Partial update of the German S3 Guideline Sleep-Related Breathing Disorders in Adults

Steering Committee

• Prof. Dr. med. Boris A. Stuck, Marburg

• Dr. med. Alfred Wiater, Cologne

1 Abstract

A review of the recommendations in the currently applicable version of the chapter “Sleep-related breathing disorders in adults” in the guideline “Non-restorative sleep/sleep disorders” conducted by the guideline steering group identified a number of recommendations and chapters requiring revision or updating in the light of new scientific or clinical evidence. These included selected chapters on diagnosis (clinical examination, polysomnography, respiratory polygraphy, and diagnosis with reduced and alternative systems) and treatment (positional therapy, surgical procedures) of sleep-related breathing disorders, as well as the chapters on obstructive sleep apnea (OSA) and dementia and on sleep-related hypoventilation/sleep-related hypoxemia. These chapters were updated according to the methodology of an S3 guideline. The remaining recommendations and chapters, on the other hand, retain their validity for the time being.

With regard to the diagnosis of sleep apnea, a number of recommendations on clinical examinations, as well as on respiratory polygraphy and polysomnography, have been specified more precisely and modified. A chapter on diagnosis has been supplemented with alternative systems: tonometry-based diagnostics has been included as an alternative for the diagnosis of sleep apnea. Positional therapy for positional OSA and tonsillectomy with uvulopalatopharyngoplasty are given a higher level of recommendation based on new randomized trials. Recommendations on OSA and dementia can no longer be made due to a lack of evidence; the recommendations on sleep-related hypoventilation/sleep-related hypoxemia have been specified with greater precision.

2 Concept of the update

The existing guideline “Non-restorative sleep/sleep disorders”—chapter “Sleep-related breathing disorders in adults”—became due for revision 3 years following its publication on 22.12.2019. After a review of the guideline content by members of the previous steering committee and the authors of the original guideline, it became evident that large parts of the guideline remain valid—thereby making a complete update unnecessary—and that its validity can instead be extended for a further 2 years. However, a number of chapters were identified as requiring revision due to new evidence. In view of the size and subject diversity of the guideline, the decision was taken not to update it in its entirety, but rather to revise only parts of the guideline. Therefore, the following update contains the revised guideline chapters. Accordingly, the unrevised chapters remain valid and can still be found under guideline Registry No. 063-001. A complete revision of the entire guideline is planned in the future. Details on the selection of chapters to be updated and the methodology used to this end can be found in the guideline protocol (see Appendix). It should be noted at this juncture that the guideline relates only to adults.

3 Diagnosis

3.1 Clinical examination

The aim of the clinical examination should be to identify anatomical changes in the upper respiratory tract or the region of the viscerocranium that could be (in part) responsible for the onset of obstructive sleep apnea (OSA). The clinical examination should include the nose, oral cavity, and pharynx, as well as craniofacial morphology. The clinical examination should be extended in cases where symptoms are reported or relevant pathologies are suspected in these regions [1, 2]. To this end, it may be necessary to consult relevant specialists (ear, nose, and throat specialists; oral and maxillofacial surgeons; orthodontists; specialized dentists). For the differential diagnostic planning of surgical interventions or the use of a mandibular advancement splint (mandibular advancement device, MAD), an examination of the pharynx under anesthesia (drug-induced sleep endoscopy [DISE]) is of particular importance in order to describe the collapse pattern, localization, and extent of relevant stenosis [3, 4].

Recommendations.

Modified:

• An examination of the oral cavity and pharynx is of great importance and shall be performed (evidence level 5, grade A recommendation).

• If treatment with a mandibular advancement device is being considered, an assessment should be made of possible mandibular protrusion, and dental status should be recorded (evidence level 5, grade A recommendation).

• The diagnostic work-up of OSA should include a clinical assessment of craniofacial morphology for orientation (evidence level 5, grade A recommendation).

• To assess airflow-related nasal structures, a clinical examination of the nose should be performed (evidence level 5, grade B recommendation); this may also include an endoscopic assessment (evidence level 5, grade C recommendation).

New:

• To determine body mass index (BMI), both height and body weight should be obtained (evidence level 5, grade A recommendation).

The recommendations on clinical examination were adopted with strong consensus (18/18).

3.2 Diagnostic polysomnography

Supervised cardiorespiratory polysomnography (PSG) represents the most important instrument and reference in sleep medicine diagnostics in the sleep laboratory. It involves recording the physiological signals required for a quantitative assessment of sleep, sleep disorders, and diseases associated with sleep according to the International Classification of Sleep Disorders—Third Edition (ICSD‑3; see Table 1).

Table 1 Channels recommended for cardiorespiratory polysomnography. The table shows the function to be investigated, the associated biosignals, and the required technique as well as its technical specifications in relation to the optimal sampling rate and filter settings

Sleep stage classification largely corresponds to the older classification by Rechtschaffen und Kales [5, 6]. This reduces ambiguities and increases reliability [7]. A chapter on central nervous activation (arousals) adopts the definitions given in an earlier recommendation paper [8]. Other chapters define the recording and evaluation of electrocardiogram (ECG) parameters [9] and leg movements. Motor patterns, such as periodic leg movements, bruxism, and rapid eye movement (REM) sleep behavior disorders, are precisely defined [10]. With regard to nocturnal breathing disorders, definitions are given for various types of apnea and hypopnea. Esophageal pressure measurement is the reference method to determine increased respiratory effort in the case of obstruction and to differentiate obstructive from central respiratory disorders. However, it is not widely available. A PSG-based algorithm that includes flattening, respiratory effort, breathing patterns, arousal, and sleep stages is also highly accurate in the differentiation of respiratory disorders.

Inductance plethysmography is recognized as a noninvasive method yielding comparable results [11]. In order to detect hypoventilation during sleep, the CO₂ concentration needs to be continuously determined. The method most widely used to this end is transcutaneous measurement of the partial pressure of carbon dioxide (tcPaCO₂) [12]. PSG also includes recording body position and a finely synchronized video recording of the sleeping subject [13]. The American Academy of Sleep Medicine (AASM) Manual was updated to a limited extent in 2012 (version 2.0) and 2014 (version 2.1), as well as in 2015 (version 2.2) and again in 2016 (2.3), to take new findings into consideration [14,15,16,17]. Further changes followed in 2017 (version 2.4) and 2018 (2.5). In 2016, a chapter on the outpatient diagnosis of sleep apnea was included, setting out evidence-based parameters for recording and evaluation [18, 19]. This also includes a subchapter on outpatient diagnosis of sleep apnea using peripheral arterial tonometry, together with the specific parameters of this technique [18, 19].

Monitored PSG makes it possible to investigate and quantitatively record sleep disorders involving changes in physiological parameters. With current computer-assisted technology, PSG sets manageable requirements in terms of equipment. It requires specially trained sleep medicine personnel to perform the measurements and evaluate the biosignals. Further training courses and qualifications in sleep medicine are established for medical technical personnel, psychologists, and natural scientists, as well as for physicians. Medical specialists in certain specialties have the opportunity to gain additional training in “sleep medicine” according to the relevant training regulations of the German State Chambers of Physicians. In addition, there is a defined further training program on the diagnosis and treatment of sleep-related breathing disorders (SRBD) according to the Fifth Book of the German Social Code. Physicians and natural scientists are able to obtain proof of qualification as a somnologist from the German Society for Sleep Research and Sleep Medicine (Deutsche Gesellschaft für Schlafforschung und Schlafmedizin, DGSM).

The AASM Scoring Manual [13, 14] enables classification into the following stages: awake, REM, N1, N2, and N3. There are national and international recommendations on the equipment and staffing of a sleep laboratory; compliance with these recommendations is a prerequisite for the accreditation of sleep laboratories by the sleep medicine societies [11].

Table 2 provides an overview of the evidence-based data on PSG. The validity and reliability of visual scoring has been demonstrated and fulfills the current requirements for the quality of visual scoring of biosignals [7, 20]. The report on sleep medicine findings must document whether the PSG was recorded and scored according to Rechtschaffen and Kales criteria [5] or AASM criteria [13, 14]. The AASM guidelines are revised approximately every 2 years, with the latest revision undertaken in 2018 (version 2.5) [21].

Table 2 Meta-analyses on cardiorespiratory polysomnography in the monitored sleep laboratory

Recommendations.

Reviewed and retained:

• PSG in the sleep laboratory monitored by personnel qualified in sleep medicine is recommended as the gold standard and reference method (evidence level 2a, grade A recommendation).

• PSG should be performed in line with current recommendations. This includes recording sleep electroencephalogram (EEG), electrooculogram (EOG), electromyogram (EMG), ECG, respiratory flow, snoring, respiratory effort, oxygen saturation, body position, and a video (evidence level 2a, grade A recommendation).

New:

• PSG should be performed in the case of inconclusive respiratory polygraphy results, persistent suspicion of SRBD despite normal respiratory polygraphy, in the differential diagnosis of disorders in sleep medicine, and/or in the case of suspected comorbid disorders relevant to sleep medicine (evidence level 2a, grade B recommendation).

The recommendations on diagnostic PSG were adopted with strong consensus (16/18).

3.3 Respiratory polygraphy for sleep-related breathing disorders

Simplified portable systems are available for the diagnosis of sleep-disordered breathing ([24,25,26]; see Table 3).

