Although efficacious in the treatment of obstructive sleep apnea (OSA), continuous positive airway pressure (CPAP) can be difficult to tolerate, with long-term adherence rates approaching 50%. CPAP alternatives clinics specialize in the evaluation and treatment of CPAP-intolerant patients; yet this population has not been studied in the literature. To better understand these patients, we sought to assess insomnia, sleep-related functional status, sleepiness, and nasal obstruction, utilizing data from validated instruments.
After approval from the Emory University Institutional Review Board, a retrospective chart review was performed from September 2015 to September 2016 of new patient visits at the Emory CPAP alternatives clinic. Patient demographics and responses were recorded from the Insomnia Severity Index, Functional Outcomes of Sleep Questionnaire-10 (FOSQ-10), Epworth Sleepiness Scale, and Nasal Obstruction Symptom Evaluation questionnaires.
A total of 172 patients were included, with 81% having moderate-severe OSA. Most of the patients demonstrated moderate-severe clinical insomnia and at least moderate nasal obstruction. FOSQ-10 scores indicated sleep-related functional impairment in 88%. However, most patients did not demonstrate excessive daytime sleepiness.
This patient population demonstrates significant symptomatology and functional impairment. Because of the severity of their OSA, they are at increased risk of complications. In order to mitigate the detrimental effects of OSA, these significantly impacted patients should be identified and encouraged to seek CPAP alternatives clinics that specialize in the treatment of this population.
Lam AS, Collop NA, Bliwise DL, Dedhia RC. Validated measures of insomnia, function, sleepiness, and nasal obstruction in a CPAP alternatives clinic population. J Clin Sleep Med. 2017;13(8):949–957.
Obstructive sleep apnea (OSA) is a common sleep-related breathing disorder with estimated global prevalence ranging from 9% to 38%.1 The consequences of OSA stem from months to years of recurrent upper airway collapse. The resulting impairment in ventilation and resultant fragmented sleep contribute to significant cardiovascular, cerebrovascular, metabolic, and neurocognitive morbidity, with an annual economic effect in the tens of billions of dollars.2
Developed in the early 1980s, positive airway pressure (PAP), usually in the form of continuous positive airway pressure (CPAP), is the current gold standard OSA therapy.3 By increasing pressures within the upper airway, CPAP aids in resisting the collapse responsible for obstruction. In the years since its introduction, CPAP has proven efficacious in reducing OSA associated symptoms and improving both the quality of life and long-term outcomes.4
Although treated patients report reductions in OSA-related symptoms such as sleepiness, anxiety, and depression, fewer workdays missed secondary to illness, and increases in quality-of-life scores, many (between 30% and 60%) discontinue CPAP treatment.5,6 These patients are at significantly greater risk of adverse events, with one study demonstrating a 10% increase in 5-year mortality, when compared with CPAP adherent patients.7
Current Knowledge/Study Rationale: CPAP alternatives clinics specialize in the treatment of patients in whom CPAP therapy fails or is declined. Because CPAP nonadherence is highly prevalent and contributes to significantly decreased quality of life and overall health, this study was undertaken to examine and describe the characteristics of patients who are CPAP nonadherent who actively seek treatment and present to a CPAP alternatives clinic.
Study Impact: This study demonstrates that patients who present to a CPAP alternatives clinic have moderate to severe, untreated OSA and exhibit significant insomnia, nasal obstruction, and functional impairment. This highlights the importance of encouraging patients such as these to seek CPAP alternatives clinics for both symptomatic improvement and avoidance of OSA sequelae.
Although many patients remain untreated or undertreated, another important group of patients actively seek alternatives to CPAP therapy. As both the public and medical field have come to recognize the consequences and burdens of inadequately treated OSA, the number of CPAP alternatives has grown. The wide array of alternatives have varying degrees of invasiveness and effectiveness, and may be tailored to address each patient's individual condition.
Characteristics and factors associated with CPAP therapy nonadherence have been previously assessed and have suggested a multitude of associations. Although these data represent factors related to CPAP adherence, no study has assessed the characteristics of patients who present to CPAP alternatives clinics seeking other treatment options. This is potentially important because these patients, seeking alternatives to CPAP therapy, may be more motivated for seeking OSA care based on factors that are currently poorly understood. Utilizing validated measurement tools, we sought to comprehensively assess the characteristics of CPAP nonadherent patients who sought treatment in a CPAP alternatives clinic.
