Therapeutic options for patients with obstructive sleep apnoea: a retrospective review of difficult-to-treat cases in a multidisciplinary approach

Abstract

Introduction

Continuous positive airway pressure (CPAP) is the first-line treatment for obstructive sleep apnoea (OSA) but long-term adherence is limited. This study evaluates the implementation and outcomes of a multidisciplinary team approach to managing CPAP-intolerant patients using alternative therapies.

Methods

A retrospective review was conducted on ‘difficult-to-treat’ OSA cases discussed at multidisciplinary team meetings at the Guy’s and St Thomas’ NHS Foundation Trust Sleep Disorders Centre (project number 17 160). Patient demographics, comorbidities, OSA severity (Apnoea-Hypopnoea Index (AHI), Epworth Sleepiness Scale (ESS), 3% Oxygen Desaturation Index (ODI)), management plans and outcomes were extracted from meeting minutes and clinical records. Statistical analysis included analysis of variance (ANOVA) to compare baseline characteristics (AHI, age, body mass index (BMI)) across treatment groups and paired t-tests to assess pre-treatment and post-treatment changes in OSA indices (p<0.05).

Results

We reviewed a cohort of 138 patients (103/138 male, age 51 (17) years, BMI 29.5 (5.1) kg/m²) with severe OSA (AHI 37.7 (21.4)/hour). Common comorbidities included hypertension (29.0%), anxiety/depression (25.4%), history of previous Otorhinolaryngology (ENT) intervention (17.4%) and cardiac diseases (16.7%). Treatment selection was guided by predefined criteria, including OSA severity, BMI, patient preferences, anatomical abnormalities and comorbidities. Treatment recommendations included mandibular advancement devices (27.5%), hypoglossal nerve stimulation pathway (24.6%), weight management (23.9%), CPAP retrial (13.8%), other ENT surgery (10.9%) and BiPAP/ASV (10.1%). Follow-up data from patients with postintervention data demonstrated significant reductions in AHI (−44.1%, n=29), ESS (−32.9%, n=13) and ODI (−67.2%, n=12) post-treatment (p<0.05).

Conclusion

A multidisciplinary approach enables structured delivery of evidence-based treatment selection for patients with OSA who failed first-line therapy. Observed improvements in a subset of patients suggest potential benefit of tailored non-CPAP strategies and super-regional accessibility.

WHAT IS ALREADY KNOWN ON THIS TOPIC

• Continuous positive airway pressure (CPAP) is the first-line treatment for obstructive sleep apnoea (OSA), but long-term adherence is often poor. Several evidence-based non-CPAP therapies are available for CPAP-intolerant patients.

WHAT THIS STUDY ADDS

• This study describes the implementation of the first UK multidisciplinary team pathways specifically designed for ‘difficult-to-treat’ patients for OSA who are intolerant of CPAP.

HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

• The findings support the role of structured multidisciplinary pathways in coordinating personalised non-CPAP treatment strategies for patients who fail first-line CPAP therapy and a broader transition away from a CPAP-only model of care towards integrated, patient-centred OSA management pathways.

Introduction

Obstructive sleep apnoea (OSA) is a common sleep disorder caused by repetitive obstruction of the upper airway during sleep. OSA is associated with worsening memory and depression,¹ increased cardiovascular and metabolic risks.^{2 3} Patients commonly present with loud snoring, excessive daytime sleepiness, unrefreshing sleep, morning headaches, impaired concentration and mood disturbance. Recurrent episodes of apnoea and hypopnoea lead to intermittent hypoxia and sleep fragmentation,⁴ which drives the sympathetic tone and, consequently, elevates heart rate and blood pressure.

