The development of a digital pathway for the assessment of residual excessive daytime sleepiness in obstructive sleep apnoea: a service evaluation

Abstract

Background

In the United Kingdom, national guidelines for obstructive sleep apnoea (OSA) highlight the importance of assessing continuous positive airways pressure (CPAP) compliance in order to identify patient cohorts of clinical importance, e.g. residual excessive daytime sleepiness (rEDS). This service evaluation reports on the implementation of clinical decision support system (CDSS) software within a bespoke pathway focusing on assessment of rEDS.

Methodology

The CDSS is a National Health Service framework-listed digital ecosystem comprising multiple clinical guided consultations encompassing the entire OSA pathway. Patients commenced on CPAP for ≥6 months and deemed “CPAP compliant” were identified from remote monitoring and subsequently underwent review using a clinical guided consultation (CGC). Patients identified with rEDS following CGC review were then discussed in an interhospital multidisciplinary team meeting, listed for a sleep study on CPAP and booked for consultant physician outpatient review.

Results

472 patients were identified as being CPAP compliant and underwent CGC review, with rEDS noted in 43 (9.7%) of them. rEDS was found to be significantly more common in those reporting chronic insomnia, symptoms of restless legs, dyspepsia, rhinitis and nocturia, as well as in current smokers, those prescribed key medications and in those not in current employment. The multidisciplinary team discussion and subsequent consultant review of the rEDS cohort was able to highlight specific factors requiring changes in patient management including those suitable for wake-promoting therapy.

Conclusion

The incorporation of CDSS technology within a bespoke clinical pathway in order to review patients post CPAP initiation ensures that sleep services are able to readily identify and manage rEDS.

Shareable abstract

The use of a clinical decision support system integrated within a bespoke pathway enables the evaluation of residual excessive daytime sleepiness complicating obstructive sleep apnoea https://bit.ly/3LHuoA3

Introduction

While continuous positive airways pressure (CPAP) is a highly effective therapy for obstructive sleep apnoea (OSA), a proportion of patients complain of residual excessive daytime sleepiness (rEDS) despite being adequately adherent to CPAP [1–4]. It is important for sleep services to identify and manage rEDS cases, as it is associated with impaired quality of life and may also herald the presence of significant underlying comorbidity requiring optimisation, the existence of coexisting sleep conditions and the need to review contributory medications [5].

National guidelines emphasise the need to review patients following the initiation of CPAP in order to identify those subgroups, such as rEDS, which may warrant specific intervention [6]. However, sleep services in the United Kingdom face challenges in meeting this requirement due to limited clinical resources often being prioritised in order to meet the increasing volume of diagnostic activity. Therefore, little is known regarding the prevalence and characterisation of rEDS in a UK population. Furthermore, if we are to truly detail the mechanistic factors at play underlying rEDS in OSA in order to improve clinical outcomes in this area, there remains a need to understand the relationship between rEDS and key symptoms volunteered by patients.

We have previously reported that the implementation of technology in the form of clinical decision support system (CDSS) software may lead to significant pathway transformation in the assessment of OSA referrals, greatly reducing the time to diagnosis, identification of “high-risk” patients and the need for specialist physician review [7–10]. This service evaluation reports the feasibility of using such software when integrated within a bespoke clinical pathway aiming for a structured review of patients post-CPAP initiation, while also encompassing a route for the identification, characterisation and management of rEDS.

Methodology of clinical pathway development

The CDSS used is a National Health Service (NHS) framework-listed and -accredited end-to-end sleep management ecosystem (SleepHealth Solutions, Swaffham, UK) used by the Liverpool Sleep Service for all referrals with suspected OSA. This technology has previously been validated for the assessment, diagnosis and management of OSA [7]. It incorporates multiple clinical guided consultations (CGCs) used during each step of the patient pathway, e.g. during OSA assessment, OSA diagnosis, CPAP set-up, CPAP compliance, dispensing of CPAP consumables, etc. Each consultation integrates a structured history with clinical examination and key investigations intelligently directing the operator to a diagnosis and prompting a management plan based on guideline-level practice. Specifically, when applied to the assessment of patients post-CPAP initiation, the use of a “CPAP compliance” CGC directs patient care by integrating clinical assessment and CPAP compliance data, ensuring that distinct subgroups of patients are identified at the end of the CPAP compliance consultation for appropriate onward review within the sleep service. For instance, those patients with rEDS are automatically identified for medical review, those CPAP-compliant patients with suboptimal control of OSA are identified for therapy adjustment and those requiring attention to CPAP compliance also marked for follow-up (figure 1). The functionality of the CDSS system includes a management dashboard which enables the operator to identify “at-risk” groups at each stage of the patient journey within the service, e.g. sleepy drivers, those operating machinery or not CPAP-compliant.