In September 2016 the Emory University Institutional Review Board (IRB) granted approval to conduct this retrospective chart-review study. A single data reviewer reviewed medical records of all new, adult patient visits to the single provider, Emory CPAP alternatives clinic, from September 8, 2015 to September 29, 2016. Inclusion criteria for patient encounters were: (1) new patient visit to the CPAP alternatives clinic, (2) patient age 18 years or older at the time of the encounter, and (3) patient history indicated that either the patient was unable to initiate PAP therapy, or the patient had a history of PAP use but was unable to tolerate and/or wished to discontinue PAP.
If all inclusion criteria were met, data regarding patient history, symptoms, intake questionnaires completed during initial CPAP alternatives clinic visit, physical examination findings, sleep studies, clinic procedures, and treatment plan were collected by a single data abstractor, from the patients' new patient visit CPAP alternatives clinic notes. Sex and date of birth were obtained from the patient medical record. Patient age at the time of CPAP alternatives clinic visit was calculated using the patient date of birth and date of the visit. All other demographic data were obtained from the patients' new patient visit CPAP alternatives clinic notes. Comorbid conditions, surgical histories, and answers to several standardized questions such as, “As of today, is sleep apnea your most important health problem?” were obtained from both the new patient visit CPAP alternatives clinic notes, and from the historical medical records. Collected intake questionnaires included the Insomnia Severity Index (ISI), Functional Outcomes of Sleep Questionnaire-10 (FOSQ-10), Epworth Sleepiness Scale (ESS), and the Nasal Obstruction Symptom Evaluation (NOSE) instruments.8–11 Standardized formatting was used for new patient visit notes and patient intake questionnaires.
Data were abstracted utilizing a standardized abstraction form (Microsoft Access v16.0, Microsoft Corporation, Redmond, Washington, United States). All available data points were collected. Data quality was assessed by the lead author prior to analysis. In the event of data discordance, or question of data validity, the senior author was consulted for resolution. In the case of data omission, complete data points from the chart were included in the analysis.
Once collected, data were stored within a secure folder in compliance with the Health Insurance Portability and Accountability Act. A complete list of collected data can be found in the supplemental material.
Insomnia Severity Index
The ISI is a short survey (7 questions) that has been validated as a tool for both insomnia screening and in evaluating treatment response.9,11 Respondents are asked a series of questions regarding the severity of their insomnia symptoms and instructed to record their responses on a 0 to 4 Likert scale. A total score is calculated by summing the responses to each question. The first three questions of the ISI may be used to evaluate the contributions of initial/onset, middle/maintenance, and terminal insomnia. For this study, only completed questionnaires were scored. Total ISI scores were interpreted as follows: no clinically significant insomnia (0–7); subthreshold insomnia (6–10); moderately severe clinical insomnia (15–21); and severe clinical insomnia (22–28).11
Functional Outcomes of Sleep Questionnaire-10
The FOSQ-10 is a validated, shortened (10 questions) version of the original Functional Outcomes of Sleep Questionnaire that was developed to assess the functional effect of a patient's daytime sleepiness.12,13 The survey is broken down into several domains, capable of assessing general productivity, activity level, vigilance, social outcomes, intimacy, and sexual relationships.12 On the questionnaire, patients are asked to rate the effect of their sleep disorder on a number of activities, with 1 indicating the highest level of effect and 4 indicating the lowest level of effect. Subjects may also answer 0, if the individual does not engage in the activity for reasons unrelated to the effect of their sleep disorder. Each of the five subscales are scored by averaging the responses from within the subscale. Responses are only included in the calculation if 1 or more. Thus, skipped questions and responses of 0 are not included in the score. The total score is calculated by averaging subscale scores with at least one response of 1 or more, and multiplying the average by 5. In this study the FOSQ-10 was scored for all individuals who completed all or part of the survey. A total FOSQ-10 score less than 17.9 was used to indicate sleep-related disease-specific functional impairment.14
Epworth Sleepiness Scale
The ESS evaluates sleepiness and is a simple method of screening for sleep-related disorders including OSA.8,15 Respondents are asked to rate the likelihood of dozing off in various situations, over the past week or so. A 0 to 3 Likert scale is used, with 0 indicating that the patient “would never doze off” and 3 indicating a “high chance of dozing.” Responses to each of the eight items are summed for a total score. In this study, only completed ESS surveys were included in analysis. Totaled ESS survey scores greater than 10 were interpreted as excessive daytime sleepiness.8 Scores were further interpreted as follows: lower normal daytime sleepiness (0–5), higher normal daytime sleepiness (6–10), mild excessive daytime sleepiness (11–12), moderate excessive daytime sleepiness (13–15), and severe excessive daytime sleepiness (16–24).16
Nasal Obstruction Symptom Evaluation
The NOSE is a validated survey used in assessing nasal obstruction.10 The 8-question survey asks respondents to rate, over the past month, how much of a problem various conditions had been for the patient. A 0 to 4 Likert scale is used, with 0 indicating “not a problem” and 4 indicating “severe problem.” For a total score, responses are summed and multiplied by 5. In this study, NOSE scores were included if all items were answered. NOSE scores were interpreted as follows: none to mild nasal obstruction (0–25), moderate nasal obstruction (30–50), severe nasal obstruction (55–75), and extreme nasal obstruction (80–100).17
SPSS Statistics v24 (IBM Corporation, Armonk, New York, United States) was utilized to evaluate data distribution, descriptive statistics, and correlations. Data distribution was assessed numerically via the Shapiro-Wilk test of normality, and graphically by Normal Q-Q plot. Geographic means were calculated and are reported for variables with non-normal distributions. For normally distributed data, and in the instance of non-normal distributions for variables with values of zero present in the dataset (apnea-hypopnea index [AHI], percent total recorded time with SpO2 < 90% [%TRT < 90%], ESS, and NOSE) the arithmetic means were calculated and are reported. Correlations were assessed via Spearman rank-order correlation.