Certain anatomical features, including a large neck circumference, a retrognathic or micrognathic mandible (lower jaw), an underdeveloped maxilla (upper jaw), enlarged tonsils, macroglossia and abnormal upper airway shapes, are associated with a higher risk of OSA.⁵ Other risk factors associated with higher OSA risk include obesity, male gender, female hormonal changes during pregnancy or menopause and age.^{2 5 6} The PALM classification⁷ describes OSA pathophysiology as a combination of anatomical and non-anatomical factors, including Pcrit (passive critical pressure), low arousal threshold, high loop gain and impaired upper airway muscle responsiveness.

Continuous positive airway pressure (CPAP) has been the current standard and first-line treatment for OSA, depending on symptoms and severity.⁸ CPAP requires the patient to sleep with an interface, either a nasal or oronasal mask that the patient places over their airway while asleep; the mask is connected with a tube to a bedside machine that produces compressed room air, which then enters the patient’s upper airway and prevents it from collapsing.⁹ Despite CPAP being successful in controlling airway patency when asleep, long-term adherence to CPAP therapy is limited. Studies report up to 75% at 3 months and less than 50% within a year.^{10 11} Early CPAP usage behaviour is highly predictive of long-term adherence, with non-adherence typically established within the first month.¹²

Treatment of OSA in adults is evolving as new therapies have been introduced in clinical practice in recent years. The National Institute for Health and Care Excellence (NICE) and European Respiratory Society (ERS) guidelines on OSA suggest various non-CPAP therapeutic options.^{8 13}

1. NICE recommends mandibular advancement devices (MADs) for patients with mild to moderate OSA and for some patients with severe OSA who cannot tolerate CPAP therapy.⁸

2. OSA is more prevalent in patients with obesity, and weight loss can improve OSA substantially. Previous trials indicate an average decrease in Apnoea-Hypopnoea Index (AHI) of 0.78 events/hour for every kilogram of weight lost.¹⁴ In some cases, a referral to the bariatric team may be indicated.

3. Positional therapy promotes sleeping in a lateral posture instead of a supine position or sleeping with the upper body elevated.¹⁵

4. Hypoglossal nerve stimulation (HNS) involves the implantation of a neurostimulator, which delivers an electric current to the distal branch of the hypoglossal nerve, thus preventing the tongue from prolapsing backwards.^16–18 A non-invasive approach to deliver electrical current to the hypoglossal nerve is transcutaneous electrical stimulation,¹⁹ where a transcutaneous patch is applied to the skin in the submental area, allowing the upper airway dilator muscles to be stimulated via the hypoglossal nerve, which keeps the lumen open during sleep.

5. Surgical options are typically recommended only for patients with a significant deviation from normal anatomy. Common surgical approaches include uvulopalatopharyngoplasty, tonsillectomy, Barbed-Wire Pharyngoplasty and maxillomandibular advancement.²⁰

6. Significant improvements have been demonstrated when switching OSA patients who poorly tolerated CPAP to BiPAP. Although CPAP remains the first-line therapy, BiPAP is particularly beneficial for individuals with overlapping conditions such as obesity or COPD-related hypoventilation and offers an effective second-line alternative.²¹

At Guy’s and St Thomas’ NHS Foundation Trust, a multidisciplinary team (MDT) pathway has been established since November 2020 for ‘difficult-to-treat’ OSA patients—those in whom first-line CPAP therapy has been unsuccessful. The MDT brings together all relevant specialties involved in OSA management, including respiratory physicians, sleep physiologists, ENT surgeons, oral and maxillofacial surgeons, dentists, neurologists and CPAP specialists. The MDT was initially held once every 3 months but transitioned to monthly meetings over the past 3 years. Patients were referred to the MDT if they had documented OSA, demonstrated CPAP intolerance or failure and required consideration of non-CPAP therapies. This approach facilitates the tailored prospective provision of evidence-based treatments and selection of alternative treatment pathways for patients in whom the first-line therapy was unsuccessful.²² This study aims to evaluate the implementation of an MDT approach to manage patients with OSA who do not tolerate CPAP, with particular emphasis on how alternative non-CPAP therapies are selected and coordinated across specialties. The study is not designed to determine whether MDT review directly improves patient outcomes.