Table 3 Studies on respiratory polygraphy for sleep apnea

Portable diagnostic systems are classified according to the number of channels recorded. These are predominantly systems with between four and six channels that do not record sleep EEGs (synonym: respiratory polygraphy systems).

Respiratory polygraphy systems with an adequate selection of biosignals, very good signal acquisition, and very good signal processing can reduce the number of false-positive diagnoses [18, 19]. Preselecting patients by means of a targeted patient history can significantly increase pretest probability, as well as the number of false-positive diagnoses. Using respiratory polygraphy, it is possible to make a basic distinction between OSA and central sleep apnea (CSA). Validations are available for some systems, but these are not yet sufficient for a high level of evidence.

Respiratory polygraphy systems for the diagnosis of sleep apnea generally record airflow with a thermistor or nasal pressure transducer, respiratory effort by means of inductance plethysmography, oxygen saturation with appropriate pulse oximetry (averaging with sufficiently high temporal resolution), pulse rate, and body position [19]. With the method used in the SCOPER system, the number of channels becomes irrelevant, with the focus being placed on measuring functions (S = sleep, C = cardiovascular, O = oximetry, P = position, E = effort, R = respiratory; [19]). There are several grades of quality for each function. Sleep can be assessed on the basis of actigraphy or other surrogate parameters, and, depending on the question being asked, does not need to be inferred from a sleep EEG. The SCOPER system is used today to grade respiratory polygraphy systems.

The respiratory polygraphy is evaluated according to the current rules of PSG [21] to allow visual scoring and processing of artifacts. It is essential to indicate in the documentation that visual scoring has been performed [31]. For the sampling rates and other technical specifications of polygraphy systems, the evidence-based recommendations for PSG are used (see Table 1).

Respiratory polygraphy systems for the diagnosis of SRBD should be used by medical specialists trained in sleep medicine who can record and evaluate pretest probability, symptomatology, and comorbidities [31]. In Germany, further training in sleep medicine according to the BUB guideline is a prerequisite for billing respiratory polygraphy according to a standardized evaluation scale (einheitlicher Bewertungsmaßstab, EBM). Respiratory polygraphy systems can be used to diagnose OSA, but not in the case of comorbid pulmonary, psychiatric, neurological, and neuromuscular disorders, nor in the case of other concomitant sleep disorders such as CSA, periodic limb movement disorder (PLMD), insomnia, circadian sleep–wake rhythm disorders, and narcolepsy [33, 34]. Respiratory polygraphy systems enable a distinction to be made between CSA and OSA. In cases where hypopneas predominate, polygraphy systems do not always enable a definitive differentiation to be made between CSA and OSA and are not validated for this purpose. Due to the absence of EEG channels (e.g., to measure sleep time and waking responses), respiratory polygraphy systems are inferior to PSG in that their assessment of the severity of SRBDs can be less accurate, they cannot reliably rule out an SRBD, and they are unable to diagnose possible differential diagnoses of an SRBD. Comparability of the number of hypopneas between respiratory polygraphy and PSG is most likely to be achieved if the following definition of hypopnea is chosen: a reduction in oronasal airflow of between 30 and 90% of baseline at an at least 4% drop in oxygen saturation [16, 35]. Physiological irregularities in breathing rhythm during the sleep–wake transition (also known as sleep-onset apneas) can be incorrectly classified as sleep apnea in the absence of EEG analysis and lead to false-positive results.

It has been shown in patients with chronic cardiovascular disease that medical history is not sufficient to detect sleep-related disorders. This group of patients requires respiratory measurement parameters to be recorded. There are no data as yet on whether systematic screening has an effect on long-term prognosis [29, 30].

Recommendations.

Modified:

• Respiratory polygraphy systems should be used to diagnose OSA if pretest probability is high (evidence level 2a, grade A recommendation).

• Respiratory polygraphy should not be used as the sole mode of detection and exclusion of OSA in the case of low pretest probability, nor in suspected concomitant sleep disorders or sleep-related comorbidities (evidence level 5, grade A recommendation).

• Respiratory polygraphy for the diagnosis of SRBDs should be performed by specialists qualified in sleep medicine who are able to record and evaluate pretest probability, symptoms, and comorbidities (evidence level 5, grade A recommendation).

• The recorded signals should be visually scored by trained personnel. Scoring by automatic scoring alone is currently not recommended (evidence level 2a, grade A recommendation).

• Cardiorespiratory PSG should be performed as the diagnostic method to exclude SRBDs; respiratory polygraphy is not sufficient for this purpose (evidence level 2a, grade A recommendation).

• Respiratory polygraphy is not sufficient for the diagnostic work-up of ventilatory failure, hypoventilation during sleep, or obesity hypoventilation syndrome; daytime blood gas parameters (paCO₂ and HCO₃₋) as well as CO₂ during sleep are needed for this (evidence level 2a, statement).

• Disease course and treatment follow-up can be performed using respiratory polygraphy. In patients with questionable treatment success, as well as in patients at cardiovascular risk and patients with other disorders that impair sleep, PSG follow-up can be performed without prior respiratory polygraphy (evidence level 2b, grade C recommendation).

The recommendations on respiratory polygraphy for SRBDs were accepted with strong consensus (the recommendation on the diagnostic work-up of ventilatory failure by 10/11, and the remaining recommendations by 10/10).

3.4 Diagnosis of sleep-related breathing disorders with reduced systems

Systems that record only between one and three channels (pulse oximetry, long-term ECG, actigraphy, oronasal airflow measurement) yield up to 17% false-negative and up to 31% false-positive findings [27]; this explains why their use is not recommended for the definitive diagnosis or exclusion of SRBDs.

Recommendations.

Reviewed and retained:

• Respiratory polygraphs that meet fewer than all the abovementioned criteria can provide an indication of whether an SRBD is present. They should not be used as the sole method to diagnose SRBDs (evidence level 2b, grade A recommendation).

Modified:

• In the case of cardiovascular diseases that increase the risk of SRBDs (e.g., arterial hypertension, heart failure, atrial fibrillation, cerebrovascular disease) in the absence of all typical symptoms, single- or dual-channel registration is possible as a screening test. If this registration prompts suspicion of OSA, further diagnostic evaluation with respiratory polygraphy or PSG should be carried out (evidence level 2b, grade A recommendation).

The recommendations on the diagnosis of SRBDs with reduced systems were adopted with strong consensus (14/17).

3.5 Diagnosis of sleep-related breathing disorders using alternative systems

Recent studies of varying quality point to possibilities for monitoring SRBDs using alternative systems that cannot be classified as classic PSG, respiratory polygraphy, or reduced-signal systems. These alternative systems are based on the SCOPER criteria [19], which specify that the physiological entities of sleep, cardiovascular parameters, oxygen saturation, position, respiratory effort, and airflow are recorded, either directly or indirectly, in varying quality. Systems with new sensor technology, such as peripheral arterial tonometry [28], breathing sound/snoring sound analysis [36, 37], noncontact high-resolution detection of body motion using radar technology [38, 39], and bed sensor technology [40, 41], are able to extract multiple entities from one signal by means of advanced biosignal analysis. A prominent example that has been published with high evidence is the successful extraction of airflow and respiratory effort from a modified finger pulse (peripheral arterial tonometry) for the diagnosis of sleep apnea [28]. Smartphone- and smartwatch-based systems are picking up on these technical developments and are generally not approved as medical devices.

Recommendations.

New:

• Systems using peripheral arterial tonometry demonstrate good evidence for the diagnosis of sleep apnea and should be used analogously to respiratory polygraphy in the case of high pretest probability in order to confirm the diagnosis and determine the severity of SRBDs (evidence level 2a, grade B recommendation).

• Systems based on smartphone and smartwatch technology with internal and external sensors are not yet approved as medical devices; moreover, validation studies are lacking or only available in exceptional cases, meaning that these devices are not currently recommended (evidence level 5).

The recommendations on the diagnosis of SRBDs with alternative systems were adopted with strong consensus (14/15).

4 Obstructive sleep apnea syndrome

4.1 Obstructive sleep apnea and dementia

Two prospective cohort studies with 2636 older men and 298 older women, respectively, without cognitive impairment at baseline investigated the effect of OSA on brain performance. In the men, a significantly greater loss of cognitive performance was found after a mean of 3.4 years in the presence of OSA with additional hypoxemia [42]. Among the women with OSA and hypoxemia, the risk of developing mild cognitive impairment (MCI) or dementia was increased by a factor of 1.85 (95% confidence interval [CI] 1.11–3.08) over a mean of 4.7 years after adjustment for other risk factors [43]. The risk for sleep apnea patients to develop dementia is increased if the apolipoprotein E ɛ4 allele is absent [44]. Furthermore, a significant decline in instrumental activities of daily living has been demonstrated in older women with untreated sleep-disordered breathing [45].

4.1.1 Treatment of obstructive sleep apnea in individuals with dementia

In a small randomized controlled trial, individuals with mild or moderate dementia and OSA (apnea–hypopnea index [AHI] > 10/h) were treated with continuous positive airway pressure (CPAP) or subtherapeutic CPAP therapy. Although acceptance of CPAP therapy was initially high, 25% of patients dropped out of the study after randomization [46, 47]; the dropout rate after the randomized study was also high [48].