Table 1 presents patient age, sex, body mass index (BMI), current and previous therapies, and sleep study data. A total of 172 patients met inclusion criteria. Within this cohort, 56.9% of patients (n = 95) were nonobese (BMI < 30). Ninety-eight percent of patients (n = 168) reported having received at least some previous therapy (PAP, oral appliance therapy [OAT], positional therapy, or surgery). Reasons for PAP failure are shown in Figure 1. The most commonly cited reasons included claustrophobia (18.6%), mask discomfort (12.8%), and sleep difficulty (12.2%). Fifty-five percent reported a history of non-PAP therapy, 95.8% of whom also reported a history of PAP treatment. Reasons for OAT failure are demonstrated in Figure 2. Sixty-four patients (37.2%) reported previous upper airway surgery for the treatment of obstruction. Eighty-one percent of patients demonstrated moderate to severe OSA (AHI or respiratory event index ≥ 15 events/h). Common comorbid conditions are listed in Table 2.
Reasons for PAP failure.
The most commonly cited reasons for PAP failure, in patients for whom this data was available (n = 156). PAP = positive airway pressure.
Reasons for OAT failure.
The most commonly cited reasons for OAT failure. OAT failure was documented in 41 patients, 77.4% of those who reported a history of OAT use. The data in this figure include data from subjects for whom the reason for OAT failure was documented. (n = 34). OAT = oral appliance therapy, TMJ = temporomandibular joint syndrome.
Treatment plan data were available for all patients. Specific therapy, or evaluation for specific therapy, was recommended in 91.9% of patients (n = 158). Plans involved some non-PAP therapy, or evaluation for non-PAP therapy in 83.1% of cases (n = 143). Forty-eight percent of subjects (n = 83) received a recommendation for either surgery or evaluation for surgery, and 29.1% of patients (n = 50) were recommended for either OAT or evaluation for OAT. Patient plans included either PAP or evaluation for PAP in 41.3% (n = 71). Further patient evaluation with drug-induced sleep endoscopy was recommended for 41.3% of patients (n = 71).
Insomnia Severity Index Results
Of the 123 new patient visit records from which ISI survey data were available, 11 were excluded due to incomplete surveys. Completed ISI survey data were available for 112 patients (65.1% of total). Total scores were 15 or higher in 66 patients (58.9%), indicating moderate or severe clinical insomnia. No clinically significant or subthreshold insomnia was demonstrated in 46 patients (41.1%).
When asked to rate their difficulties (0 = none to 4 = very severe) falling asleep (initial insomnia), staying asleep (middle insomnia), and waking up too early (terminal insomnia), patients noted greatest difficulty with staying asleep (average item score 1.94 ± 1.39). Additional data from individual ISI items as well as ISI scores and interpretations can be found in Table 3.
Functional Outcomes of Sleep Questionnaire-10 Results
FOSQ-10 data was available for 120 patients (69.8%) from whom the majority (88.3%) demonstrated sleep-related disease-specific functional impairment (total score less than 17.9). With the exception of the activity level subscale, all calculated subscale averages were less than 3.0, indicating functional impairment. Complete FOSQ-10 data are listed in Table 4.
Functional Outcomes of Sleep Questionnaire-10.
Functional Outcomes of Sleep Questionnaire-10.