FIGURE 1.

Excerpt from a patient report following the “continuous positive airways pressure (CPAP) compliance” guided consultation. OSA: obstructive sleep apnoea; ESS: Epworth Sleepiness Scale; AHI: apnoea–hypopnoea index; DVLA: Driver and Vehicle Licensing Agency.

The clinical pathway used in this evaluation is displayed in figure 2.

FIGURE 2.

Clinical pathway for the assessment of residual excessive daytime sleepiness (rEDS). CPAP: continuous positive airways pressure; AHI: apnoea–hypopnoea index; CDSS: clinical decision support software; ESS: Epworth Sleepiness Scale; MDT: multidisciplinary team.

Patients commenced on CPAP for ≥6 months and who were deemed compliant with therapy and where therapy was deemed effective in controlling OSA (CPAP use >4 h for ≥70% of nights within the previous 3 months and with a residual apnoea–hypopnoea index (AHI) <5 events·h⁻¹ as confirmed via CPAP interrogation with mask leak <30 L·min⁻¹) were identified through a remote monitoring database (AirView; ResMed) by a sleep physiologist. This patient group was subsequently invited to undertake a “CPAP-compliance” CGC review for further characterisation. Each CGC review was performed by telephone by a member of a practice support service (PSS) team (paramedical administration staff, who were trained in undertaking the CGC “compliance” review), with each consultation lasting <20 min. We have reported previously on validating the PSS team undertaking CGC review when earlier in the clinical pathway during the diagnosis of OSA [9]. Each CGC compliance consultation completed by the PSS team was then reviewed virtually by two consultant sleep physicians (S. Craig, B. Chakrabarti) and a sleep physiologist (B. Freeman).

rEDS as based on the CGC review was defined by the presence of all the following criteria:

• Epworth Sleepiness Score of ≥11;

• CPAP-compliant (CPAP use >4 h for ≥70% of nights within the previous 3 months);

• AHI <5 events·h⁻¹ (as confirmed via CPAP machine interrogation);

• absence of hypoventilation on initial diagnostic sleep study (>20 min <90% saturation on the sleep study at OSA diagnosis);

• absence of a pre-existing primary sleep disorder, e.g. known diagnosis of central sleep apnoea, hypersomnia, known diagnosis of restless legs syndrome;

• absence of non-CPAP therapy either as primary or adjunctive therapy for OSA syndrome.

Those cases identified as rEDS were then selected for discussion in a monthly interhospital multidisciplinary team (MDT) meeting between Liverpool Sleep and Ventilation Centre and Guys and St Thomas’ NHS Foundation Trust, comprising three consultant sleep physicians (S. Craig, J. Steier, B. Chakrabarti) and a specialist pharmacist with an interest in sleep disorders (J. Kaler). Following discussion at the interhospital MDT meeting, those patients with rEDS were then invited to attend a consultant outpatient clinic in Liverpool Sleep and Ventilation Centre for review and further workup including undertaking a domiciliary sleep study (WatchPAT; Itamar Medical, Caesarea, Israel) on CPAP in order to determine satisfactory control of OSA.

Formal ethical approval to conduct the analysis was obtained from the health research authority (REC reference 24/NW/0154). Statistical analysis was performed using SPSS 28.0 (IBM, USA). Data are presented as mean±sd, unless otherwise stated. Statistical significance was defined as a p-value <0.05. The independent-sample t-test was used to identify significant differences in continuous variables and the Chi-squared test for categorical variables.

Results

472 patients (mean±sd age 53±13 years; 66% male; mean ESS 4.3±4.3 points) underwent CGC review, with rEDS noted in 9.7% (n=46; mean ESS 14.3±2.7 points) between May and September 2024. At CGC review, rEDS was found to be significantly more common in those complaining of chronic insomnia, those reporting symptoms of restless legs, dyspepsia, rhinitis and presence of nocturia (defined as waking with nocturia more than twice nightly). The presence of rEDS at CGC review was reported to be significantly more common in current smokers, those taking key medications (i.e. painkillers/anti-depressants/antipsychotics) and in those not in current employment, with a nonsignificant trend observed between rEDS and female gender (table 1).