After 3 weeks of treatment, patients receiving CPAP therapy showed a clinically nonrelevant improvement in daytime sleepiness compared with patients receiving subtherapeutic CPAP therapy (∆ Epworth Sleepiness Scale [ESS] CPAP versus subtherapeutic CPAP: −1.6) [47]. A significant improvement was seen in various neuropsychological test scores [46].

CPAP therapy can improve daytime sleepiness [47], sleep profile [49], and cognitive performance [48] only in selected patients with mild or moderate dementia in nonrandomized trials. Long-term CPAP therapy is possible in selected patients with mild or moderate dementia and OSA [46,47,48,49].

The randomized studies conducted to date have not been able to demonstrate that CPAP therapy relevantly improves symptoms and cognitive performance in patients with mild to moderate dementia. The studies were too small to demonstrate effects of this kind (Table 4). Therefore, it is currently not possible to make specific recommendations on the treatment of sleep apnea in dementia patients.

Table 4 Treatment of patients with OSA and dementia

Pharmacological treatment of dementia using donepezil can lead to a relevant improvement in obstructive sleep apnea [50].

4.1.2 Nocturnal positive-pressure ventilation

The recommendations in the existing chapter on nocturnal positive-pressure ventilation remain valid. This section merely specifies with greater precision the statement on CPAP/APAP initiation without PSG monitoring, which is now as follows: “With regard to managing therapy initiation, there is a growing body of RCTs (randomized controlled trials) showing that in certain subgroups of patients, CPAP/APAP discontinuation can be carried out on an outpatient basis at a sleep medicine center without PSG monitoring in a sleep laboratory.”

4.2 Non-CPAP methods in obstructive sleep apnea

4.2.1 Positional therapy

The term “position-dependent” or “positional” OSA is usually used when the AHI is at least twice as high in a supine position compared to other positions [51]. The prevalence of positional sleep apnea is estimated to be 25–30% of all patients with OSA. The probability that a patient has positional sleep apnea diminishes with increasing severity of sleep apnea and obesity [51].

In principle, positional therapy (preventing a supine position) has therapeutic potential especially in patients whose respiratory disorders manifest only in the supine position or whose AHI in other body positions is lower or does not require treatment. The methods or devices used for positional therapy differ significantly in terms of design and the reliability with which the supine position is prevented. A distinction can be made in positional therapy between passive methods (e.g., the so-called tennis ball technique, as well as special pillows and backpacks) and active methods (positional trainer, vibration alarm devices controlled by a special algorithm) [52]. Here, as with other devices for sleep apnea treatment, the problem of compliance and adherence also arises.

A recent meta-analysis includes three randomized studies comparing active procedures using a positional trainer versus CPAP therapy. However, it is important to note that these meta-analyses investigated primarily only one system. Two of these studies showed that CPAP resulted in a greater reduction in AHI, with a mean difference of 6.4 events per hour (95% CI 3.00–9.79) compared to positional therapy [52]. Positive effects on daytime sleepiness were comparable under positional therapy and CPAP [52]. It was not possible to draw conclusions regarding adverse events, quality of life, or long-term effect [52]. In a recent randomized crossover study in which each treatment lasted 6 weeks, the mean nocturnal duration of use was approximately 1 h longer with a positional trainer than with CPAP therapy [51]. No severe adverse events occurred in either group. The rate of adverse events was lower under positional therapy compared to automatic CPAP therapy (4.2% versus 18.8%) [51]. Patients receiving positional therapy reported pain in the back, shoulder, and neck (3.4%, 4/117), as well as abdominal skin irritation (< 1%, 1/117). While both treatments significantly improved daytime sleepiness, neither achieved a relevant improvement in sleep-related quality of life (as measured with the Functional Outcomes of Sleep Questionnaire, FOSQ) [51].

In another randomized trial, active positional therapy was equivalent to treatment with a mandibular advancement device (MAD) for positional OSA [53], while in yet another study, a combination of the two therapies was shown to be significantly more effective than the individual therapies [54].

Although active positional therapy was equal to passive positional therapy in terms of preventing a supine position, it showed significantly better compliance, mean disease alleviation, sleep quality, and quality of life [55]. In a long-term study on active positional therapy (positional trainer), the discontinuation rate within 1 year was 16% (9/58) [56]. Active positional therapy, therefore, appears to represent a realistic long-term treatment method for positional OSA. This was not the case for most passive positional therapies (e.g., tennis ball method). Table 5 provides an overview of the studies presented here.

Table 5 Positional therapy in patients with OSA

Recommendation.

Modified:

In patients with mild to moderate positional OSA, positional therapy with validated systems should be considered (evidence level 1b, grade B recommendation).

The recommendation on positional therapy was adopted with strong consensus (9/11).

4.2.2 Surgical treatment methods

In terms of surgical treatment, a distinction is made between resective and nonresective surgical methods and procedures that modify the facial skeleton (osteotomies). Resective methods include all surgical procedures aiming to remove or correct anatomical obstructions which are diminishing the airflow. This goal is achieved by resection or by modifying the upper airway with different kinds of surgical techniques. Overall, resective procedures carry a peri- and postoperative risk, but this risk is described as low in more recent studies and depends on the procedure [62]. Tracheotomy is another procedure that reliably eliminates OSA [63] but should be seen as a last choice. In general, surgical treatment forms are more difficult to evaluate to a high level of evidence, since surgical techniques are individually selected depending on the anatomy and function of the upper airway. This makes standardization challenging. Moreover, only a few surgical centers master the entire spectrum of the different surgical approaches [64]. Due to the nature of many surgical techniques, blinding is not feasible. Nevertheless, an ever-growing number of controlled and randomized studies comparing surgical treatment methods with CPAP, placebo, or waiting are available. Data on long-term effects are not available for all the procedures presented in this manuscript.

A sleep diagnosis is needed if surgical procedures are used in OSA patients. The predictors of surgical success—and thus the selection criteria for selecting the appropriate surgical procedure in OSA patients—need to be elaborated separately for each intervention. They can differ significantly and are not available for all surgical procedures. Obesity is a negative predictor but has different and to some extent as yet unknown threshold values, depending on the intervention.

Surgeries to improve nasal breathing need to be considered in a different context since they generally do not improve the breathing disorder, but they do reduce daytime sleepiness, snoring, as well as the required CPAP pressure, and improve general acceptance of CPAP therapy [65]. Since the underlying studies were conducted in patients with existing nasal airway impairment—so due to an existing rhinological indication—surgical procedures to improve nasal airway function are limited to this population.

Tonsillectomy, uvulopalatopharyngoplasty, and modifications.

Two randomized studies evaluated the combination of tonsillectomy (TE) and uvulopalatopharyngoplasty (UPPP) in groups of patients that had exclusively or predominantly tonsils which were obstructing the oropharynx. In these groups, surgery was highly significantly superior to a 6-month waiting period in terms of reduction in respiratory events. For example, an AHI reduction was seen in 60% (surgery group) versus 11% (waiting group). Furthermore, treatment success (Sher criteria: reduction of AHI > 50% to an AHI < 20) was shown in 59% (surgery group) vs. 6% (waiting group) [66]. The AHI was reduced from 35.7 ± 19.4 to 28.6 ± 19.3 in the control group and 33.7 ± 14.5 to 15.4 ± 14.1 in the surgical group [67]. Accordingly, a systematic review of these studies also demonstrated significant superiority of TE with UPPP in terms of AHI, snoring, and daytime sleepiness compared to an untreated control group of patients with clinically diagnosed isolated oropharyngeal obstruction [68].

Positive predictors of treatment success include defined anatomic factors (e.g., tonsil size), clinical scoring systems derived from these, BMI, and severity of sleep apnea [68, 69]. For patients with very large tonsils, TE alone can be helpful and effective [70]. For unselected patients, the efficacy of TE combined with UPPP has been shown only with a lower level of evidence to date [68]. A significant and long-term improvement in daytime sleepiness, physical and mental wellbeing, and quality of sleep has recently been demonstrated for TE-UPPP in randomized controlled trials [71,72,73,74]. Cohort studies achieved normalization 3 months after successful UPPP of serum leptin and nitric oxide levels, as well as endothelium-dependent vasodilation [75, 76]. Furthermore, tumor necrosis factor‑α (TNF-α) and interleukin‑6 (IL-6), as inflammatory markers, were also significantly reduced 3 months after UPPP with septoplasty compared to a healthy control group [77]. In an epidemiological cohort study of 444 patients, OSA mortality was significantly reduced by UPPP compared to mortality in an untreated control group. This result of mortality was comparable to that of CPAP therapy [78].

Whether the initial effect of surgery diminishes over time is the subject of debate due to inconsistent data [79, 80]. However, in contrast to previous published data, the results of more recent 24-month data do not support an attenuation of the effect over time, at least in the case of TE-UPPP [81].

Persistent side effects, such as voice changes, foreign body sensation, and difficulty in swallowing, are possible depending on the surgical procedure [62, 68]. Therefore, relevant reviews recommend TE-UPPP in patients with mild to moderate OSA, especially if they do not tolerate CPAP therapy and have suitable anatomy (tonsillar hyperplasia with clinically suspected oropharyngeal obstructions), since it is superior to “no therapy” and has acceptable side effects [70]. Whether UPPP offers a benefit compared to TE alone cannot be determined on the basis of the currently available data.