Epworth Sleepiness Scale Results
One hundred twenty-eight new patient visit records included ESS questionnaire data. Data from 11 of these were excluded secondary to incomplete surveys. Completed ESS data were available for 68.0% of patients, of whom the majority (56.4%) demonstrated normal daytime sleepiness (total score 10 or lower). Excessive daytime sleepiness (total score at least 11) was demonstrated in 43.6%. Complete ESS data may be found in Table 5.
Epworth Sleepiness Scale and Nasal Obstruction Symptom Evaluation.
Epworth Sleepiness Scale and Nasal Obstruction Symptom Evaluation.
Nasal Obstruction Symptom Evaluation Results
Data from NOSE questionnaires were available from 127 new patient visits. Of these, two questionnaires were incomplete and were excluded. Eighty patients (64%) indicated at least moderate nasal obstruction (total score at least 30), whereas 45 (36.0%) indicated no or mild nasal obstruction (score range 0–25). NOSE questionnaire data are listed in Table 5.
Spearman rank-order correlations were performed to assess relationships between validated measures. There were moderate correlations between scores of all tests, with the exception of the correlation between FOSQ-10 and ISI, which was strong. All correlations were statistically significant. Correlations are listed in Table 6.
Correlations/Spearman rho for validated measures.
Correlations/Spearman rho for validated measures.
Prior to the current study, there existed no analysis of the CPAP alternatives clinic patient population. These patients are at increased risk of potentially associated cardiovascular complications—greater than 80% of this cohort demonstrated moderate to severe OSA with approximately half having comorbid hypertension. The majority of patients in this cohort suffered from moderate to severe clinical insomnia and nasal obstruction, and approximately 90% demonstrated sleep-related functional impairment. Yet, most of these patients did not experience excessive daytime sleepiness. The majority of this cohort, in need of OSA treatment, were able to receive recommendations during their initial visit.
General demographics of this cohort, patients referred to a CPAP alternatives clinic, are relatively similar to other OSA cohorts.18,19 Interestingly, although all patients presented seeking alternatives to CPAP therapy, more than 30% of patients were using CPAP at the time of their encounter. Nearly all patients reported having previously attempted some form of OSA treatment, most frequently CPAP (95%).
Recent investigation has indicated an association between CPAP nonadherence and nonobesity in OSA.20 In the current report, more than half of studied patients were nonobese (BMI < 30 kg/m2). It is postulated that up to 90% of these non-obese patients demonstrate low arousal thresholds for airway narrowing.20 Gray et al. hypothesize that low arousal threshold, characterizing these patients, contributes to frequent sleep disruption that is exacerbated by CPAP masks and positive pressure.20 Clearly, this phenotype represents a challenging group to treat. Although CPAP remains the first-line therapy for most patients with OSA, it is reasonable to believe that patients such as these could benefit from therapy that does not increase arousals.
Validated instruments were utilized in this study to measure specific sleep-related qualities. The ISI, one such instrument, was used to evaluate for insomnia.9 Most of these patients exhibited significant clinical insomnia, defined as moderate to severe by ISI score. Specifically, 58.9% of patients met these criteria. In this cohort, the prevalence of insomnia was on the higher end of reported insomnia rates of patients with general OSA and considerably more than 15.2% of healthy patients with insomnia in other published studies.21–23 OSA and insomnia demonstrate a synergistic effect on mood disorders, and patients with these combined comorbidities have been shown to exhibit increased sleep apnea symptom severity.21 Though a topic of continued debate, it has been suggested that insomnia itself influences CPAP adherence.24 CPAP mask discomfort and mechanical constraints may exacerbate preexisting insomnia and, in this way, pose a substantial challenge in the treatment of OSA.
