Barriers to positive airway pressure adherence among veterans with sleep apnea: a mixed methods study

Abstract

Rates of adherence to positive airway pressure (PAP) for sleep apnea are suboptimal. Though previous studies have identified individual factors associated with PAP nonadherence, few projects have investigated a wide range of possible barriers directly from the patient perspective. We examined the range of factors that patients identify as barriers to optimal use of PAP as well as the solutions most commonly offered by providers. We employed a mixed methods design including semistructured interviews and medical record review at a US Department of Veterans Affairs Medical Center. Thirty patients identified as nonadherent to PAP participated. Patients were asked to report on current sleep problems, reasons for nonadherence, and solutions proposed by providers. Chart review was used to identify untreated apnea severity, Epworth Sleepiness Scale score, therapy hours, and residual apnea severity. Patients described physical and psychological barriers to adherent use at approximately equal rates: Mask leaks and dry throat/nose were common physical barriers, and anxiety/claustrophobia and worsening insomnia were common psychological barriers. Untreated apnea severity, residual apnea severity, and daytime sleepiness were not associated with therapy hours. Solutions offered by providers most frequently addressed physical barriers, and solutions to psychological barriers were rarely proposed. The most common solution offered by providers was trying different masks. We recommend individualized assessment of each patient’s barriers to use as well as increased involvement of behavioral health providers in sleep medicine clinics.

Implications.

Practice: Increased involvement of behavioral health providers and greater attention to psychological barriers to PAP use may increase PAP adherence.

Policy: Effective Sleep Medicine programs should prioritize integration of behavioral sleep medicine providers who are trained to address common barriers to PAP adherence, such as insomnia and claustrophobia.

Research: Future research should investigate the efficacy of behavioral programs designed to increase PAP adherence, including Motivational Interviewing and PAP desensitization.

Introduction

The prevalence of obstructive sleep apnea (OSA) in the U.S. population is approximately 34% in men 30–70 years old, and 17% in women 30–70 years old, though significantly more U.S. adults are considered to be at risk of having OSA based on known risk factors of age, gender, comorbidities, and body mass index [1–4]. OSA is associated with reduced quality of life, increased cardiovascular risk, and mortality [5, 6].

Though alternative treatments for sleep apnea are evolving, positive airway pressure (PAP) remains the most common treatment for OSA and is very effective when patients are adherent [7, 8]. PAP use is also associated with an improvement in daytime sleepiness and function, medical and psychological comorbidities, and mortality [9–11] particularly when underlying OSA is severe [9, 10]. Despite the potential benefits, adherence to PAP therapy remains challenging, with nonadherence rates of up to 50% within the first week of treatment [12]. Of those who do use PAP during the first week, 12%–25% stop using therapy by year three [12].

Lack of adherence to PAP is of particular concern for military veterans and the providers who treat them, as there has been a rapid increase in the diagnosis of sleep apnea in this population [13]. Sleep apnea is the most common sleep disorder among veterans enrolled in VA healthcare [14]. Increased diagnosis and treatment demand for OSA among veterans are likely to only increase in the coming years; clinical encounters for OSA care in U.S. military service members increased by 517% from 2005 to 2014 [15]. Compared with veterans without OSA, veterans with OSA are at higher risk of mortality, incident coronary heart disease, strokes and chronic kidney disease [16].

Several studies have examined factors influencing PAP adherence, including insomnia, inconsistent pretreatment bedtime, level of daytime sleepiness, claustrophobia, mask type, and cognitive factors [17–23]. However, most of these studies focus on a single variable rather than considering the multitude of factors that collectively contribute to the outcome of PAP nonadherence. A lack of patient-centered research investigating barriers to PAP adherence has prevented the development of relevant strategies to improve the quality of sleep apnea care; understanding the relative importance of several potential barriers to PAP use may help providers more readily identify common challenges when problem-solving PAP adherence and address these challenges with greater knowledge. The purpose of the current study was to examine the range of factors that military veterans identify as barriers to optimal PAP use, as well as to evaluate the most commonly suggested solutions providers recommend to improve PAP adherence. We employed a mixed methods design to maximize data capture of barriers and allow us to compare response styles based on closed- versus open-ended assessment strategies. Based on factors identified in the research literature as well as during clinical practice, data were collected from participants’ medical records as well as a self-report checklist of barriers. Qualitative responses ensured there were not any barriers missed and allowed participants to elaborate on any issues that were not captured through quantitative data. Together these data sources allowed us to more comprehensively understand barriers to PAP use in a way that would not otherwise be possible.