TABLE 1.

Factors associated with the presence of residual excessive daytime sleepiness

	Not residually sleepy on CPAP	Residually sleepy on CPAP	p-value
Subjects	426 (90.3)	46 (9.7)
Epworth Sleepiness Scale score	3.23±2.77	14.28±2.74	<0.001
Age years	53±13	53±11	ns
Gender
Female	182 (42.7)	26 (56.5)
Male	244 (57.3)	20 (43.5)	0.07
BMI kg·m−2	36.54±8.78	36.83±7.19	ns
Diabetes	70 (16)	10 (22)	ns
History of cardiovascular disease	93 (21.8)	12 (26.1)	ns
History of cerebrovascular disease	13 (3.1)	1 (2.2)	ns
Diagnosis of hypertension	192 (45.1)	22 (47.8)	ns
On regular antidepressant/antipsychotic/analgesics	301 (70.7)	42 (91.3)	0.002
Heartburn symptoms	131 (30.8)	21 (45.7)	0.04
Restless legs symptoms	141 (33.1)	23 (50)	0.02
Bruxism symptoms	113 (26.5)	17 (37)	0.13
Nocturia (waking up ≥2 times per night)	74 (17)	18 (39)	<0.001
Symptoms of blocked nose/nasal congestion	202 (47.4)	31 (67.4)	0.01
Current or prior diagnosis of depression/anxiety disorder	233 (54.7)	30 (65.2)	0.11
Insomnia symptoms	106 (24.9)	19 (41.3)	0.02
Not in current employment (excluding retirement)	71/338 (21)	16/32 (50)	<0.001
Shift worker in those in currently employed	52/267 (19.5)	3/16 (18.8)	ns
Smoking status			<0.001
Current	26 (6.1)	11 (23.9)
Ex-smoker	185 (43.4)	16 (34.8)
Never-smoker	211 (49.5)	19 (41.3)

Data are presented as n (%) or mean±sd, unless otherwise stated. CPAP: continuous positive airways pressure; BMI: body mass index; ns: nonsignificant.

Out of these 43 patients identified as rEDS, 34 attended a face-to-face outpatient clinic held by a consultant physician. Eight (24%) patients were felt to have significant chronic pain issues likely to explain the rEDS, while in a further four patients, rEDS was felt to be specifically due to life circumstances, obesity and shift work. Chronic insomnia was felt to be a main driver of the rEDS in three patients, with specific pharmacotherapy for insomnia initiated in two of these. 26% (nine out of 34 patients) were found to have inadequate control of OSA (AHI >5 events·h⁻¹) on the sleep study performed on CPAP, and thus required adjustment of therapy. Two rEDS patients were deemed not to require any specific changes in treatment, while five (15%) patients were deemed appropriate for wake-promoting agents. Three patients reported that their symptoms of rEDS had now improved (confirmed by an ESS score <10) between the CGC review and consultant physician review remaining compliant with CPAP, and thus no additional therapy was needed.

Prior to the implementation of this pathway, none of the rEDS cases had been identified or reviewed in the outpatient clinic.

Discussion

We have reported previously that the implementation of validated CDSS technology within a large sleep service at the point of OSA referral reduces the need for specialist review by 85% by streamlining referrals and reducing diagnostic delay while also delivering health economic benefit [9, 11, 12]. This evaluation demonstrates that such technology may also be successfully utilised “downstream” to OSA diagnosis. The CGC was able to stratify patients post CPAP initiation and highlighted key subgroups for further attention, e.g. suboptimal CPAP compliance, rEDS, requiring CPAP adjustment and those suitable for non-CPAP therapies [13, 14]. The use of a guided consultation linked to a management dashboard ensures that patients falling into such groups would be identified in “real time” and then diverted to medical, paramedical or physiologist review as appropriate, resulting in a more effective and efficient service delivery. This is highlighted by the fact that prior to this pathway and under “usual care”, none of the patients diagnosed with rEDS had been identified and reviewed. The remote capability of the initial triage process and CGC review with only those patients requiring specific review or intervention needing a “face-to-face” appointment would reduce the carbon footprint in line with the NHS net zero programme [15]. In addition, this evaluation illustrates the functionality of the CGC review being performed adequately by trained clerical personnel, which would also address staffing and workforce issues in health services already struggling to match capacity with demand [16]. Further studies are needed to determine the applicability of large language models in such a pathway, enabling self-completion of the CGC leading to further pathway efficiencies.

