Detection of sleep apnea by case-finding and home monitoring with Somnolter^®: a pilot study

Sleep apnea is a pathological phenomenon which can result in clinical symptoms, e.g. daytime fatigue, excessive sleepiness, impaired concentration, depression. The combination of symptoms and proven breathing abnormalities during polysomnography (PSG) is historically framed in more or less distinguished syndromes. The obstructive sleep apnea syndrome (OSAS) is originally described by Gulleminault in 1976[1]. When not only apneas but also hypopneas are taken into account, the name changes into obstructive sleep apnea hypopnea syndrome (OSAHS). Also central sleep apnea-hypopnea syndrome (CHAHS), Cheyne-stokes breathing syndrome (CSBS) and sleep hypoventilation syndrome (SVHS) are recognized as pathological entities. The upper airway resistance syndrome (UARS) is a more controversial term that covers the part of the spectrum without oxygen desaturation.

Of these sleep related breathing disorders, OSAHS is the most common worldwide[2]. Despite the high prevalence, 70-80% of patients with OSAHS remains undiagnosed[3–5]. Numerous studies show prevalence rates of 4% in men and 2% in women[6, 7] while other large population studies reveal increased apnea-hypopnea indexes (AHI) in up to 26% of participants[6, 8–10]. These prevalence rates differ substantially because of inadequate syndrome definitions leading to difficulties to correctly compare results. Also in children OSAHS is common. A literature review of Marcus et al. demonstrates that up to 5.7% of the screened pediatric population has OSAHS[11]. Therefore, even children and adolescents with a high risk of having OSAHS, should be screened. Important studies concerning the prevalence of OSAHS are mentioned in Additional file1: Appendix 1.

In 1999 the American Academy of Sleep Medicine (AASM) Task Force developed recommendations for OSAHS syndrome definition and diagnostic measurement techniques[12]. These diagnostic criteria were established to minimalize the difficulties of comparing study results, e.g. prevalence rates. Diagnostic criteria are cited in Additional file1: Appendix 2. The purview and limitations of these criteria are further explained in Additional file1: Appendix 3.

It is crucial for health care professionals, especially when working in primary care, to gain the necessary knowledge about risk factors for OSAHS in order to adequately identify high risk patients. Both modifiable and non-modifiable risk factors associated with OSAHS are summarized in Additional file1: Appendix 4. From all risk factors, excess body weight seems to be one of the most important risk factors for OSAHS[5, 13, 14]. Therefore long-term weight reduction might be a helpful factor in the conservative treatment of obstructive sleep apnea.

OSAHS is also associated with a number of negative consequences on health. Studies show an increased risk of traffic accidents, arterial hypertension, stroke, depression, cardiovascular diseases and type 2 diabetes[15]. The associated morbidity and its strength of evidence is described in Additional file1: Appendix 5[16–18]. Some inconsistency about the association between OSAHS and coronary artery disease, arrhythmias, heart failure, type 2 diabetes and depression remains. On the other hand a remarkable independent association with arterial hypertension, stroke, cardiovascular mortality and death from all-causes is clearly proven. Some of the underlying pathophysiological mechanisms are described in Additional file1: Appendix 6.

The aim of this pilot study was to set up a case-finding strategy to identify high risk patients in whom it could be useful to screen for OSAHS as it is a highly prevalent disorder with possible dangerous sequelae and very effective treatment options (education and continuous positive airway pressure)[19]. In a selected number of high risk patients, a recently developed device, the Somnolter^®, was then used to investigate whether obstructive sleep apnea could be detected. The Somnolter^® is proven by Cheliout-Heraut et al. in 2011 to be a valuable alternative with a similar sensitivity and specificity as an in-hospital PSG, the current gold standard, for diagnosing sleep apnea syndrome[20, 21]. Both PSG and the Somnolter^® record the traditional parameters essential for the diagnosis of sleep apnea syndrome but only the Somnolter^® additionally records mandibular movements (Figure 1). These movements represent the respiratory effort during sleep and will reflect the number of apnea episodes which are very important for the diagnosis of sleep apnea syndrome. The mandibular movements also allow to assess the total sleeping time. Together with the other features of the device, the recording of mandibular movements make the Somnolter^® feasible to detect sleep apnea in a comfortable environment, such as the patients home.