Methods

Thirty patients at a US Department of Veterans Affairs Medical Center were recruited to participate in a semistructured interview about their sleep and PAP experience. All participants were diagnosed with OSA, had been prescribed PAP for at least 30 days, and were identified as PAP nonadherent (i.e., less than 4 hr nightly use 70% of days during the clinically relevant period of interest, as per Medicare standard [24]) by their sleep clinic provider or were contacted after missing a PAP follow-up appointment and self-identified as PAP nonadherent. Patients identified by their provider were referred to the interviewer via warm hand-off at the end of the in-person PAP clinic appointment or agreed to be contacted by the interviewer later that week if a warm hand-off was not convenient. This study was approved as a nonresearch quality improvement project and therefore was exempt from local institutional review board review [25].

Interviews were conducted individually by the first author—a psychology postdoctoral fellow (L.A.G.)—using an original, semistructured interview instrument, developed and piloted in collaboration with (San Francisco Veterans Affairs Health Care System) providers familiar with OSA (see Supplementary Material). Items were generated from this fellow’s direct observation of patient encounters with multiple staff (e.g., physicians, nurse practitioners, medical instrument technicians) during sleep clinic appointments over the previous year. Development and revisions of the interview guide were made under the guidance of an experienced health services researcher with a background in qualitative interviewing and interview guide development. Interviews were audio-recorded with permission, and detailed notes were taken by hand during all interviews. When recording was declined (n = 2), the interviewer read the detailed notes back to the interviewee after each question to confirm accuracy.

Interviews lasted an average of 15 min. Patients were asked to describe their general problems with sleep and were screened for insomnia (early insomnia: difficulty falling asleep, middle insomnia: difficulty staying asleep, or late insomnia: waking up early). Next, patients rated their sleep quality over the past week on a scale from 0 (not at all restful) to 100 (extremely restful). They were asked about their initial experiences using PAP: if they tried PAP at home, and how long they tried PAP before discontinuing regular use. Patients were then asked an open-ended question regarding barriers to PAP use. After responding, patients were presented with a checklist of 27 reasons why people may not use PAP (see Table 1). Items were derived from the literature on PAP adherence and/or suggested by providers familiar with PAP nonadherence. Patients were invited to add any additional barriers that were not listed. In addition, each patient was asked to indicate which two reasons were most important for not using PAP. The final section of the interview focused on solutions to barriers to using PAP: Patients were asked if they had spoken with a healthcare provider about their problems with PAP and what solutions had been suggested. Follow-up questions were asked where relevant to generate additional detail for qualitative data collection. Basic patient demographics were collected for descriptive purposes.

Table 1.

Checklist of barriers to PAP use presented to patients

Item	Theme	n	Percent
The mask leaks air.	Physical	18	60
The mask causes sores or irritates my skin.	Physical	6	20
The mask is too big.	Physical	1	3
PAP gives me a headache.	Physical	1	3
The air pressure feels too high.	Physical	9	30
The air pressure feels too low.	Physical	5	17
PAP gives me dry throat or dry nose.	Physical	17	57
My nose is too stuffy to use PAP.	Situational	15	50
I can’t fall asleep with PAP on.	Psychological	7	23
I wake up, and it’s too hard to fall back asleep with PAP on.	Psychological	12	40
I take my mask off in the middle of the night and don’t realize it.	Psychological	12	40
I feel claustrophobic when I use PAP.	Psychological	8	27
I feel anxious when I use PAP.	Psychological	4	13
My nightmares are worse with PAP.	Psychological	0	0
I’m not motivated to use PAP.	Psychological	10	33
PAP doesn’t help me.	Psychological	3	10
My sleep apnea will get better by itself.	Psychological	3	10
I don’t need PAP because I’ll just lose weight.	Psychological	0	0
I don’t know how my PAP works	Knowledge	3	10
My PAP is too much trouble.	Psychological	9	30
I forget to put my PAP on.	Psychological	3	10
I fall asleep before I can put my PAP on.	Psychological	9	30
The machine is too noisy.	Physical	6	20
I go to the bathroom too much to use PAP.	Situational	7	23
It’s too hard to travel with PAP.	Situational	5	17
My bed partner does not like PAP.	Situational	0	0
I’m embarrassed to use PAP.	Psychological	0	0
Other:

Note: Responses to “Other” not reported in this table. PAP positive airway pressure.