Methods and patients

We conducted a retrospective cohort review following institutional review (registered number: GSTT/2025/17160) of all cases discussed in the ‘difficult-to-treat’ OSA MDT (DTT-OSA-MDT) meetings between November 2020 and March 2025 at Guy’s and St Thomas’s NHS Foundation Trust, London, UK (Sleep Disorders Centre). The referral criteria for the OSA MDT included: documented diagnosis of OSA, prior CPAP trial with intolerance or failure and referral for MDT discussion within the study period. No exclusion criteria were applied. The study size was, therefore, determined by the full MDT caseload; no formal sample size calculation was performed due to the retrospective service-evaluation design.

For each patient, demographic data (age, sex, body mass index (BMI)), OSA severity and symptom indices (AHI, Epworth Sleepiness Scale (ESS) and 3% Oxygen Desaturation Index (ODI) scores), comorbidities and relevant anatomical findings were extracted from the MDT documentation and EPIC medical records. AHI and ODI values were obtained from the most recent in-laboratory polysomnography or home sleep apnoea test performed before the first MDT discussion. ESS scores were taken from the patient-reported value at the most recent clinical consultation before MDT referral.

MDT discussions, therapeutic recommendations and clinical outcomes were documented for each case. Follow-up data for ESS, AHI, ODI and therapy adherence were extracted from post-treatment assessments where available, and the most recent follow-up values were used for analysis for all cases. In patients discussed at multiple MDT meetings, baseline values were taken from the most recent pre-referral assessment before the first meeting, while all treatment recommendations made across meetings were included. Additionally, the selection criteria for various treatment options discussed during the MDT meetings were documented and analysed.

Statistical analysis

Following testing for normality, normally distributed variables were summarised as mean±SD, and non-normally distributed variables as median with IQRs (first–third quartile). Subgroup analyses were undertaken to compare age, BMI and AHI across non-CPAP therapy subgroups were assessed using one-way analysis of variance (ANOVA). For patients with available longitudinal data, paired t-tests were used to compare baseline and follow-up ESS, AHI and ODI. Statistical significance was set at p<0.05. Sensitivity analyses were not undertaken due to sample-size limitations and variability in follow-up availability.

Handling of missing data

Missing data were handled using available-case analysis. No imputation was performed.

Patient and public involvement

Patients and members of the public were not involved in the design, conduct, reporting or dissemination plans of this research.

Addressing confounding

To minimise confounding, baseline and follow-up values were standardised by using the most recent pre-MDT assessment and the latest available post-treatment records, respectively. However, we acknowledge residual confounding inherent to the retrospective design in our analysis, including variable follow-up intervals and inconsistent data availability.

Results

We evaluated a cohort of 138 individuals, including 103 males. The mean age was 51 years (SD 17), and the mean BMI was 29.5 kg/m² (SD 5.1). All patients were referred to the Difficult-to-treat OSA MDT, with a mean AHI of 37.7 events/hour (SD 21.4), lying in the severe range. The predominance of middle-aged males with elevated BMI is consistent with recognised risk factors for OSA (table 1).

Table 1. Baseline patient demographics and OSA severity indices.

Patient characteristics (n=138)
Male: female	103:35
Age (years)	50.7±17.3
BMI (kg/m²)	29.5±5.1
AHI (/h)	37.8±21.4
ESS	12.4±6.1
3% ODI (/h)	36.1±22.6

BMI: n=126; AHI: n=133; ESS: n=113; ODI: n=64.

AHI, Apnoea-Hypopnoea Index; BMI, body mass index; ESS, Epworth Sleepiness Scale; ODI, Oxygen Desaturation Index; OSA, obstructive sleep apnoea.