The value of the numerous modifications of UPPP (e.g., uvulopalatal flap, Z‑palatoplasty, relocation pharyngoplasty, Han’s UPPP, lateral pharyngoplasty) has not yet been elucidated. Hence, corresponding studies should be performed [82]. A systematic review was able to show superiority over UPPP only for expansion sphincter pharyngoplasty in the case of suitable anatomy [83]. A meta-analysis showed that anterior palatoplasty significantly improves AHI, snoring, and daytime sleepiness in patients with mild OSA [84].

Laser-assisted uvulopalatoplasty (LAUP) reduces AHI to below 10/h in 50% of cases, while postoperative side effects are seen in up to 60% of cases [85]. Similar results were seen in the long-term over a period of 11 years on average [86]. Therefore, LAUP is also not recommended in international publications.

Hypoglossal nerve stimulation.

In the past decade, stimulation of the hypoglossal nerve (HN) has become firmly established in the treatment of OSA. Stimulation of the nerve fibers generates a contraction of the genioglossus muscle, which results in an active opening of the upper airway [87]. There are currently different methods of hypoglossal nerve stimulation (HNS), which, due to their clinical results, also need to be considered in a differentiated manner [88, 89]. The largest amount of evidence is available for the method of selective breathing cycle-controlled HNS. Extensive clinical data are available from a controlled multicenter study with randomized treatment withdrawal and also from a 5-year follow-up period. Many other controlled studies as well as routine care studies exist [90,91,92,93,94,95]. All studies show sustained effects with selective breathing cycle-controlled HNS in reducing respiratory events in patients with moderate to severe OSA. A high usage time could be detected and led to a significant improvement in health-related quality of life. Patient age and previous upper airway surgery have an impact on the efficacy of HNS. Procedure-related morbidity can be described as low [88]. The rate of revision surgery over a period of 5 years in the first clinical trials was 6% [89, 90] and could be reduced to 1% according to the data of the subsequent large registry study (ADHERE registry) [88].

Selective breathing cycle-controlled HNS is only beneficial in patients without tonsillar hypertrophy and moderate to severe OSA. Since previous studies showed that complete concentric collapse at the level of the soft palate during drug-induced sleep endoscopy (DISE) was associated with a much higher nonresponder rate, patients showing this obstruction pattern at the velum need to be excluded [96]. Former studies showed that treatment response was dependent on BMI and AHI. Although the STAR study primarily included patients with a BMI < 32 kg/m² and an AHI between 20 and 50/h, results of cohort studies and the international registry study (ADHERE) suggest that the indication criteria could be broader (BMI of 35 kg/m² instead of 32 and AHI of between 15 and 65/h instead of 20 and 50/h) [87, 88, 90, 97, 98].

Targeted hypoglossal neurostimulation (continuous non-breathing-dependent stimulation of the hypoglossal nerve) consists primarily of upper airway stabilization and not—in contrast to breathing-controlled selective stimulation—active opening [89, 99]. Targeted hypoglossal neurostimulation can be considered in patients with an AHI of 20–65/h and a BMI of up to 35 kg/m², as well as in patients with concentric collapse of the upper airways.

A success rate according to Sher criteria (≥ 50% reduction in AHI and AHI < 20/h under treatment) was achieved at 12 months for the individual systems in 72% of 211 patients receiving selective breathing cycle-controlled HNS and in 77% of 13 patients receiving targeted hypoglossal neurostimulation [89]. There are no randomized comparative studies addressing the different stimulation methods or comparing with other therapies (e.g., MAD) [89].

Osteotomies.

Osteotomies to advance the maxilla and mandible (maxillomandibular advancement) enlarge the pharyngeal airway and increase the pharyngeal muscle tone. Both effects synergistically reduce the collapsibility of the upper airway. Osteotomies can be highly effective in the treatment of OSA, not only in the case of congenital malformations (e.g., Pierre Robin sequence, Crouzon syndrome, Apert syndrome), but also for specific anatomical features of the pharyngeal airway such as microgenia (small mandible), mandibular retrognathia (retroposition of the mandible in relation to the anterior skull base), or the associated narrow sagittal craniofacial structures. Osteotomies can also be performed in normognathic patients. Advancement by 10 mm is considered necessary. In a meta-analysis of 627 patients, a substantial improvement in AHI of 86% was reported, and an AHI < 5 was achieved in 43.2% [100]. The treatment effect was stable for more than 2 years in the series of long-term data. In cohort studies and in a randomized study, no difference in effectiveness compared to ventilation therapy could be detected [101, 102]. A meta-analysis in 2016 showed an overall reduction in AHI by 47.8/h (±4.7) [103]. The preoperative ESS of 13.5 (±2.8) was reduced to 3.2 (±3.2) postoperatively (median minimum follow-up time of 6 months, range 2–6 months). In terms of adverse effects, transient paresthesia (altered perception of sensation) of the second and/or third trigeminal branch is often reported [99]. This can be found in 14% of patients at 12 months. More than 90% of patients report positive or neutral aesthetic effects [100].

Other surgical procedures.

Radiofrequency ablation (RFTA) and soft palate implants, which are minimally invasive surgeries, are better tolerated than resective procedures. However, both surgeries appear to be significantly inferior to the latter in terms of efficacy [104,105,106]. To date, RFTA has been shown to result in only 31%—and only short-term—reduction in sleepiness measured by the ESS [106]. For example, a placebo-controlled study showed RFTA as single procedure to have no effects on OSA [107]. With regard to RFTA tongue base reduction, robust evidence of an effect is lacking to date, apart from isolated studies [108]. Radiofrequency surgery of the soft palate, at least in OSA patients, appears to have no effect on AHI, daytime sleepiness, and subjective snoring [109]. With a high level of evidence, soft palate implants have been shown to reduce snoring and OSA with a low to moderate effectiveness. Furthermore, extrusion was reported in nearly 10% of patients [105, 110].

Multilevel surgery is currently widely promoted, but controlled studies are missing to demonstrate its success. All case reports of a series of patients show a relatively constant response rate of 50–70% (weighted mean 66.4%) [111]. Response rates do not differ when tongue suspension is used as the retrolingual procedure instead of the most often used procedure of genioglossus advancement with or without hyoid suspension [112]. Another study also showed no difference between lingual tonsillectomy (using CO₂ laser) and hyoid suspension in combination with radiofrequency therapy of the tongue base [113]. In a randomized trial by Babadamez et al. [114], the authors were unable to show a significant difference between three different techniques of tongue base resection combined with UPPP [114]. Also, the response rate may decline over a period of 5 years [115].

The low level of evidence for hyoid suspension, midline glossectomy (cryosurgery, laser, robot-assisted), and lingualplasty has been already reported in a previous version of this guideline. Nothing has changed during the newly investigated time period [85, 116]. There is insufficient evidence for the following treatment procedures: laser-assisted soft palate surgery, uvula capping, cautery-assisted palatal stiffening operation, injection snoreplasty, radiofrequency surgery of the tonsils, transpalatal advancement pharyngoplasty, and isolated genioglossal advancement.

An overview of studies on surgical treatment is presented in Table 6.

Table 6 Surgical treatment in patients with OSA

Recommendations.

Reviewed and retained:

• Surgeries to improve nasal breathing should be considered in cases of obstructed nasal breathing and consequent CPAP intolerance (evidence level 4, grade B recommendation).

• If anatomical findings are appropriate, i.e., small mandible and narrow craniofacial structure, maxillary and/or mandibular advancement (bimaxillary advancement) should be considered, particularly if other treatments (CPAP, MAD) are not possible or not sufficiently tolerated (evidence level 1b, grade A recommendation).

Modified:

• In the case of clinically suspected oropharyngeal obstruction due to tonsillar hyperplasia, tonsillectomy with UPPP should be considered, particularly if other treatments (CPAP, MAD) are not possible or not sufficiently tolerated (evidence level 1a, grade A recommendation).

• Procedures for neurostimulation of the hypoglossal nerve should be considered in patients with CPAP intolerance or ineffectiveness with an AHI 15–65/h and a BMI up to 35 kg/m² in the absence of anatomic abnormalities and moderate to severe OSA (evidence level 1b, grade B recommendation).

The recommendations on surgical treatment methods were accepted with strong consensus (recommendation on TE-UPPP with 11/12, and remaining recommendations with 13/13).

5 Sleep-related hypoventilation/sleep-related hypoxemia

The ICSD‑3 [124] differentiates between sleep-related hypoventilation and sleep-related hypoxemia. For sleep-related hypoventilation, a distinction is made between six entities, whereas no subdivision is proposed for hypoxemia (see Table 7). According to the ICSD‑3, sleep-related hypoxemia is present when either PSG or nocturnal pulse oximetry records oxygen saturation of ≤ 88% for ≥ 5 min in the absence of sleep-related hypoventilation.

Table 7 Sleep-related hypoventilation syndrome/hypoxemia syndrome (according to ICSD-3)

Sleep-related hypoxemia is usually caused by an internistic or neurological disease and cannot be explained by a sleep-related breathing disorder, which, however, may be concomitantly present.

Some patients with sleep-related hypoxemia also exhibit daytime hypoxemia. Due to the clinical significance obesity hypoventilation syndrome and sleep-related hypoventilation due to a medical disorder, this section deals exclusively with these entities.

5.1 Obesity hypoventilation syndrome (OHS)

The diagnostic criteria are as follows:

• a. Hypercapnia (daytime paCO₂ ≥ 45 mm Hg)

• b. BMI > 30 kg/m²

• c. Hypoventilation is not primarily defined by another disorder

Some definitions additionally stipulate the presence of an SRBD, most commonly OSA (90%).