Objective PAP adherence data were obtained via device or remote data download from the most recent available week prior to interview, or if unavailable the week closest to the interview within a 2-month period. ­Twenty-four patients had available data. Additional data known to be associated with PAP adherence were also collected from the medical record. Baseline Apnea Hypopnea Index (AHI), residual AHI, and Epworth Sleepiness Scale (ESS) scores were obtained from the medical record to evaluate the relationship between these variables and objective PAP adherence. Baseline AHI obtained via sleep study was available for 28 participants. This represents the number of times per hour each patient experienced apneas or hypopneas during sleep apnea testing and is a descriptor of sleep apnea severity when untreated. Residual AHI was obtained for 22 patients; residual AHI describes how many apneas and/or hypopneas per hour the patient had while using PAP and thus is a measure of treatment efficacy. The most recent ESS score from within a year of interview date was obtained from the medical record, with 27 scores available total. This 8-item self-report Likert-type measure presents eight scenarios and prompts respondents to note their likelihood of dozing in each situation, with 0 representing “never” and 3 representing a “high chance” of dozing. This measure represents daytime sleepiness, with higher scores indicating higher likelihood of dozing across different scenarios [26]. For descriptive purposes, body mass index was obtained from the past or following year, with most recent value used; body mass index scores were obtained for all 30 participants.

Pearson correlations were calculated to evaluate the relationships between PAP adherence and ESS, baseline AHI, residual AHI, and sleep quality. Qualitative data analysis was conducted using an inductive approach to thematic analysis, adapted for the rapid health services research context [27]. Microsoft Excel software was used to enter data and track themes. After all interviews were completed, the interviewer listened to the recordings, reviewed the detailed notes, and prepared summaries of responses to each individual question. These summaries for all interviews were then reviewed by the interviewer to develop a preliminary set of codes. Audio files and written data files were then reexamined and data manually coded, facilitating the iterative identification of themes, refinement of theme descriptions and assessment of their relative prevalence across all interviews. Material categorized into each code and all identified themes/theme descriptions were reviewed by a second rater (board-certified sleep medicine physician; K.F.S.). Questions and discrepancies were discussed to ensure accuracy and consistency in the identification and description of themes.

Results

Participant characteristics

Thirty interviews were completed. Twenty-four interviews were completed in a private examination room on Medical Center grounds; the remaining six interviews were conducted via telephone. Twenty-eight participants consented to audio recording, and two declined to be recorded. Participants’ ages ranged from 28 to 85, with an average age of 58 years (SD = 17, median = 64). Most patients were male (90%). About half were Caucasian (53%). Additional participant demographics and descriptive statistics are presented in Table 2. Average untreated AHI of the sample was 30.21/hr (SD = 19.02; range 4.2–70.5); two-thirds of participants had untreated AHI scores in the moderate (i.e., 15–29) or severe (i.e., 30 and higher) range.

Table 2.

Sample demographics

Characteristic	Frequency, no. (%)
Age
20–29	1 (3)
30–39	6 (20)
40–49	2 (7)
50–59	3 (10)
60–69	8 (27)
70–79	8 (27)
80–89	2 (7)
Gender
Male	27 (90)
Female	3 (10)
Race
Caucasian	16 (53)
Black/African American	3 (10)
Asian/Native Hawaiian or Other Pacific Islander	3 (10)
Hispanic	2 (7)
More than one race	6 (20)
Marital status
Single	10 (33)
Married	15 (50)
Divorced	3 (10)
Widowed	1 (3)
Education
Less than high school	1 (3)
High school graduate	3 (10)
Some college/Associates degree	16 (53)
Bachelors degree	5 (17)
Graduate/professional school	3 (10)
BMI
<25	3 (10)
25–29.9	6 (20)
30 or above	21 (70)

Note: Certain items were missing for some patients: race (n = 1), marital status (n = 1), and education level (n = 2). Percentages may not add up to 100 due to rounding. BMI body mass index.

