected patient population showed minimal to no improvement in treatment adherence and conflicting data on patient preference. However, several groups observed APAP treatment in patients with CPAP intolerance or those with high pressure variability to determine if these subgroups of OSA patients may benefit from APAP therapy. In one study, 27 patients with an AHI ≥ 20 and a high variability index, which was a calculated value based on the variability of pressures required during a 2-week run-in period on APAP, were randomized to 8 weeks of CPAP or APAP followed by crossover to the other treatment [27]. Machine usage time was not significantly different between the groups; however, mean pressure and ESS scores were lower in the APAP group and a greater number of participants chose APAP over CPAP when simply asked which device they preferred. Another study looked at a mixed population of patients in which CPAP was not the optimal treatment, as defined by patients with CPAP intolerance, CPAP requirements ≥ 12 cm H₂O, or central respiratory events accounting for ≥ 10 % of events [22]. Patients were randomized in a crossover fashion to either BPAP or APAP for 6 weeks each. Both modalities significantly improved respiratory parameters (although mean residual AHI while on PAP in each group was > 5) and sleep quality. Patients again preferred APAP treatment as a long-term therapy for their sleep-disordered breathing . Although the heterogeneity of this study population prevents any firm conclusions, the search for subpopulations that may benefit from APAP therapy, as illustrated in these two studies, is an important strategy in improving patient’s acceptance of PAP therapy, thereby improving overall outcomes.

APAP devices rely on brand-specific proprietary algorithms that use a variety of methods for detecting respiratory events and adjust pressure settings differentially. An early trial evaluated devices guided by “flow limitation (APAP_fl)” versus “FOT (APAP_fot)” for determining the presence of respiratory events [28]. In a group of 30 patients with moderate to severe OSA (AHI ≥ 20 events per hour) treated with each device in a “split-night” setting (4 h/device), there was a trend towards lower AHI in the APAP_fl group, higher overall pressure in the APAP_fl group, and greater pressure variability in the APAP_fot group. Although based on this study it is unclear which detection system is more accurate in identifying obstructive respiratory events, most current devices use a combination of flow limitation and forced oscillation in order to identify both obstructive and non-obstructive events. Several studies have compared devices produced by competing manufacturers [29, 30]. It is apparent from these studies that there is variability in delivered pressure and respiratory parameters between the various devices but the clinical significance is unclear.

Although the concept of APAP was well conceived in attempting to lower average positive pressure and improving patient compliance with PAP devices, the evidence does not support this. Overall usage of APAP was slightly increased when all studies are taken into account, although these results were highly dependent on the type of study and the patient population studied. In an attempt to combine other technology with APAP , two recent studies looked at APAP with expiratory PR (which will be discussed in the next section) to further improve patient comfort [31, 32]. When compared with fixed CPAP, APAP with expiratory PR (see the following section) was essentially equivalent with respect to sleep/respiratory parameters and outcomes. Consistent with prior studies, the patients in one study were satisfied with both regimens but preferred APAP treatment [32], while in the other study, patients were less satisfied and felt they derived less benefit with APAP therapy [31]. Without clear and convincing evidence supporting the superiority of APAP over CPAP, this treatment modality cannot be recommended at this time except on an individual basis.

Many device manufacturers have developed technologies to decrease applied positive pressure during the expiratory phase in order to improve compliance with the device by encouraging more physiologic transitions between inspiration and expiration. Brand-specific algorithms differ in the degree and timing of such Pressure Relief (PR) [33, 34]. The pressure decrement is either flow based (pressure drop proportional to patient’s flow amplitude) or prespecified (i.e., 1 cm H₂O, 2 cm H₂O or 3 cm H₂O) . Likewise, the pressure decrement is either in the early portion of the expiratory phase (returning to set pressure prior to inhalation) or throughout the entire expiratory phase. PR may be utilized in most modes of PAP, including CPAP, APAP, and BPAP. Although when in CPAP or APAP mode PR appears to be very similar to a bi-level mode of ventilation, the main difference is that the pressure differential between IPAP and EPAP is usually lower than 3 cm H₂O in PR mode and PR can be set in a stepwise manner (increase settings for more PR) as compared with the completely independently set IPAP and EPAP of BPAP therapy.