A total of 365 questionnaires were filled out by patients consulting the co-operating GPs practices or by patients who heard from the study by one of the co-workers. The mean age of this population was 53.67 ± 14.57 years old. The sex ratio (male/female) was 6/10. The mean ESS score was 6.56 ± 4.56 on a total of 24 points. A hundred and sixty-eight patients out of 365 responded positively to only one of the screening questions, 65 patients responded positively to two screening questions and 61 answered positively to all three questions (Figure 3). Of 127 patients with ≥ 2 positive screening questions, 66 patients had an ESS < 10 and 61 patients had an ESS ≥ 10.

Between the 61 patients of the eligible population, 32 patients also met the other inclusion criteria (age ≥ 25 and BMI ≥ 25) for diagnostic evaluation with Somnolter^®. Seven of them were excluded based on the exclusion criteria mentioned earlier in this article. Nineteen subjects agreed to be tested by home-monitoring. Registration with Somnolter^® failed in one patient, resulting in a study population of 18 patients.

There appeared to be a heavy family burden in the study population. A family history of OSAHS, prior familial cardiovascular events and type 2 diabetes was reported by 6, 10 and 11 patients respectively.

Eight patients had an obstructive apnea index (OAI) ≤ 1 per hour, 6 had an OAI between 1 and 5 per hour, 1 person had an OAI between 5 and 10 per hour and the other 3 participants had a severe rise in OAI of 10.7, 35.3 and 46.5 respectively.

Other measured parameters were respiratory related arousal index (RERA I), respiratory disturbance index (RDI), respiratory arousal index (RAI) and the cumulative time below 90% oxygen saturation (SaO₂ CT 90). Though these parameters varied a lot between the different patients as shown in Table 2, some form of association could still be seen. In patients with a SaO₂ CT90 ≥ 1 hour (ranging from 1h29 minutes to 4h18 minutes), AHI, OAI, RDI and RAI increased equably with increasing SaO2 CT90 (Additional file1: Appendix 9). This trend was not seen for RERA I. When SaO₂ CT 90 was corrected for the total sleeping time, an even stronger trend could be seen (Additional file1: Appendix 10).

This pilot study revealed 11 new and 3 borderline cases of OSAHS on a total of 18 patients successfully tested with the Somnolter^®. This suggests that the used case-finding strategy and questionnaire could adequately detect high risk patients. The implementation of the proposed strategy in the GPs practice is easy and little time-consuming. The three screening questions about snoring, daytime somnolence and difficulties to concentrate, can easily be embedded in a general risk determination. In this study, only the patients with ≥ 2 positive answers on the 3 screening questions were further questioned using the ESS. Further research is needed to see if it is also useful for patients with only 1 positive question to fill out the ESS. To determine if this proposed case-finding strategy increases the pretest probability in a statistically significant way, case control studies will be needed. These results also support the statement mentioned before, namely that a high percentage (up to 70-80%) of patients with OSAHS is undiagnosed in the general population. Some explanations are suggested. At first, GPs rarely think of OSAHS as a possible explanation for persisting complaints of fatigue. If no somatic explanation is found at first sight, the complaints are often attributed to mental causes such as depression or chronic fatigue syndrome. For instance, patient 15 received antidepressants from his GP to treat his daytime somnolence and lack of energy. The Somnolter^® showed that this patient had severe OSAHS, which could explain his symptoms. On the other hand depression might also be a consequence of OSAHS[17], as stated before. Secondly, when not handed out directly and systematically to patients, fewer brochures were filled out. This might reflect some kind of threshold for the patients to consult their GP about snoring, excessive daytime sleepiness or difficulties to concentrate during the day. This study specifically asked for these symptoms through 3 screening questions. A final explanation for this high number of undetected OSAHS is the fact that in-hospital PSG appears to discourage patients. In the future the Somnolter^® might be a feasible alternative for the current PSG in the hospital, possibly resulting in a higher number of OSAHS diagnosis. When the diagnosis of OSAHS is confirmed, patients would only need one admission to test and regulate the pressure of the CPAP-device. By reducing the need for admission, waiting times for overnight PSG as well as health care costs can be reduced.