Disease severity

The majority of patients in the cohort demonstrated severe OSA based on AHI classification, with only a small number in the normal or mild categories who were specifically referred from other Trusts for discussion at this meeting (table 2). The mean ESS score was 12.4 (SD=6.1), indicating a moderate level of daytime sleepiness. 70 patients had an ESS score over 10, indicating excessive daytime sleepiness. This distribution likely arises from a combination of factors, including suboptimal adherence to CPAP therapy. Poor compliance may result in insufficient treatment efficacy, thereby allowing OSA to remain severe and symptomatic. The mean ODI was 36.1/hour (SD=22.6), indicating the presence of significant nocturnal oxygen desaturation (table 1).

Table 2. Baseline OSA severity distribution.

Severity	Patients (%)
Normal (AHI<5)	4 (3.0)
Mild (AHI 5–15)	18 (13.6)
Moderate (AHI 15–30)	31 (23.5)
Severe (AHI≥30)	80 (60.2)
Missing	5 (3.8)

Four patients had AHI values within the normal range at referral but were nonetheless discussed at the MDT due to the presence of other sleep-related comorbidities (eg, comorbid insomnia, COMISA), previously abnormal AHI values, or intolerance to CPAP therapy.

AHI, Apnoea-Hypopnoea Index; CPAP, continuous positive airway pressure; MDT, multidisciplinary team; OSA, obstructive sleep apnoea.

Comorbidities

The cohort presented with a wide range of medical and psychiatric conditions (online supplemental table 1). Hypertension (n=40, 29.0%) and anxiety/depression (n=35, 25.4%) were the most frequently reported comorbidities, followed by ENT or upper respiratory issues that required previous surgery (n=24, 17.4%), heart disease (n=23, 16.7%) and other psychiatric or neurological conditions (n=17, 12.3%). A notable prevalence of nasal structural abnormalities (n=16, 11.6%) and sleep-related movement disorders, Periodic Leg Movement Disorder or Restless Leg Syndrome, were found (n=15, 10.9%). A smaller proportion of patients had metabolic, respiratory or neurological conditions, while a few (n=13, 9.4%) had no identifiable comorbidities.

Management recommendation

A variety of treatment recommendations were made, and 60 patients (43.5%) received more than one recommendation. Most frequently, the use of a MAD was recommended (n=35, 25.4%), followed by patients placed on the HNS pathway (n=34, 24.6%), and weight or metabolic management (n=32, 23.2%). Further diagnostic procedures, including drug-induced sleep endoscopy (DISE, n=28, 20.3%) and repeat sleep studies (n=28, 20.3%), were often required. Other notable options were Otorhinolaryngology (ENT) review (n=24, 17.4%), specific postural advice (n=22, 15.9%), CPAP therapy retrial (n=19, 13.8%), ENT surgery (n=15, 10.9%) and other ventilation therapies such as bilevel ventilation (n=14, 10.1%). DISE was required for determining patient eligibility for HNS or ENT surgical interventions. Among the 28 patients for whom the MDT recommended DISE, 23 subsequently proceeded to HNS. Most patients recommended for HNS had either undergone DISE before the MDT meeting or were advised to complete it in the MDT meeting (table 3).

Table 3. Recommended management options by the MDT.

Management options	Patients (%)
MAD	35 (25.4)
HNS	34 (24.6)
Weight Management*	32 (23.2)
Specific postural advice	22 (15.9)
CPAP retrial	19 (13.8)
ENT surgery†	15 (10.9)
Bilevel ventilation	14 (10.1)
Psychiatric/psychological treatment	9 (6.5)
Others	14 (13.0)
Further diagnostic procedures
DISE	28 (20.3)
Repeat sleep study	28 (20.3)
ENT review‡	24 (17.4)
Cardiac review	6 (4.4)

Most patients were recommended more than one management option.

^*Weight loss programme/referral to metabolic team

^†Includes maxillomandibular advancement (MMA), barbed-wire pharyngoplasty, tonsillectomy, thyroidectomy, nasal polyp removal, etc.