This study demonstrates the need for clinicians to characterise rEDS cases holistically [17]. 26% of patients with rEDS were found to have an elevated AHI on CPAP despite the “device AHI” indicating <5 events·h⁻¹. Patients with rEDS should thus undergo a formal sleep study on CPAP and further research is required to determine the accuracy of CPAP-AHI compared with AHI determined by polysomnography ahead of recommending any pathway modification [18]. In terms of symptoms associated with rEDS, the presence of insomnia was reported in just over 41% of rEDS patients, reinforcing the need for comprehensive clinical assessment and consideration of tests such as actigraphy. Longitudinal studies are required to determine the incidence and pattern of insomnia reported in patients pre- and post-CPAP initiation, focusing on the rEDS subgroup, and to study their response to specific insomnia-based therapy such as cognitive behavioural therapy and licensed pharmacotherapy [19]. Symptoms of rhinitis, dyspepsia and nocturia were also significantly associated with rEDS, supporting the need for assessment of comorbidity which may also be contributing to sleep fragmentation. Such an assessment should also consider the interplay with psychosocial factors, as rEDS was observed more commonly in those not in current employment and in current smokers. Furthermore, any medical review of rEDS should also include the reconciliation of medications that may be linked to daytime sedation, which our study found to be more commonly associated with rEDS. Just over 90% of rEDS patients reporting taking medications such as antidepressants, long-term analgesics and antipsychotics, and furthermore, chronic analgesic use was felt to primarily responsible for rEDS symptoms in 24% when reviewed face-to-face. We report that following a detailed consultant review and MDT discussion after taking these factors into consideration, 11% of rEDS patients were deemed appropriate for wake-promoting agents. This is of particular importance given the emergence of licensed wake-promoting therapies targeted at rEDS and supports the need for sleep services to develop a structured pathway incorporating MDT discussion for the management of rEDS cases and the need for “real-world” observational data determining patient outcomes in response to wake-promoting agents [20–22]. Future studies incorporating this and similar CDSS technology would utilise artificial intelligence, aiming to better characterise those patients who would be predicted to respond to wake-promoting agents and any other emerging potential rEDS therapies so that such patients would be identified earlier in the clinical pathway.

There are limitations with the implementation of this pathway. We did not objectively confirm EDS using a multiple sleep latency or maintenance of wakefulness test. However, this would present challenges in terms of resource if undertaken for every rEDS case as well as in correlating the significance of such tests post CPAP initiation [23, 24]. In this pathway, each CGC review was conducted remotely by nonmedical staff who had received appropriate training in the CDSS software. However, each CGC review subsequently underwent verification by sleep specialist staff with each rEDS case further discussed at an interhospital MDT discussion. While the history taken during the CGC review did incorporate common categories of medication with the potential to cause sedation, the CGC review did not incorporate specific details regarding the routine prescription of other categories of medications such as β-blockers or sedating antihistamines, which is an area for software refinement. However, the pathway specifies that the CGC review should be followed by a formal physician assessment in rEDS cases, although nine out of 43 patients did not complete the outpatient appointment. Ideally in a CPAP-compliance pathway, as per guideline level care, CGC reviews should take place ideally within 1 month of CPAP with subsequent follow-up dependent on the outcome of this review [6]. In this feasibility evaluation, we did not perform CPAP compliance reviews within this period, but all patients had been established on CPAP with adequate compliance for >6 months, thus presenting a sufficient time interval for making an accurate diagnosis of rEDS. While a history of depression or anxiety was not reported as being associated with rEDS, unlike some previous studies, a formal depression and anxiety scale was not utilised during the CGC review which again is an area warranting possible future pathway modification [25].

In summary, this study supports the feasibility of implementing clinical decision support software technology enabling the creation of a structured clinical pathway enabling the identification of rEDS complicating CPAP therapy in OSA.

Data availability

The datasets generated during and analysed during the study are not publicly available due to the General Data Protection Regulation.