The prevalence of OHS among patients with OSA ranges, depending on the study, from 4 to 50%; individuals with a BMI > 30 kg/m² can be expected to have OHS with a frequency of 10–50%.

5.1.1 Main findings

Since 90% of OHS patients have OSA, they often also complain of symptoms of OSA, such as nonrestorative sleep, daytime sleepiness, and impaired concentration. This can result in symptoms being attributed exclusively to OSA, and the diagnosis of OHS being overlooked. Compared to OSA or obese patients, OHS patients more frequently experience shortness of breath and are more likely to present with clinical symptoms of peripheral edema, pulmonary hypertension, and cor pulmonale. Compared to eucapnic patients with a BMI > 30 kg/m², hospitalization rates, morbidity, and mortality are higher in OHS patients. In addition to respiratory complications such as increased need for invasive ventilation in hospital, in particular cardiovascular sequelae, such as arterial hypertension, heart failure, cor pulmonale, and angina pectoris, contribute to the greater morbidity. As a result, quality of life in OHS patients is significantly limited.

5.1.2 Diagnosis

In the case of a BMI > 30 kg/m² and corresponding clinical suspicion, blood gas analysis is performed to detect daytime hypercapnia. However, hypoventilation already begins to manifest with nocturnal hypercapnia before the full picture is achieved, making nocturnal determination of pCO₂ (arterial, capillary, transcutaneous, end-tidal) necessary in the case of a BMI > 30 kg/m² [125,126,127]. PSG is required to confirm sleep-disordered breathing.

Recommendation.

Reviewed and retained:

• In the case of clinical suspicion or predisposing underlying disease, the diagnosis of sleep-related hypoventilation should be made by means of nocturnal arterial or capillary blood gas analysis or nocturnal transcutaneous or end-tidal CO₂ measurement. For the diagnosis of OHS, daytime arterial blood gas analysis is required. To diagnose sleep-related hypoxemia, nocturnal oximetry should be performed in conjunction with CO₂ measurement at night (evidence level 1b, grade A recommendation).

Modified:

• In patients with a BMI of > 30 kg/m² and symptoms of an SRBD, determination of venous bicarbonate while the patient is awake as well as arterial or capillary pCO₂ or transcutaneous/end-tidal CO₂ should be performed to exclude concomitant hypoventilation during sleep (evidence level 1b, grade A recommendation).

The recommendations on the diagnosis of OHS were adopted with strong consensus (13/13).

5.1.3 Treatment

With regard to instrument-based treatment, a differentiation should be made between OHS patients with relevant OSA and OHS patients without upper airway obstruction. In stable OHS patients with severe OSA, both CPAP therapy and noninvasive ventilation (NIV) proved to be equally effective treatments in two recent randomized studies [128, 129]. In the study by Masa et al., both instrument-based therapies were superior to conservative therapy alone in terms of symptom improvement and PSG parameters [129]. Also, when considering long-term effects, CPAP and NIV were equally effective in OHS patients with severe OSA [130]. If a relevant rise in pCO₂ is seen during treatment initiation under CPAP titration, or a relevant persistent drop in oxygen saturation occurs (SaO₂ < 80% for ≥ 10 min), switching to NIV is required [131,132,133]; this is also the treatment of choice in OHS without concomitant relevant OSA [132,133,134]. If CPAP therapy has been started, clinical follow-up after 3 months is also judicious here. If a clinical improvement and normocapnia are seen, CPAP therapy is continued; otherwise, a switch to NIV is considered [132, 133].

Oxygen can reduce respiratory drive and increases transcutaneous CO₂, at least in acute situations. Evidence for the chronic setting is not available [135]. In contrast, NIV improves respiratory response, blood gases, the micro- and macrostructure of sleep, quality of life, hemodynamic parameters, and survival in OHS patients. Both fixed-pressure support NIV and volume-targeted NIV have been shown to be effective. Comparative studies show differing results. In addition, the expiratory pressure can also be automatically adjusted [136].

Weight reduction should be considered an important measure to address cause in OHS, although ventilatory therapy should not be delayed. If conservative approaches to weight reduction fail, bariatric surgery represents a possible treatment alternative. This has been shown to reduce body weight and improve lung function and blood gases.

Recommendations.

Reviewed and retained:

• In the case of OHS, bariatric surgery should be considered once weight loss measures have been exhausted (evidence level 2c, grade B recommendation).

Modified:

• In patients with OHS and relevant OSA, CPAP therapy should be attempted (evidence level 1b, grade B recommendation).

New:

• In OHS patients with severe hypercapnia or without concomitant upper airway obstruction, NIV should primarily be initiated (evidence level 1b, grade B recommendation).

The recommendations on the treatment of OHS were adopted with strong consensus (9/9).

5.2 Sleep-related hypoventilation due to a medical disorder

Typical disorders in the following respective disease groups are given in italics in parentheses: obstructive airway diseases (e.g., COPD), restrictive lung disease (e.g., interstitial lung disease), obesity (OHS), and neuromuscular or chest wall diseases (kyphoscoliosis, post-tuberculosis [TB] syndrome, post-polio syndrome, muscular dystrophy) predispose affected individuals to the development of sleep-related hypoventilation.

5.2.1 Main findings

Patients’ symptoms are uncharacteristic and often obscured by those of the underlying disease. Since the impaired ventilation predominates, patients typically complain of dyspnea on exertion, decreased exercise capacity, often leg edema, and, as a result of hypercapnia, headache. Sleep maintenance disorders and waking with shortness of breath are the most common sleep-related symptoms. Daytime sleepiness can also be a prominent symptom. Systematic investigations on the leading and sleep-related symptoms are not available.

5.2.2 Onset, course, and complications

The underlying disease causes reduced capacity and/or increased load on the respiratory system, which can still be compensated for in the early stages of the disease. As the underlying disease progresses, hypoventilation and hypercapnic phases occur initially in REM sleep, leading to metabolic compensation in the form of bicarbonate retention; this subsequently also reduces the respiratory response to hypercapnia. In the further course, hypoventilation/hypocapnia in non-REM sleep develops, as does, ultimately, the full picture of awake hypercapnic respiratory failure.

5.2.3 Diagnosis

The diagnosis of manifest daytime alveolar hypoventilation is made, by definition, using arterial blood gas analysis. For the further diagnostic work-up while awake, lung function testing and measurement of respiratory muscle strength and load are helpful. ECG, laboratory work-up, and chest X‑ray, as well as echocardiography where necessary, are performed depending on medical history and clinical findings. Irrespective of the underlying disease, daytime hypercapnia is regularly preceded by hypoventilation in REM sleep, and later also in non-REM sleep [137,138,139], which may worsen the prognosis of these patients [140, 141]. Observational studies suggest that nocturnal hypercapnia is an indicator for disease severity and long-term prognosis [142].

Since the uncharacteristic symptoms of alveolar hypoventilation are often wrongly attributed exclusively to the underlying disease, there is a risk of overlooking the early phase of chronic ventilatory failure with its episodes of hypoventilation only at night, thereby delaying appropriate treatment. Therefore, if patients are at risk of developing secondary alveolar hypoventilation, measurement of nocturnal breathing at regular intervals is indicated. The risk rises significantly from a vital capacity of < 50% predicted in restrictive disorders [139]. Pulse oximetry alone is not sufficient to detect sleep-related hypoventilation. Arterial paCO₂ measurements at night are not feasible. Therefore, transcutaneous or end-tidal pCO₂ measurement in combination with respiratory polygraphy is required to confirm sleep-related hypoventilation. Hypercapnia measured in transcutaneous capnometry directly demonstrates hypoventilation [143]. Continuous CO₂ measurement alone has the one disadvantage that it remains unclear whether the patient has reached REM sleep. Therefore, the procedure is good for the positive detection of hypoventilation, but not suited to its exclusion. Thus, in the case of nocturnal symptoms in the absence of confirmation of hypoventilation or long-term capnometry on respiratory polygraphy, PSG is indicated.

Recommendations.

Reviewed and retained:

• Being the most sensitive method to detect sleep-related hypercapnia, transcutaneous capnometry is recommended. This can be performed in conjunction with respiratory polygraphy or polysomnography (evidence level 2, grade C recommendation).

• With regard to the initiation of ventilation therapy in patients with neuromuscular or chest wall disorders and a vital capacity of < 50%, hypoventilation in sleep should be excluded (evidence level 1b, grade A recommendation).

• Polysomnography is the diagnostic standard for the exclusion and differential diagnosis of SRBDs in the context of sleep-related hypoventilation or hypoxemia (evidence level 2, statement).

The recommendations on the diagnosis of hypoventilation syndrome were adopted with strong consensus (13/13).