PAP use

All patients endorsed trying PAP at least once since receiving the device. Many patients had difficulty identifying how long they tried PAP before discontinuing regular use, but just over half (n = 17; 57%) endorsed using it for 1 month or less. Based on PAP device data, nine patients (38%) had zero use in the past 7 days. Of the remaining 15 patients with any PAP use data, average nightly use was 102.53 min (SD = 72.03, median = 66, range 27–270). One patient met criteria for adherence to PAP use in the week leading up to the interview; this patient’s data was included for analysis given their report of nonadherence much of the time. Average reduction from initial AHI was 26.79/hr (SD = 16.89). There was no association between residual AHI and therapy hours (r = .25, p = .29).

Primary sleep complaints and insomnia

Patients described a wide range of sleep problems. The most common problem was frequent and/or extended middle of night awakenings (n = 13; 43%). The next two most endorsed complaints were trouble falling asleep (n = 7; 23%) and not getting enough sleep (n = 6; 20%). Several responses related to various complaints about PAP, such as mask removal while asleep (n = 3; 10%) and mask shifting while asleep (n = 3; 10%). When screened for insomnia, most patients (n = 25; 83%) endorsed at least one insomnia symptom. Middle insomnia was most prevalent, (n = 16, 53%), and early and late insomnia were endorsed by 13 patients each (43% each). Average ESS score was 9.93 (SD = 4.81), and average sleep quality rating was 42.67/100 (SD = 22.66). There was no association between PAP therapy hours and any of the following variables: baseline AHI (r = −.04, p = .84), ESS (r = −.05, p = .81), or sleep quality (r = −.05, p = .81).

Reasons for not using PAP

Patients responded with an average of three responses (SD = 2) to the open-ended question about barriers to PAP. In evaluating qualitative responses to our open-ended inquiry about reasons for not using PAP, four primary themes were identified: (a) physical problems with PAP equipment, (b) psychological barriers to PAP use, (c) gaps in knowledge preventing PAP use, and (d) situational barriers. Physical problems and psychological barriers were each endorsed 39 times, situational barriers endorsed 18 times, and knowledge gaps two times. The most common physical problems were mask leaks (n = 7, 23%), movement during the night shifting the mask (n = 7, 23%), and dry throat/nose (n = 5, 17%). The most common psychological barriers were claustrophobia (n = 5, 17%), worsening of insomnia (n = 5, 17%), and perception of suffocation with PAP (n = 5, 17%). There were several situational barriers endorsed by three patients each (10%): nose too stuffy, frequent bathroom trips, interference of another medical problem, difficulty scheduling appointments, and duration of wait time for equipment to arrive.

The average number of responses endorsed on the barriers to PAP checklist was seven (SD = 3). On the checklist of barriers to PAP use, items endorsed by one third or more of patients included: mask leak (n = 18, 60%), dry nose/throat (n = 17, 57%), stuffy nose (n = 15, 50%), difficulty falling back asleep in the middle of the night with PAP (n = 12, 40%), taking the mask off during the night without recollection (n = 12, 40%), and not feeling motivated to use PAP (n = 10, 33%). When anxiety and claustrophobia were combined into one category, 10 patients (33%) endorsed these problems.

Primary barriers to PAP adherence

Once patients completed the checklist of barriers to PAP use, patients were asked to identify the two most important reasons why they did not use PAP regularly. Two patients did not narrow their checklist responses to two main reasons, and one patient endorsed only one reason, leaving 55 responses for primary barriers to PAP use. Physical problems were most commonly endorsed (n = 27; 45%), then psychological problems (n = 20; 36%). Situational problems were endorsed eight times (13%), and no knowledge gaps were endorsed. Dry nose/throat (n = 12, 22%) was the most commonly endorsed primary reason for not using PAP. Five responses each (9% each) were provided for the mask leaking, pressure feeling too high, and claustrophobia.

Qualitative descriptions of barriers to PAP use

Physical problems with PAP hardware

Patients reported that mask leaks were problematic by creating noise or uncomfortable physical sensations (e.g., air blowing into eyes). These leaks often led to unwanted awakenings. Leaks were also viewed as contributing to dry mouth (another frequently endorsed barrier to PAP use). Many patients were unaware of potential solutions to dry mouth, such as using the humidifier, getting a mask refitted, using a chin strap, or reviewing medications that may cause dry mouth with their primary care provider.