^‡Referral to ENT for further review and surgical options other than HNS

CPAP, continuous positive airway pressure; DISE, drug-induced sleep endoscopy; ENT, Ear, Nose and Throat (Otorhinolaryngology); HNS, hypoglossal nerve stimulation; MAD, mandibular advancement device; MDT, multidisciplinary team.

Age distributions were similar across recommended treatment groups, while the BMI was elevated across all therapeutic recommendations. Patients referred for weight management had significantly higher BMIs, while patients recommended for HNS had lower BMIs, consistent with established selection criteria of a BMI <32–35 kg/m².

Interventions, such as bilevel ventilation, surgery, HNS and CPAP retrial, were associated with higher baseline AHI values, aligning with the clinical rationale of escalating treatment for patients with more severe OSA. MAD, weight management and postural advice were generally recommended for patients with lower AHIs, reflecting a more conservative approach suited to milder cases (table 4).

Table 4. Baseline patient characteristics by therapy groups.

	Age (year)	BMI (kg/m²)	AHI (/h)
MAD	52.4±16.2	29.7±5.0	28.4±15.4
HNS	53.5±12.9	28.2±3.7	46.7±15.5
Weight management	54.5±15.2	34.9±4.3	43.0±22.4
Postural advice	50.5±16.2	29.3±4.0	24.5±16.4
CPAP retrial	51.6±16.4	29.8±4.2	49.9±21.5
ENT surgery	47.3±18.7	30.0±4.4	49.3±21.0
Bilevel ventilation	53.3±12.0	31.6±5.9	53.2±24.7
P value	0.922	<0.0001*	<0.0001*

Plus-minus values are mean±SD.

^*Significant findings (p<0.05).

AHI, Apnoea-Hypopnoea Index; BMI, body mass index; CPAP, continuous positive airway pressure; DISE, drug-induced sleep endoscopy; ENT, Ear, Nose and Throat (Otorhinolaryngology); HNS, hypoglossal nerve simulation; MAD, mandibular advancement device.

Follow-up

Among patients with follow-up data (ESS: n=13; AHI: n=29; ODI: n=12), clinically meaningful improvements were observed across all measured parameters following implementation of MDT-recommended non-CPAP therapies. AHI declined for an average of over 40%; 21 (72%) patients demonstrated a greater than 50% reduction in AHI compared with baseline. Of these, 13 patients (45%) fulfilled the SHER15 criteria for treatment success, defined as both a >50% reduction and a residual AHI of <15 events per hour.^{23 24} Similarly, ESS and ODI both showed significant reductions (table 5). Although follow-up data were not uniformly available for all outcome measures across every patient, the consistent improvement across physiological and patient-reported outcomes supports the effectiveness of MDT-guided non-CPAP therapy selection.

Table 5. Outcome measures: follow-up and baseline ESS, AHI and ODI.

Variables	Baseline	95% CI	Follow-up	95% CI	MD (%)	P value
ESS	12.4±6.1	11.3 to 23.6	8.3±5.8	4.8 to 13.1	4.1 (−32.9%)	0.016*
AHI (/hour)	37.8±21.4	33.7 to 71.3	21.1±16.4	14.9 to 36.0	−16.7 (−44.1%)	<0.0001*
3% ODI (/hour)	36.1±22.6	30.5 to 66.6	11.8±8.1	6.7 to 18.5	−24.3 (−67.2%)	<0.0001*

± values are mean±SD.

^*Significant findings (p<0.05).

AHI, Apnoea-Hypopnoea Index; ESS, Epworth Sleepiness Score; MD, mean difference; ODI, Oxygen Desaturation Index.

HNS follow-up data were available for nine patients. The mean usage was 4.9 hours/week (SD=2.8), with a median of 5.6 and an IQR of 3.4.

The available non-CPAP therapeutic options highlight the flexibility and value of the MDT in combining multiple therapeutic options tailored to individual patient characteristics (online supplemental figure 1).