5.2.4 Treatment

In the case of chronic underlying diseases, their treatment alone is generally not sufficient to eliminate hypoventilation. From a treatment perspective, therefore, NIV during sleep via a mask is performed with the aim of increasing alveolar ventilation and preventing hyperventilation. The main criteria for initiating long-term NIV therapy in sleep-related hypoventilation due to a medical condition include symptoms and sequelae of ventilatory failure, such as dyspnea and edema, as well as reduced quality of life in the form of nonrestorative sleep as a result of sleep maintenance disorders or hypersomnia. Furthermore, the guideline of the German Society of Pneumology and Respiratory Medicine [132, 133] on noninvasive and invasive ventilation in chronic respiratory failure formulates the following criteria:

• In the case of sleep-related hypoventilation due to lower airway obstruction:

  • a diurnal paCO₂ ≥ 50 mm Hg or

  • a nocturnal paCO₂ ≥ 55 mm Hg or

  • an increase in nocturnal CO₂ ≥ 10 mm Hg, as measured transcutaneously, with a diurnal paCO₂ of between 45 and 50 mm Hg; in addition, long-term NIV should be initiated after an acute exacerbation requiring ventilation if daytime hypercapnia > 53 mm Hg persists 14 days after discontinuation of acute ventilation.

• In the case of sleep-related hypoventilation due to a chest wall disorder:

  • a diurnal paCO₂ ≥ 45 mm Hg or

  • a nocturnal paCO₂ ≥ 50 mm Hg or an increase in nocturnal CO₂ ≥ 10 mm Hg, as measured transcutaneously

• In the case of sleep-related hypoventilation due to a neuromuscular disease:

  • a diurnal paCO₂ ≥ 45 mm Hg or

  • a nocturnal paCO₂ ≥ 45 mm Hg or

  • a nocturnal CO₂ ≥ 50 mm Hg, as measured transcutaneously, or

  • a rapid decrease in forced vital capacity (FVC)

The aim of ventilation is normocapnia by eliminating hypoventilation under ventilation in sleep and reducing paCO₂ to daytime normocapnia. Ventilation can be initiated during the day and at night. Over the course of the initial set-up, the effectiveness of ventilation must be assessed by blood gas analysis both under spontaneous breathing and under ventilation, and supplemented with nocturnal measurements.

As a general rule, therapy is administered as NIV via a mask covering the nose or nose and mouth during the entire period of sleep. Since REM sleep represents a particularly critical phase, the effectiveness of ventilation in sleep should be documented by means of transcutaneous CO₂ measurement (tcpCO₂) plus respiratory polygraphy. In the case of uncertainty regarding nocturnal hypoventilation during REM sleep, PSG is indicated.

NIV can be administered as assisted, assist-control, or as purely controlled ventilation. There are no data on the superiority of one or other of these modes. Patients with neuromuscular and thoracic cage disorders often tolerate the controlled mode subjectively very well, while COPD patients generally prefer the assisted mode. Achieving an optimal, individually tailored set-up is crucial to good acceptance and treatment success. Only a handful of methodologically high-quality controlled studies are available on the effects of NIV (see Tables 8 and 9). In slowly progressive muscle diseases, kyphoscoliosis, and previous history of tuberculosis, NIV achieves a dramatic clinical improvement both in the acute and in the long-term setting, meaning that controlled trials in these conditions are now questionable from an ethical perspective. Whereas in the past these patients died from respiratory failure, life expectancy with NIV can be virtually normal [144,145,146,147,148,149,150]. As a result, quality of life under NIV is also massively improved, the hospitalization rate drops, and symptoms are reduced [124, 151]. In some cases, physiological parameters such as blood gas and lung function may even normalize [131, 141, 152]. The abovementioned diseases carry an unrestricted indication for treatment. For rapidly progressive neuromuscular diseases such as amyotrophic lateral sclerosis (ALS) and Duchenne muscular dystrophy, there is one controlled study and several case series showing a significant survival benefit with NIV [149, 151, 153, 154]; however, progression of the underlying disease limits the beneficial effects of NIV. The indication for ventilation therapy needs to be made on a case-by-case basis in these patients. As part of this, the ethical discussion on patient acceptance of any invasive ventilation that may be required needs to be had as early as possible. Diaphragm stimulation to treat ventilatory failure in ALS should not be performed, since a randomized study showed increased mortality with this procedure [155].

Table 8 Studies on NIV therapy in alveolar hypoventilation while awake or asleep in the setting of neuromuscular disorders (NMD); restrictive thoracic disorders and obesity hypoventilation syndrome (OHS)

Table 9 Studies on NIV therapy in alveolar hypoventilation while awake or asleep in the setting of COPD

COPD patients represent the largest group of patients that fulfill the indication criteria for NIV. Shorter-term controlled investigations demonstrated improved quality of life, a reduction in hospitalization rate, improved sleep quality, as well as improved physical training and blood gases [159, 163, 165,166,167,168]. Several controlled studies, albeit with significant shortcomings, found no reduction in mortality for the NIV-treated group of patients [160,161,162, 164]. A meta-analysis [158] demonstrated no differences in terms of BGA, pulmonary function, and quality of life for seven studies. An important point of criticism of these studies was the lack of substantial reduction in paCO₂ as a result of NIV therapy. In a randomized controlled study, in contrast, life expectancy in COPD patients significantly improved under NIV, while quality of life was reduced [169]. The German multicenter study in stable GOLD stage IV COPD patients with daytime hypercapnia confirmed these improved mortality figures as a result of NIV [170]. In that particular study, ventilation therapy led to a significant reduction in daytime paCO₂. On the basis of this evidence, an attempt at NIV therapy should be undertaken in COPD if the abovementioned indication criteria are met. Treatment effect and compliance should be assessed at around 3 months, and a decision should be made regarding treatment continuation.

Following acute COPD exacerbation requiring ventilation, continuous NIV should be initiated if hypercapnia (paCO₂ > 53 mm Hg) persists following discontinuation of acute ventilation [157, 158].

For further details on the treatment of chronic respiratory failure, the reader is referred to the German S2 guideline of the German Society for Pneumology and Respiratory Medicine “Treating chronic respiratory failure with invasive and non-invasive ventilation” [132, 133].

Recommendations.

Reviewed and retained:

• If nocturnal hypoventilation persists under CPAP, noninvasive fixed pressure support ventilation (without or with target volume) should be initiated (evidence level 1b, grade B recommendation).

• Oxygen therapy alone is not recommended in OHS (evidence level 2c, grade A recommendation).

Modified:

• Initiation of NIV is recommended in symptomatic patients with lower airway obstruction, neuromuscular disorders, or chest wall disorders with hypercapnia while awake (paCO₂ ≥ 50 mm Hg in diseases with lower airway obstruction or ≥ 45 mm Hg in neuromuscular or chest wall disorders) or asleep (paCO₂ ≥ 55 mm Hg in disorders with lower airway obstruction, ≥ 50 mm Hg in chest wall disorders, or ≥ 45 mm Hg in neuromuscular disorders, or a tcpCO₂ rise ≥ 10 mm Hg in lower airway obstruction and chest wall disorders, or tcpCO₂ ≥ 50 mm Hg in neuromuscular disease) (evidence level 1b, grade A recommendation).

The recommendations on the treatment of hypoventilation syndrome were adopted with strong consensus (9/9).

6 Appendix

6.1 Guideline report

6.1.1 Scope and purpose

The present update relates to the S3 guideline “Non-restorative sleep/sleep disorders” and specifically to the chapter “Sleep-related breathing disorders” therein. The above guideline on SRBDs was published in 2015 in the journal Somnologie [171].

Since the last publication of the guideline, a number of relevant findings relating to some aspects of this guideline have been published, thereby necessitating a reassessment of individual guideline statements and recommendations; the present update has been prepared for these chapters. All other contents, chapters, and recommendations in the guideline for which no update has been submitted remain valid in their current form (see also Chap. “Concept of the update”).

This guideline is aimed at medical and nonmedical professional groups (e.g., psychologists, natural scientists), nursing staff, self-help groups, and interested lay persons; it also serves as an information source for MTAs.

6.1.2 Composition of the guideline group and interest group involvement

Steering committee.

• Dr. med. Alfred Wiater, Cologne

• Prof. Dr. med. Boris A. Stuck, Marburg

Authors.

• Prof. Dr. med. Boris A. Stuck, Marburg

• Prof. Dr. med. Michael Arzt, Regensburg

• Prof. Dr. med. Ingo Fietze, Berlin

• Prof. Dr. med. Wolfgang Galetke, Hagen

• Dr. med. Holger Hein, Hamburg

• Simon D. Herkenrath, Solingen

• Priv. Doz. Dr. med. Benedikt Hofauer, Freiburg

• Prof. Dr. med. Joachim T. Maurer, Mannheim

• Prof. Dr. rer. physiol. Thomas Penzel, Berlin

• Prof. Dr. med. Maritta Orth, Mannheim

• Prof. Dr. med. Geert Mayer, Schwalmstadt-Treysa

• Prof. Dr. med. Winfried Randerath, Solingen

• Prof. Dr. med. J. Ulrich Sommer, Wuppertal

• Prof. Dr. med. Armin Steffen, Lübeck

• Prof. Dr. med. Clemens Heiser, Munich

Scientific medical societies.

