Not there yet; the challenge of treating sleep-disordered breathing in people living with spinal cord injury/disease

Abstract

In the current edition of SLEEP, Badr et al. [1] report an important study that aimed to improve positive airway pressure (PAP) therapy adherence in Veterans living with spinal cord injury (SCI) or disease (SCI/D) and sleep-disordered breathing (SDB). Acute cervical SCI (tetraplegia) causes severe, persistent obstructive sleep apnea (OSA) [2] with a resultant population prevalence of at least mild disease estimated as 83% [3]. Central and mixed sleep apnea events, especially at sleep onset, have been reported in tetraplegia [4], but OSA predominates in larger community samples [5]. The Badr et al. study is a randomized controlled trial (RCT) comparing the efficacy of a combined sleep and PAP adherence program, (the “BEST” program; Best practices PAP + patient education + ongoing support and training). The study aimed to compare usual care with usual care plus BEST on PAP adherence, as defined by the number of nights where usage was above 4 hours per night during the 3-month (90-day) trial intervention period. The co-primary aim was to examine the impact of the program on sleep quality (measured by the Pittsburgh Sleep Quality Index [PSQI]). Secondary outcomes included general functioning, respiratory functioning, quality of life, depression, fatigue, and sleepiness [1]. The authors are commended for being the first to investigate a comprehensive and evidence-based intervention that aimed to improve PAP adherence in SCI. Despite the extremely high prevalence of OSA in SCI/D [3], access to screening, testing, diagnosis, and treatment is poor [6], and uptake and adherence with PAP are uniformly reported as challenging [7]. The behavioral component of the BEST intervention was based on a previous RCT, similarly targeting the Veteran population but in those without SCI/D living with co-morbid insomnia and newly diagnosed OSA [8]. That experiment reported improvements in PAP adherence, sleepiness, and sleep quality, similar to other studies targeting PAP adherence with supportive psychological and/or behavioral interventions [9]. The control intervention in the current study was an attentional control (no additional education and general sleep advice only) which was completed by the same proportion of participants as the intervention arm. Diagnosis and pressure determinations were made overnight in a sleep laboratory for both groups, reflecting current guidelines. Despite this close attention to the design of the intervention, and evidence of fidelity to delivery, no difference in PAP usage between the intervention and control participants was observed. Most PAP studies in people living with SCI/D papers report low usage, but usage is reported and categorized variably across the literature. Arguably reporting average hours of use over a set time period is a more accurate, comparable, and granular method of describing PAP usage. Average hours of use per night throughout the Badr trial ranged from approximately 2.3 at 1 month to 1 (6 months), with an average primary adherence, defined as at least 4 hours per night over the 3-month trial period, of 23% (data contained in their Supplementary Table S4) [1]. In our group’s RCT (the COSAQ study) of treating OSA with auto-titrating (A)PAP in acute tetraplegia, overall adherence was 33% (n = 26/79), and overall average APAP use was 2.9 hours per night [10]. In COSAQ, 48 people were excluded prior to randomization for failing to achieve >4 hours per night on a 3-night APAP run-in. If included as “non-adherent,” COSAQ adherence falls to 21% [11]. In a prospective cohort of 16 people with chronic tetraplegia and OSA, adherence with continuous (C)PAP at 1, 6, and 12 months was 38%, 25%, and 25%, respectively, with average nightly use of 3.1, 2.6, and 2.1 hours [7]. More recently, Di Maria et al. [12] retrospectively analyzed CPAP use in a clinical sample of people with SCI/D and moderate to severe SDB. Using the same >4 hours per night PAP adherence definition, and with inclusion of those who refused CPAP or were lost to follow-up (conservatively assumed to be non-users), the 6- and 12-month adherence rates in this study would be 31/80 (39%) and 24/80 (30%) [12]. As detailed above, PAP adherence in SCI/D ranges from 20% to 40%, comparable with CPAP adherence in the non-SCI/D which ranges from 17% to 54% [13]. Alternatives to PAP for OSA have been developed in the general population. A systematic review comparing effectiveness of CPAP and Mandibular Advancement Devices (MAD) for OSA in the general population concluded that while CPAP more effectively lowered the apnea–hypopnea index, adherence was significantly lower than MAD, resulting in no discernible differences in quality of life, cognitive, and functional outcomes [14]. We are unaware of any publications examining MAD use in SCI/D; an obvious area of opportunity. The traditional 4-hour usage target, which arose from data in the general OSA literature, is clearly difficult to achieve for many people with SCI and OSA. Further, it is accepted that CPAP confers a dose–response relationship [15]. Indeed, Badr et al. demonstrated this, finding significant associations between hours of CPAP use and important outcomes, despite the very low usage overall [1]. These data will prove very useful for exploring the minimum dose of PAP for improvements in meaningful outcomes such as sleepiness, fatigue, and mood in the SCI/D population. We strongly support the assertion by Badr et al. that inadequate management of SDB in SCI is another example of healthcare inequity impacting people with disability. Eighteen percent (23/127) of people with SCI/D screened for inclusion in their study were using PAP [1]; similar to treatment rates reported in other studies [16, 17]. Furthermore, 97% (63/65) of enrolled participants were diagnosed with SDB; over one-third had severe SDB. This represents an enormous unmet need in SCI/D, and a failure of contemporary care models to ensure accessible, evidence-based management of a common and deleterious disorder. To combat these issues, the authors rightly suggest more collaborative, intensive, and individualized management of sleep disorders in SCI/D. We agree, but further assert that to overcome barriers to obtaining care, we must also investigate alternative, non-specialist models of managing non-complicated OSA. Our research in this area has shown that the predominant care pathway in SCI involves referring people with suspected SDB to specialist sleep services, but that this model is often inaccessible and/or unsuitable for people living with SCI/D [7, 18]. To overcome these barriers, we have been investigating non-sleep specialist models of managing uncomplicated OSA in SCI. A shift toward non-specialist models of managing sleep disorders is well underway in primary care where numerous RCTs have demonstrated non-inferiority of ambulatory, home-based diagnosis and CPAP initiation [19, 20]. Furthermore, level 1 evidence demonstrates non-inferiority of primary care management of non-complicated OSA compared with traditional, sleep specialist models [21]. Several SCI rehabilitation centers have adopted a similar approach to independently diagnosing and treating non-complicated OSA [18]. We have recently adapted, implemented, and pilot-tested a similar model that also screens for and refers out more complicated cases, such as hypoventilation [22]. Sleep problems are commonly ranked as important impairments to a healthy life by people living with SCI/D and are among the least likely of all secondary health problems to be treated [23–26]. Badr and colleagues are to be congratulated for directly addressing this prevalent, unmet need. While their study intervention was not better than usual care, their research highlights that any PAP use is better than none, and they have provided important signposts on the journey towards better sleep care for people living with SCI/D.