Besides the 14 new diagnoses of OSAHS, other interesting data regarding the sleep quality of patients were analyzed by the Somnolter^®. This allowed the authors to draw supplementary conclusions. First, for some patients the authors assigned another degree of OSAHS than would be given by the AASM. Patient 1 for example would normally be diagnosed as having mild OSAHS, but because of a very low OAI (0.4/h) and a SaO₂ CT90 of only 2 minutes, both related to supine position, she was finally diagnosed with positional OSAHS. Although an AHI of only 15.9/h, patient 3 was considered to have severe instead of moderate OSAHS and this because of long and deep desaturations (minimum SaO₂ of 65%) and a SaO₂ CT90 of more than 3 hours. These 2 cases indicate that besides AHI other parameters such as OAI, body position, SaO₂ CT90, maximum desaturation etc. should be considered when assigning the degree of OSAHS.

Secondly, the Somnolter^® allows to distinguish the more severe cases of OSAHS and to make a suggestion towards an appropriate treatment. In patient 15 the SaO₂ CT90 was no less than 4 hours and 18 minutes with a minimum SaO₂ of 67%. This patient for example has very severe OSAHS and should undoubtedly be referred for further treatment. It is remarkable that this patient, the most severe case of OSAHS in this study population, is a patient with Down syndrome. As mentioned before, Down syndrome significantly increases the risk of OSAHS[23]. On the other hand, in patient 1 the apneas were related to supine position. Taking measures to prevent this patient from sleeping in supine position might be sufficient to minimalize the apneas and deep desaturations and to decrease symptoms.

Finally, in patients with a SaO₂ CT90 ≥ 1 hour (ranging from 1h29 minutes to 4h18 minutes), AHI, OAI, RDI and RAI increased equably with increasing SaO₂ CT90. When SaO₂ CT 90 was corrected for the total sleeping, an even stronger trend could be seen. Further research is needed to establish a significant correlation.

The Somnolter^® is a new generation device since the Jawac sensors analyze jaw movements. The Jawac sensors are a powerful tool giving an image of a persons’ snoring pattern as well as gasping movements during sleep. Figure 1 shows gasping movements typically appearing after a period of apnea and desaturation, and followed by a slow increase of SaO₂. In the future, fewer sensors might be sufficient to adequately diagnose OSAHS. For instance the Jawac sensor to detect gasping combined with a pulse oxygen meter to detect desaturations. Devices measuring only these two parameters are already developed but further research is needed to confirm them as an adequate diagnostic tool.

It is important to stress that the questionnaire used in this study cannot replace a thorough history taking. For example, one of the patients was in a mourning phase which could cause her troubled sleep and explain her complaints of fatigue. This was only known after testing with the Somnolter^®. Another patient also reported, after testing, that her fatigue could possibly be explained by her recent motherhood. The used questionnaire did not pick up these causes of fatigue. Therefore, an extra question might need to be added to the questionnaire. For example: ‘Do you have any idea why you are extremely tired?’.

The authors are well aware of the limitations of this pilot study. In 3 of the 5 general practices, the brochure could not be distributed in a systematic way. As a consequence, in these general practices only prior interested individuals filled out the document, resulting in a possible selection bias. In two GPs practices, the brochures and questionnaires were handed out systematically by a secretary, with very high response rates. Because of this variable approach, the data cannot be considered to be of epidemiologic value. Another limitation of the brochures was an incomplete or incorrect completion by the patient. Because of the short time period during which this pilot study took place, in addition to having only one Somnolter^®-device available, a limited number of patients could be tested.

This pilot study set up a case-finding strategy to identify high risk patients in whom it could be useful to screen for OSAHS. Although the proposed strategy still needs to be optimized, the authors concluded that their strategy could definitely be helpful in selecting high risk patients. They detected OSAHS in 14 out of 18 patients tested with the Somnolter^®, suggesting a high percentage of undiagnosed OSAHS in the general population.

In the future the Somnolter^®, already proven to have a similar sensitivity and specificity as the gold standard in-hospital polysomnography, might be a feasible alternative to diagnose OSAHS. Patients experience the Somnolter^® as less discouraging than polysomnography since it is more comfortable and can be done in a home-setting.