Discussion

We have shown that evidence-based therapeutic recommendations for OSA can be effectively applied in patients who are intolerant of CPAP therapy within a structured clinical pathway, including multidisciplinary input where available. With an increasing prevalence of OSA in the general population, demand for tailored non-CPAP treatments is likely to increase. Given the high prevalence of comorbidities, including hypertension, depression/anxiety and heart diseases, a multidisciplinary approach can support optimising treatment selection, compliance and outcomes as well as provide wider super-regional accessibility to therapies.

We compare the demographic and comorbidity profile of our cohort of patients with OSA with that reported in broader OSA populations, particularly Pinto et al and Bakare et al for comparison, as they provide complementary perspectives: Pinto et al offer a focused clinical snapshot of comorbidities in a defined OSA group, while Bakare et al present large-scale case–control data highlighting demographic and modifiable risk factors.^{25 26} Both studies highlight common comorbidities typically associated with OSA, including high BMI, cardiovascular and metabolic conditions, and depression, providing a relevant reference for comparison.

Compared with these studies, our cohort demonstrated a lower prevalence of cardiometabolic comorbidities, including hypertension (29.0% vs 39%) and diabetes (9.4% vs 15%), and a lower mean BMI (29.5 kg/m²), reflecting referral patterns that favour younger, less medically complex patients for surgical or oral appliance interventions. In contrast, psychiatric comorbidities were more common within our cohort (depression 25.4% vs 19%), which may reflect the psychological burden of prolonged unsuccessful standard therapy. ENT surgical history and nasal structural abnormalities were also frequent, reflecting anatomical considerations specific to this referral group not captured in broader studies.

Age distributions were broadly similar across treatment groups, indicating that age was not a major determinant of therapy selection in this cohort. In contrast, BMI and baseline AHI had a clear influence on treatment allocation. Patients referred for weight management had the highest BMIs, whereas those recommended for HNS had lower BMIs, in line with established eligibility criteria (BMI <32–35 kg/m²).¹⁰ Patients referred for non-invasive bilevel ventilation, surgery, HNS or CPAP retrial had significantly higher AHIs, consistent with guideline recommendations to escalate therapy for more severe disease.^{13 27} In contrast, MAD and postural therapy were preferentially recommended for lower-AHI patients, reflecting a conservative approach suited to milder OSA.

The MDT meetings particularly clarified the selection criteria for MAD and HNS at Guy’s and St Thomas’. To be eligible for MAD, patients had to have good dental health, including healthy teeth and no active severe periodontal disease or dental decay (with regular dental and hygienist review). Additionally, an assessment of the upper airway and facial anatomy was essential to ensure anatomical suitability for MAD.⁸ Patients with generalised tonic-clonic seizures were excluded (online supplemental figure 1).

Eligible candidates for HNS included those with moderate to severe OSA, defined by an AHI between 15 and 65 events/hour, who were CPAP intolerant. A BMI of less than 35 kg/m² was used as a threshold, with preference for patients with a BMI below 32.¹³ DISE findings to assess eligibility for HNS included complete anteroposterior (A-P) collapse of the velum, tongue base and epiglottis in both supine and lateral positions (Inspire system).²⁸ HNS was also considered for patients with Down syndrome, in whom it has been shown to have certain advantages over CPAP therapy, for example, related to carer-burden, in previous studies.²⁹ Patients were excluded for HNS if DISE detected complete concentric collapse of the soft palate. Other exclusion criteria included complex sleep-disordered breathing, such as central or mixed sleep apnoea (>25% of respiratory events of central origin) or if OSA had a substantial positional component (normalisation of AHI in non-supine posture; in some cases, with greater than 50% reduction); in those cases, postural advice and conditioning were recommended¹⁸ (online supplemental figure 1).