Psychological barriers to PAP use

Claustrophobia and anxiety caused problems for patients both as they tried to fall asleep and during awakenings. Of those who endorsed claustrophobia, most stated there were other situations in which they felt similarly claustrophobic or anxious. Many patients found the mask disorienting or took off the mask in the middle of the night without recollection. Some patients with claustrophobia expressed a preference for a nasal mask over an oronasal mask (i.e., one covering both nose and mouth). Insomnia worsening with PAP was cited as a reason to avoid PAP, particularly for those who found it harder to fall asleep with PAP on. Insomnia as interfering with PAP was endorsed both in terms of primary insomnia as well as in the context of anxiety and depression. Additionally, one patient reported that substance use interfered with remembering to use the device.

Gaps in knowledge preventing PAP use

Knowledge gaps included not understanding how the machine worked, both in general and in reference to specific device settings. This was particularly true of the humidifier: Multiple patients indicated they were unaware that their device had a humidifier, were not sure if they were using it, or did not know how to operate it. Cleaning and maintenance of the device were also confusing to many patients. Further, many patients reported believing the PAP device required supine sleep.

Situational barriers to PAP use

Medical conditions frequently interfered with PAP use. Nasal congestion was a common interfering factor. Pain caused frequent awakenings or position changes and, in one case, prevented the patient from plugging in the device and positioning the mask due to rheumatoid arthritis. Frequent nighttime urination was cited as another disruption to PAP use. One patient objected to the idea of wearing the mask while going to the bathroom, fearing it would scare family. One patient reported that caregiving for family prevented adherence.

Solutions for barriers to PAP use

Almost all patients had spoken with a healthcare professional about difficulty using PAP (n = 25, 83%). Solutions offered were grouped into six different themes: (a) adding or changing PAP equipment, (b) changing PAP settings, (c) instituting an alternative treatment for OSA (e.g., oral appliance, surgery), (d) sleeping laterally, (e) behavioral strategies, and (f) enhancing motivation. The most commonly endorsed theme was adding/changing PAP equipment (n = 20), and the most commonly suggested solution within that theme was trying other mask(s) (n = 14). Seven responses involved changing PAP settings. Seven suggestions involved motivational strategies, though the preponderance of these strategies involved encouragement without specific instructions for how to implement these ideas (e.g., “try harder,” “you really need to use it.”). Behavioral strategies were suggested five times and included desensitization, taking a “PAP nap,” wearing PAP while working, engaging in relaxation exercises, and meeting with mental health staff for behavioral therapy. Alternative approaches to OSA management were suggested three times and lateral sleep four times. When asked if they would be potentially interested in attending weekly coaching to help with PAP, 43% of patients (n = 13) responded affirmatively without qualification. An additional five patients expressed interest but noted concerns about clinic location; they were open to coaching at their local satellite clinic and/or via telehealth technology.

Discussion

Though barriers to PAP use have been evaluated in previous research, most studies have evaluated either one variable or a small subset of factors at a time, and few have invited open-ended exploration of patients’ reasons for not using PAP. We took a patient-centered approach to understanding military veterans’ barriers to PAP use, allowing them the opportunity to describe their barriers to PAP use in their own words, and then applying an inductive approach to thematic analysis, which generates themes from the data itself, rather than a researcher’s preconceived categorization scheme. We presented patients with a checklist of barriers to PAP use as well, to expand the possible barriers considered by each patient. We also asked patients about their sleep problems in general, which allowed us greater information about the problems they find most relevant, rather than defining their most salient problem as “PAP nonadherence,” which is often the clinical target for providers. The patient interviews conducted for this study demonstrate that many factors can lead to suboptimal use of PAP for OSA. Notably, physical and psychological barriers were endorsed most frequently and at approximately equal rates, but solutions offered by providers tended to address physical barriers. This may work for many patients, but we suggest a more individualized approach to increasing adherence. Additionally, patients, on average, endorsed three barriers when prompted for open-ended responses, and seven barriers when presented with a checklist of common barriers to use. Overall, similar barriers and themes were reported across both types of assessment, but more barriers were reported in checklist format. Of note, six patients reported anxiety and/or claustrophobia as barriers to PAP use in response to the open-ended prompt, but an additional four patients endorsed this when presented with the checklist. Possible reasons for this discrepancy include mental health-related stigma or simple deprioritization of this particular barrier. However, these differences in responding suggest that using different methods of evaluating barriers to PAP use (e.g., using mixed methods by supplementing a standard clinical conversation with a checklist) may garner a more comprehensive understanding of each patient’s situation and may prompt the development of more personalized solutions to overcoming barriers.