To ensure the representativeness of the guideline group, the following societies, professional groups, and patient representatives were invited to collaborate:

• German Society for Sleep Research and Sleep Medicine (Deutsche Gesellschaft für Schlafforschung und Schlafmedizin), DGSM

• German College of General Practitioners and Family Physicians (Deutsche Gesellschaft für Allgemeinmedizin und Familienmedizin), DEGAM

• German Society of Internal Medicine (Deutsche Gesellschaft für Innere Medizin), DGIM

• German Association of General Practitioners (Deutscher Hausärzteverband)

• Deutsche Gesellschaft für Geriatrie e. V., DGG

• German Society of Otorhinolaryngology, Head and Neck Surgery (Deutsche Gesellschaft für Hals-Nasen-Ohren-Heilkunde, Kopf- und Hals-Chirurgie e. V.), DGHNO-KHC

• German Society for Cardiology—Heart and Circulatory Research (Deutsche Gesellschaft für Kardiologie, Herz- und Kreislaufforschung e. V.), DGK

• German Society for Pediatric and Adolescent Medicine (Deutsche Gesellschaft für Kinder- und Jugendmedizin e. V.), DGKJ

• German Society for Oral and Maxillofacial Surgery (Deutsche Gesellschaft für Mund-, Kiefer- und Gesichtschirurgie e. V.), DGMKG

• German Society for Neurology (Deutsche Gesellschaft für Neurologie e. V.), DGN

• German Respiratory Society (Deutsche Gesellschaft für Pneumologie und Beatmungsmedizin e. V.), DGP

• German Association for Psychiatry, Psychotherapy and Psychosomatics (Deutsche Gesellschaft für Psychiatrie und Psychotherapie, Psychosomatik und Nervenheilkunde e. V.), DGPPN

• German Psychological Society (Deutsche Gesellschaft für Psychologie), DGPS

• German Society of Dental Sleep Medicine (Deutsche Gesellschaft für Zahnärztliche Schlafmedizin), DGZS

• German Office of the German Orthodontic Society (Deutsche Geschäftsstelle der Deutschen Gesellschaft für Kieferorthopädie e. V.), DGKFO

• German Society for General and Visceral Surgery (Deutsche Gesellschaft für Allgemein- und Viszeralchirurgie e. V.), DGAV

• German Society of Anaesthesiology and Intensive Care Medicine (Deutsche Gesellschaft für Anästhesiologie und Intensivmedizin e. V.), DGAI

• Professional Association of German Neurologists (Berufsverband Deutscher Neurologen), BDN

• Federal Association of Physicians for Chest, Sleep, and Mechanical Ventilation Medicine (Bundesverband der Pneumologen, Schlaf- und Beatmungsmediziner), BdP

• German Association of Psychologists (Berufsverband Deutscher Psychologinnen und Psychologen e. V.), BDP

• German Association of Cardiologists in Private Practice (Bundesverband niedergelassener Kardiologen e. V.), BNK

• Professional Association of German Psychiatrists (Berufsverband Deutscher Psychiater), BVDP

• German Professional Association of Otolaryngologists (Deutscher Berufsverband der Hals-Nasen-Ohrenärzte e. V.), BVHNO

• German Sleep Society (Bundesverband Schlafapnoe und Schlafstörungen Deutschlands e. V.), BSD

• Bundesverband Gemeinnützige Selbsthilfe Schlafapnoe Deutschland e. V. (German association of not-for-profit sleep apnea self-help) GSD

• German Restless Legs Association (Deutsche Restless Legs Vereinigung e. V.), RLS e. V.

• German Narcolepsy Association (Deutsche Narkolepsie Gesellschaft e. V.), DNG

• Selbsthilfegruppe Schlafapnoe/chronische Schlafstörungen Solingen e. V. (Solingen self-help group for sleep apnea/chronic sleep disorders)

The following societies, professional groups, and patient representatives expressed their willingness to collaborate and named a representative for the nominal group process:

• German Society for Sleep Research and Sleep Medicine (Deutsche Gesellschaft für Schlafforschung und Schlafmedizin), DGSM (representatives: Prof. Stuck, Dr. Wiater)

• German College of General Practitioners and Family Physicians (Deutsche Gesellschaft für Allgemeinmedizin und Familienmedizin), DEGAM (representative: Frau Prof. Dr. Baum, proxy: Herr Dr. Stephan Hoffmann)

• German Association of General Practitioners (Deutscher Hausärzteverband; representative: Frau Prof. Dr. Baum, proxy: Herr Dr. Stephan Hoffmann)

• German Society of Otorhinolaryngology, Head and Neck Surgery (Deutsche Gesellschaft für Hals-Nasen-Ohren-Heilkunde, Kopf- und Hals-Chirurgie e. V.), DGHNO-KHC (representative: Prof. Dr. Clemens Heiser, proxy: Priv. Doz. Dr. Michael Herzog)

• German Society for Cardiology—Heart and Circulatory Research (Deutsche Gesellschaft für Kardiologie, Herz- und Kreislaufforschung e. V.), DGK (representative: Herr Prof. Dr. Arzt)

• German Society for Oral and Maxillofacial Surgery (Deutsche Gesellschaft für Mund-, Kiefer- und Gesichtschirurgie e. V.), DGMKG (representative: Herr Prof. Dr. Dr. Hans Pistner)

• German Society for Neurology (Deutsche Gesellschaft für Neurologie e. V.), DGN (representative: Prof. Dr. Geert Mayer)

• German Respiratory Society (Deutsche Gesellschaft für Pneumologie und Beatmungsmedizin e. V.), DGP (representative: Herr Dr. N. Büchner, proxy: H. Woehrle)

• German Association for Psychiatry, Psychotherapy and Psychosomatics (Deutsche Gesellschaft für Psychiatrie und Psychotherapie, Psychosomatik und Nervenheilkunde e. V.), DGPPN (representative: Dr. Lukas Frase)

• German Psychological Society (Deutsche Gesellschaft für Psychologie), DGPS (representative: Frau Prof. Dr. Angelika Schlarb)

• German Society of Dental Sleep Medicine (Deutsche Gesellschaft für Zahnärztliche Schlafmedizin), DGZS (representative: Herr Dr. Markus Heise)

• German Society for General and Visceral Surgery (Deutsche Gesellschaft für Allgemein- und Viszeralchirurgie e. V.), DGAV (representative: Herr Dr. Matthias Schlensak)

• German Society of Anaesthesiology and Intensive Care Medicine (Deutsche Gesellschaft für Anästhesiologie und Intensivmedizin e. V.), DGAI (representative: Herr Dr. Martin Rösslein)

• German Society of Internal Medicine (Deutsche Gesellschaft für Innere Medizin), DGIM (representative: Herr Prof. Dr. Hendrik Bonnemeier) Bundesverband der Pneumologen, Schlaf- und Beatmungsmediziner, BdP (representative: Herr Dr. Christian Franke)

• German Association of Psychologists (Berufsverband Deutscher Psychologinnen und Psychologen e. V.), BDP (representative: Frau Dr. Johanna Thünker, proxy: Herr Prof. Dr. Erich Kasten)

• German Professional Association of Otolaryngologists (Deutscher Berufsverband der Hals-Nasen-Ohrenärzte e. V.), BVHNO (representative: Dr. med. Andreas H. Schmidt)

• German Sleep Society (Bundesverband Schlafapnoe und Schlafstörungen Deutschlands e. V.), BSD (representative: Herr Werner Waldmann)

• Bundesverband Gemeinnützige Selbsthilfe Schlafapnoe Deutschland e. V. (German association of not-for-profit sleep apnea self-help) GSD (representative: Herr Ralf Kiehne)

• German Restless Legs Association (Deutsche Restless Legs Vereinigung e. V.), RLS e. V. (representative: Herr Dr. Joachim Paulus)

• German Narcolepsy Association (Deutsche Narkolepsie Gesellschaft e. V.), DNG (representative: Herr Tobias Schmid)

• Selbsthilfegruppe Schlafapnoe/chronische Schlafstörungen Solingen e. V. (Solingen self-help group for sleep apnea/chronic sleep disorders; representative: Herr Hartmut Rentmeister)

The following societies, professional groups, and patient representatives expressed their willingness to collaborate, but failed to appoint a representative for the nominal group process despite multiple requests to do so:

• Deutsche Gesellschaft für Geriatrie e. V., DGG

• Professional Association of German Neurologists (Berufsverband Deutscher Neurologen), BDN

• German Association of Cardiologists in Private Practice (Bundesverband niedergelassener Kardiologen e. V.), BNK

• Professional Association of German Psychiatrists (Berufsverband Deutscher Psychiater), BVDP

The German Society for Pediatric and Adolescent Medicine (Deutsche Gesellschaft für Kinder- und Jugendmedizin e. V.), DGKJ, did not appoint a representative after being informed, upon inquiry, that the guideline refers to adults.

The German Office of the German Orthodontic Society (Deutsche Gesellschaft für Kieferorthopädie e. V.), DGKFO, (representative: Herr Prof. Bert Braumann) canceled its participation in the consensus conference after reviewing the handouts.

6.1.3 Methodological procedure

The update of the guideline was guided by the Association of the Scientific Medical Societies in Germany (Arbeitsgemeinschaft der Wissenschaftlichen Medizinischen Fachgesellschaften e. V., AWMF) throughout the entire development process.

In a first step, the chapters that required updating were identified. Chapters requiring revision were defined as those in relation to which new scientific insights, among other reasons, could potentially result in a change to the recommendations contained therein. Accordingly, no update was undertaken when no new scientific insights were available or these did not conflict with the current recommendations. To this end, the guideline coordinator (B. A. Stuck) inserted all guideline recommendations, with references to the corresponding chapters, into a table and made these available to the members of the steering committee of the previous guideline, as well as to the authors responsible for the update, if available. These included the following persons: J. T. Maurer (Mannheim), I. Fietze (Berlin), G. Mayer (Schalmstadt-Treysa), W. Galetke (Hagen), W. Randerath (Solingen), B. A. Stuck (Marburg), J. U. Sommer (Wuppertal), T. Penzel (Berlin), B. Hofauer (Freiburg), C. Heiser (Munich), and A. Steffen (Lübeck). The abovementioned individuals assessed each chapter of the previous version of the guideline to determine whether an update was needed, as described above. A chapter was selected when more than 25% of the abovementioned persons deemed one or more of the recommendations in the chapter to require updating.