Functional impairment is a common problem among untreated patients with OSA. In a previous study, Weaver et al. found 81.6% of patients with untreated, moderate to severe OSA (AHI ≥ 15 events/h) exhibited sleep-related disease-specific functional impairment.14 For the current study, the validated FOSQ-10 was used to measure the functional effect of patients' disordered sleep.12 Nearly 90% of this cohort demonstrated significant functional impairment, with the greatest effect noted in social outcomes. This implies that quality of life is substantially affected in nearly all of these patients, although this can be improved with appropriate therapy.4,25
Sleepiness is largely regarded as a characteristic symptom of OSA.26,27 The reported rates of excessive daytime sleepiness in OSA are variable with a reported prevalence of between approximately 30% and 87%.28,29 This may be compared to 18% in the general population.30 The ESS was used as a measure of excessive daytime sleepiness for patients in this cohort. Though falling within the reported rates of excessive sleepiness in OSA patients, more than half of patients in this study exhibited normal daytime sleepiness (ESS ≤ 10). Previous reports have demonstrated an association between CPAP nonadherence and lower ESS scores with less clear benefit of CPAP in nonsleepy individuals.31,32
Nasal congestion is one of the most frequently reported side effects of CPAP use and has known association with CPAP nonadherence.33 The NOSE instrument was used in this study as a measure of nasal obstruction. For perspective, the average NOSE score in a healthy patient is approximately 15.34 In this study, the mean NOSE score in the cohort was 43, indicative of moderate nasal obstruction. Nearly two-thirds of this cohort exhibited obstruction of at least moderate severity. Surgical correction of nasal obstruction has been associated with decreased CPAP device pressures, increased CPAP use and adherence, improved AHI, and better sleep quality.35–38
In a cluster analysis of the Icelandic Sleep Apnoea Cohort (ISAC), Ye et al. proposed that patients with OSA can be subdivided into three distinct clusters according to characteristics of their clinical presentations.39 The three groups are as follows: minimally symptomatic, excessive daytime sleepiness, and disturbed sleep. The disturbed sleep group comprised approximately one-third of included patients, had lower ESS scores than the excessive daytime sleepiness group (the most prevalent group), and were most likely to experience symptoms of insomnia. Ninety percent of these patients experienced middle insomnia, and most of them complained of nocturnal nasal congestion. This cluster appears most similar to this cohort, most of them whom did not demonstrate excessive daytime sleepiness but did experience significant clinical insomnia and nasal obstruction.
Moderate-severe OSA contributes to significant morbidity, and substantially increases the risk of cardiovascular and cerebrovascular disease, as well as all-cause mortality.40 Within this population, increased disease severity has been shown to further increase the risk of mortality.41 Patients in this study presented, having failed CPAP, with moderate to severe, symptomatic OSA—nearly all were given recommendations to be assessed for or to undergo some form of OSA treatment. Both the Adult Obstructive Sleep Apnea Task Force of the American Academy of Sleep Medicine and the American College of Physicians recommend treatment for these patients.25,42 According to these bodies, as well as the American Academy of Otolaryngology-Head and Neck Surgery, if CPAP therapy is declined or failed it is appropriate to consider alternative treatment options, and in these patients is essential for improving outcomes.25,42,43
This study has limitations. Encounters were reviewed from a single provider clinic within one academic center. Thus, neither sampling bias nor referral bias can be excluded and characteristics of this population may not represent patient characteristics at similar clinics. Additionally, this study was conducted within a training institution in which multiple providers are frequently involved in the patient encounter. As such, there were several individuals responsible for the input of patient information in the medical record. Although encounter notes were completed in a standardized format and patients completed their own questionnaires, there was likely some variability in the amount and detail of information in the medical record. This study was designed as a descriptive study and was noncomparative by nature; there was no control group. Therefore, unclear transferability limits the validity of comparisons with other cohorts. Finally, four patients (2%) in this study had not previously tried PAP due to refusal or medical contraindication (eg, history of skull base surgery).
Clearly, issues related to sleep are having a significant effect on the daily lives of individuals in the studied cohort. In many regards, patients in this cohort were similar to patients in other OSA cohorts, and before presenting to a CPAP alternatives clinic, nearly all of these patients had attempted PAP and other OSA therapies. Many patients had even undergone previous upper airway surgical intervention. Nonetheless, these patients continued to exhibit significant insomnia, sleep-related functional impairment, and nasal obstruction. In this study, patients referred to a CPAP alternatives clinic demonstrated severe symptomatology and are likely more impaired than other OSA subpopulations.
Utilizing validated instruments to measure the effect and symptoms of OSA, this study is the first to characterize patients of a CPAP alternatives clinic. It is evident that these are mostly patients with moderate to severe OSA, exhibiting impairment from daytime function to nighttime sleep quality, not necessarily daytime sleepiness. Such patients should be encouraged to seek CPAP alternatives clinics to mitigate the detrimental effect of OSA on both cardiovascular risk and quality of life.
All authors have seen and approve this manuscript. Austin Lam has received no financial support for this work and has no conflicts of interest to disclose. Nancy Collop has received no financial support for this work and receives royalties from UpToDate and grant monies from Jazz Pharmaceuticals. Donald Bliwise has received no financial support for this work. He is a Consultant for Ferring, Merck and Respicardia. Raj Dedhia is a practicing sleep otolaryngologist who staffs a CPAP alternatives clinic. Raj Dedhia has no additional disclosures and received no financial support for this work.