Initial outcome data suggest significant symptom improvement among the cohort of patients with available data, as evidenced by reductions in ESS, AHI and ODI scores. Despite the observed AHI reduction, therapeutic control fell short of the widely used criteria for treatment success, either by normalising breathing at night or the SHER15/20 criteria.^{23 24} This suggests that while non-CPAP therapies may be effective in reducing disease burden, a residual degree of OSA frequently persists and must be monitored; it also indicates that most non-CPAP therapeutic options may still be less effective than CPAP therapy. Nevertheless, these non-CPAP options would be more beneficial than no therapy. HNS adherence figures indicate relatively good adherence among the small patient subset, which meets the threshold of ≥4 hours/day (p=0.174), comparable to adherence reported in previous trials.^{18 30}

We recognise that several non-CPAP therapeutic pathways, particularly HNS and MAD, follow protocolised selection criteria independent of MDT input. As such, clinical outcomes are likely to reflect the efficacy of the intervention itself rather than the MDT process.

The generalisability of this MDT model should be considered in the context of resource and workforce constraints. The structured, multispecialty MDT described in this study reflects a tertiary centre setting with access to a broad range of expertise across several specialties, which may not be readily available in many UK sleep centres. As such, implementation of similar MDT frameworks on a regular basis may be challenging in routine clinical practice; however, it may also offer super-regional accessibility to patients whose local centres do not provide tertiary care. Importantly, the management of patients with OSA who do not tolerate first-line therapy should continue to rely on established, protocol-based pathways, which remain the cornerstone of care across all settings. These pathways, including stepwise escalation to non-CPAP therapies based on clinical criteria, can be effectively delivered within standard sleep services. In this context, MDT input may be most valuable in selected complex or refractory cases, where coordination across specialties and individualised decision-making are required, rather than as a universal requirement for all patients.

Limitations

Our study has several limitations. First, its retrospective design relies on existing clinical documentation from MDT discussions and clinic letters, which may be incomplete or inconsistent. Follow-up data for treatment outcomes and adherence were only available in a subset of patients, limiting the generalisability of our findings. More structured follow-up and patient-reported outcome tracking are needed for definitive conclusions. Medication data on the follow-up cohort are incomplete and were not part of this service review, particularly for psychiatric and neurological comorbidities. Second, there was a default selection bias in the studied cohort of patients, since the cohort had failed first-line treatment. This may not reflect the broader population of patients with OSA. Third, the recommendation of multiple and stepwise interventions for a single patient (eg, weight loss and postural advice) unavoidably introduces confounders of outcomes, making it difficult to attribute observed improvements to any single therapy. Fourth, the timing of follow-up assessments was not standardised, which may affect the comparability of outcome measures such as ESS, AHI and ODI. Importantly, as this study did not include a comparator group of patients managed without an MDT pathway, it is not possible to determine whether MDT review itself improves outcomes compared with standard protocol-driven care. Finally, the study was conducted in a single MDT at the Sleep Disorders Centre at Guy’s and St Thomas’ NHS Foundation Trust, and results may not be generalisable to community or non-specialist settings. However, the MDT extends regularly to five large sleep centres in the Greater London area, combining clinical expertise and access to NHS healthcare resources, thus providing outstanding specialist input.

Conclusion

In this cohort of patients with OSA who failed usual care, a MDT approach facilitated the selection and delivery of tailored, evidence-based non-CPAP therapies, leading to observable improvements in OSA indices and patient-reported outcomes and symptoms. While there is frequently a residual AHI of relevance with non-CPAP therapies, the reductions in AHI, ESS and ODI suggest that MDT-guided management is shown to meaningfully reduce disease burden in complex cases. However, this needs to be systematically and prospectively assessed to define the health economic benefit for such an approach. Furthermore, due to the retrospective design of this study, the incomplete follow-up and overlapping of treatment recommendations, causality cannot be firmly established. Future studies need to address core clinical outcomes systematically to help open a new chapter in the transition away from a CPAP-only sleep service and assess whether wider accessibility may improve patient care across the NHS.

Data availability statement

Data are available upon reasonable request.