Each selected chapter was then assigned to a group of authors, who revised the chapter on the basis of the present text after selecting and evaluating the literature (see below). Already at this stage, attention was paid to achieving author groups with an interdisciplinary composition.

Independently of this, a text passage was identified that had led to misinterpretation in the clinical implementation of the previous guideline and had often been the subject of discussion, despite the fact that the passage was not a recommendation. The relevant passage has been modified in this update for greater precision.

Voting in the consensus conferences was carried out using the nominal group technique and was moderated by Priv. Doz. Dr. H. Sitter, Marburg. In advance of the conferences, all those involved in the consensus-development process received an invitation specifying objectives and the conference venue. Priv. Doz. Dr. H. Sitter’s function ensured independent moderation. Handouts included the manuscript of the necessary updates as well as the recommendations. During the consensus conferences, the statements/recommendations were presented, and positions on these were recorded using a round-robin procedure and summarized by the moderator. A pre-vote and discussion on the individual comments took place, and a ranking list was created. It was ensured that the individual points raised could be discussed. A final vote on each recommendation or all alternatives was held.

The nominal group process (consensus conference) took place in Frankfurt on 23.01.2020. The following mandate holders or specialist medical societies participated:

Priv. Doz. Helmut Sitter (AWMF), Prof. Boris A. Stuck (DGSM), Dr. Alfred Wiater (DGSM), Dr. Stephan Hoffmann (DEGAM/Deutscher Hausärzteverband), Prof. Dr. Clemens Heiser (DGHNO-KHC), Priv. Doz. Dr. Michael Herzog (DGHNO-KHC), Prof. Dr. Arzt (DGK), Prof. Dr. Dr. Hans Pistner (DGMKG), Dr. N. Büchner (DGP), Holger Woehrle (DGP), Dr. Lukas Frase (DGPPN), Prof. Dr. Angelika Schlarb (DGPS), Dr. Matthias Schlensak (DGAV), Dr. Martin Rösslein (DGAI), Prof. Dr. Hendrik Bonnemeier (DGIM), Dr. Christian Franke (BdP), Dr. Johanna Thünker (BDP), Dr. med. Andreas H. Schmidt (BVHNO), Ralf Kiehne (GSD), Dr. Joachim Paulus (RLS e. V.), Herr Hartmut Rentmeister (GSD-SH), and, as representatives of the author group, Prof. Dr. Wolfgang Galetke, Simon Herkenrath, Priv. Doz. Dr. Benedikt Hofauer, Prof. Dr. Joachim Maurer, Prof. Dr. Maritta Orth, Prof. Dr. Thomas Penzel, Prof. Dr. Winfried Randerath, Prof. Dr. Ulrich Sommer, and Prof. Dr. Ingo Fietze.

Following the review of the nominal group process, questions that had not been definitively dealt with were put to a final consensus in a nonanonymized Delphi process. The authors as well as participants in the nominal group process were invited to take part in the Delphi process. This process was opened on 21.02.2020 and closed on 15.03.2020. Responses were collected by B. A. Stuck and A. Wiater and put to a vote on 16.03.2020. All representatives appointed by the medical societies and organizations involved were invited to vote. The voting procedure was completed on 31.03.2020. The following mandate holders or specialist medical societies participated in the voting procedure: Prof. Stuck/Dr. Alfred Wiater (DGSM), Prof. Dr. Clemens Heiser/Priv. Doz. Dr. Michael Herzog (DGHNO-KHC), Prof. Dr. Arzt (DGK), Prof. Dr. Dr. Hans Pistner (DGMKG), Dr. Lukas Frase (DGPPN), Prof. Dr. Angelika Schlarb (DGPS), Dr. Christian Franke (BdP), Dr. Johanna Thünker (BDP), and Dr. med. Andreas H. Schmidt (BVHNO). Based on the results of the vote, appropriate changes were made to the text; one round was sufficient to reach a consensus.

Since this is an S3 guideline, the consensus process included the following elements:

Logical appraisal (clinical algorithm), formal consensus development, evidence-basing, and decision analysis. In the case of an S3 guideline, a solution is found using conditional logic (if... then... logic) in several steps based on a clearly defined question. Clinical studies and meta-analyses are included for evidence-basing. Using the graph algorithms, the procedure should be presented in a clear, simple, and concise manner.

6.1.4 Writing the guideline/consensus process

A preliminary draft of the update was written under the guidance of the person responsible for the guideline, Boris A. Stuck, and by the authors of the individual chapters.

6.1.5 Systematic literature search

The literature search was conducted for all studies published in the PubMed database from April 2014 onwards (the period of the literature search for the original guideline ended in 04/2014). The following were defined as inclusion criteria:

Publications in German or English, prospective or retrospective clinical trials, randomized controlled trials, controlled clinical trials, systematic reviews, meta-analyses, guidelines of the AWMF as well as of the European and North American medical societies (Practice Guidelines, Guidelines) in German or English. The following were defined as exclusion criteria: original articles published in a language other than English or German, experimental animal studies, letters to the editor, case reports, expert opinions, reviews that provided only a general overview of the topic rather than a systematic summary of the literature. The search was restricted to adults (19+).

To this end, the following search strategy was used: “sleep apnea” OR “snoring” with the following limits: publication date 01042014-30042019, humans, clinical trial, meta-analysis, practice guideline, guideline, randomized controlled trial, review, controlled clinical trial, guideline, English, German, adult: 19+years.

6.1.6 Appraisal

The literature was appraised by two independent reviewers according to the Oxford Centre for Evidence-based Medicine Levels of Evidence (2009). In addition to the selection of literature and classification of the level of evidence, a critical appraisal of the literature/studies was also carried out by the authors.

6.1.7 Clinical algorithms

In all, two clinical algorithms were updated (see Sect. 6.2 “Algorithms”):

• Diagnostic approach in obstructive sleep apnea

• Treatment approach in obstructive sleep apnea

6.1.8 External review and adoption

Following completion of the consensus process, DEGAM called for more extensive changes and additions to the content, which could not be made in the context of the current process. As a result, DEGAM stated that it could not endorse the guideline.

Once the consensus process had been completed, the DGAI concluded that anesthesia issues had largely been left unconsidered in the update and, for this reason, also did not approve the guideline.

The chairpersons of the remaining medical societies deliberated on and approved the guideline in June 2020.

6.1.9 Editorial independence

Financing of the guideline.

The guideline was developed and the consensus conferences held exclusively with funds from the German Society for Sleep Research and Sleep Medicine (DGSM).

Disclosing and managing conflicts of interest.

In order to avoid bias in the guideline process due to conflicts of interest, a number of preventive mechanisms were put in place: (1) systematic recording and appraisal of conflicts of interest of all persons involved in the guideline process, as well as resulting measures; (2) the interdisciplinary composition of the guideline group; and (3) the publication of a consultation version in order to allow those not directly involved in the guideline process the opportunity to comment.

All authors and participants in the nominal group process disclosed potential conflicts of interest using the form made available by the AWMF for this purpose. The two members of the steering committee reviewed participants’ conflicts of interest, and graded these as low, moderate, or high in line with AWMF recommendations. Equivocal cases were discussed within the whole guideline group. The guideline group reviewed the self-assessment of the conflicts of interests of the two steering committee representatives.

In the case of minor conflicts of interest, simple disclosure was sufficient; in the case of moderate conflicts of interest on a range of topics, the person in question was excluded from the respective voting process; and in the case of major conflicts of interest, the person in question was additionally excluded from the discussion process. Direct connections to industry or other interest groups associated with relevant direct or indirect financial contributions (e.g., to conduct clinical trials or cover travel expenses/speakers’ fees) relating to questions dealt with in the guideline were deemed to be moderate conflicts of interest.

The individual conflict of interest declarations are held by the guideline coordinator (Stuck), and a summary is available to the public on the homepage of the AWMF.

6.1.10 Dissemination and implementation

The guideline is disseminated via the specialist journal Somnologie and can be viewed online on the homepage of the AWMF (http://www.awmf.org).

6.1.11 Period of validity and update procedure

The updated chapters, or this update, remain valid for 3 years from the time of publication. A revision of the entire guideline on SRBD is scheduled for the end of that 3‑year period at the latest.

6.2 Algorithms

(Figs. 1 and 2)

Fig. 1

Diagnostic algorithm for suspected obstructive sleep apnea. AHI apnea–hypopnea index, OSA obstructive sleep apnea, CSA central sleep apnea, PSG polysomnography

Fig. 2

Treatment algorithm for obstructive sleep apnea. *In the case of CPAP intolerance and impaired nasal breathing, consider surgery to improve nasal breathing. **Particularly if alternative therapy (CPAP, MAD) is not possible or not sufficiently tolerated. AHI apnea–hypopnea index, OSA obstructive sleep apnea, CPAP continuous positive airway pressure, APAP automatic CPAP, TE-UPPP tonsillectomy and uvulopalatopharyngoplasty, BMI body mass index, MAD mandibular advancement device