Physical barriers to PAP use were commonly cited both on the checklist and in open-ended responses. Though nasal masks have been associated with greater adherence than oronasal masks in the literature [22], few complaints were made about oronasal masks in our sample. Oronasal masks were, however, cited as problematic among patients with anxiety or claustrophobia. Of the patients who had tried a nasal mask, many reported preferring this mask but endorsed dry mouth or throat, a common complaint among those who open their mouth while using a nasal mask. Although dry mouth was the most commonly endorsed primary reason for not using PAP, few patients were aware of the need to keep one’s mouth closed while using a nasal mask. In addition, many patients stated they had not used the humidifier or were even aware their device had a humidification option. Notably, while mask leaks were commonly endorsed as a barrier to PAP use, they were rarely endorsed as a primary reason for not using PAP. As the most common solution offered by providers was trying a new mask, there was an obvious discrepancy between this solution and the most frequently endorsed primary barrier to adherence. Poor mask fit may contribute to dry mouth, but there are several other potential causes of this, as well. We recommend providers consider whether other physical aspects of PAP use could be interfering with adherence rather than automatically distributing a new mask as a first step.

Psychological barriers were endorsed at the same rate as physical barriers during the open-ended inquiry, with anxiety and/or claustrophobia endorsed in a third of the sample. Switching from an oronasal to nasal mask was the most common solution offered to address anxiety due to PAP, but anxiety often persisted even after switching masks. In this case, referral to a behavioral sleep medicine provider would be indicated to work with the patient on PAP desensitization. Although desensitization or other psychological methods may reduce anxiety associated with PAP use, desensitization was rarely offered as a solution to the patients in our sample. Desensitization capitalizes on strategies based in behavioral principles to allow the patient to habituate to the new stimulus (i.e., PAP), which could offer the patient the opportunity to acclimate to the optimal device configuration for that patient’s health. An in-person or telehealth-based desensitization appointment, with a behavioral sleep medicine specialist, followed by repeated practice at home over several days, may reduce anxiety or claustrophobia and thus promote increased PAP use. Most patients in this study were open to working with a behavioral sleep medicine provider for more intensive coaching around PAP adherence.

In the USA, approximately 30%–50% of people with OSA report symptoms of insomnia [28]. Eighty-three percent of our sample endorsed symptoms of insomnia. The increased rate of insomnia symptoms in our sample may reflect a unique quality of a veteran population (e.g., trauma exposure, history of disrupted work/sleep schedules) or may reflect previous findings that insomnia is associated with PAP nonadherence [17, 18]. When asked for their chief complaint related to sleep, the most common response was frequent and/or extended middle of night awakenings. These awakenings can be considered a consequence of untreated OSA, insomnia, or both. Moreover, PAP may help reduce middle of night awakenings, but it generally does not address early insomnia. Adjusting to PAP may exacerbate difficulty falling asleep, and patients with concurrent insomnia may experience persistence of insomnia symptoms even with PAP. A consideration of whether OSA or insomnia is the more fundamental problem will allow the provider to assess whether a PAP adjustment is most prudent, or whether behavioral care for insomnia treatment may better suit the patient’s concerns. In addition, we recommend providers emphasize the role of PAP in potentially reducing middle of night awakenings to increase patient motivation.

Situational barriers were endorsed less frequently but are still important considerations. Medical problems often interfered with PAP use, including arthritis, nasal congestion, and use of medications causing increased nighttime urination. Both sleep medicine and primary care providers can play a role in the management of these medical conditions. For example, nasal congestion was reported by half the participants and is a problem readily treated by primary care providers. Referral to sleep medicine may be indicated in certain cases, such as for patients with reduced hand dexterity due to arthritis, who may benefit from masks that utilize magnets to connect to the face or hose, or those who may be treated with an alternative to PAP.

Gaps in knowledge were rarely reported as a hindrance to PAP use, although many patients indicated they thought the initial education about PAP should have been augmented. Several patients were unsure of how their machine worked or how PAP can improve sleep apnea. Improved educational materials and programs, including written materials to take home and online resources, may thus enhance PAP adherence. This may be particularly effective if common barriers to PAP use are anticipated and addressed as part of initial education. Patients also seemed unaware that their complaints may be common. We advise educating patients about the most common problems with PAP upon receipt of the device to normalize their experience and encourage them to seek follow-up care as needed.

While the focus of this project was on barriers to PAP use, several patients made positive comments about PAP over the course of their interviews, and we encourage providers to help patients identify any positive aspects of PAP they have experienced or expect to experience. Some patients were excited to start PAP at first and felt hopeful about potential improvement in sleep. Some patients remarked that they felt better when they used PAP, particularly in terms of daytime sleepiness. One patient noted that his spouse liked PAP because it reduced his snoring. Helping patients identify positive outcomes that may be achieved with PAP use may increase adherence by enabling providers to tailor their approach to increasing motivation. This empowers patients to set their own goals of therapy and aligns the provider in a supportive rather than prescriptive role in achieving these goals. However, future research is necessary to confirm the utility of this approach.

There are several limitations of this study. First, this was a relatively small sample limited to one US VA sleep clinic. This program was neither AASM accredited nor had a board-certified sleep physician at the time of interview, both of which are associated with higher adherence rates [29]. However, PAP intolerance is commonplace in sleep programs and the reasons themselves are unlikely to be caused by the absence of accreditation and board certification. Procedures related to PAP prescription and education vary greatly across providers, and the findings of this study may not generalize to clinics with a different structure or approach to OSA treatment or to non-veteran patients. Additionally, we utilized a pragmatic recruitment strategy and consequently, as is common in mixed methods research, the interviewed sample may not be representative of all veterans; due to the nature of referral to interview, we do not know if there are key differences between veterans who opted to participate and those who declined. There are some aspects of our sample that mirror findings of other veteran studies, such as our sample’s average ESS score being in range of other VA-based studies, and similar untreated AHI [30, 31]. Moreover, like the U.S. veteran population, our study sample was 90% male [32]. We encourage further research on barriers to PAP use among underrepresented groups, such as women veterans [33], as well as non-veterans.

Another limitation is the nature of self-reported symptoms in this study. For example, anxiety and claustrophobia related to PAP were assessed by self-report rather than diagnostic clinical interview. Participants were asked detailed questions regarding insomnia, but a full diagnostic interview determining duration of insomnia or daytime distress or dysfunction was not conducted. There is also the possibility that unconscious bias affected the generation of themes, as the first coder had been trained clinically from a behavioral perspective. We attempted to mitigate this potential bias by having the themes reviewed by and discussed with a sleep medicine physician, rather than another psychologist. The variability in the timing of data obtained outside self-report, such as objective PAP use and ESS, is another limitation of this study, and future research collecting these data as part of research procedures, rather than clinical procedures (as was the case in this study) may promote greater consistency across participants. Nonetheless, there are several strengths of this study. Notably, patients were provided an opportunity to describe obstacles to PAP use from their own perspective.

Conclusions

Nonadherence to PAP is a multifaceted problem that can harm the health and quality of life of patients with OSA. Patients identified a wide range of barriers to optimal PAP use. The barriers most commonly reported were physical and psychological, although solutions offered by providers overwhelmingly focused on physical barriers. We recommend that providers discuss common potential barriers to PAP use with patients—both at the start of treatment, as well as at follow-up—as well as offer a checklist format to encourage more detailed reporting. Collaborative care with behavioral health providers may help increase motivation to use PAP and address comorbid conditions interfering with PAP use, such as insomnia or claustrophobia.

Funding

This study was supported by the Office of Academic Affiliations, Advanced Fellowship Program in Mental Illness Research and Treatment, U.S. Department of Veterans Affairs and the Rehabilitation Research and Development Service, U.S. Department of Veterans Affairs (grant number IK2 RX002952).

Compliance with Ethical Standards

Conflict of Interest: All authors declare that they have no conflicts of interest.

Ethical Approval: This project was evaluated through a San Francisco VA Medical Center Research Determination process and was determined to be a nonresearch quality improvement project and therefore exempt from local institutional review board review.

Informed Consent: Interviewees were apprized of the scope of the project and verbally consented to their participation. Interviewees who consented to audio recording of the interview provided written consent.

Welfare of Animals: This article does not contain any studies with animals performed by any of the authors.

Data Availability

This study was not formally registered. The analysis plan was not formally preregistered. Deidentified data from this study are not available in a public archive. Due to the nature of the data, limited deidentified data may be made available (as allowable according to institutional standards) by emailing the corresponding author. Analytic code used to conduct the analyses presented in this study are not available in a public archive. They may be available by emailing the corresponding author. Materials used to conduct the study are available in Table 1 and the Supplementary Material provided at the journal’s website. Additional detail may be available by emailing the corresponding author.