Increased respiratory effort in both children and adults can be recognized with the help of simple esophagus pressure measurements [745]. If these measurements are combined with the registration of the oro-nasal air flow, it is then possible to differentiate between central, mixed, and obstructive respiratory events [752]. Together with the demonstration of frequent arousals in polysomnography, the esophageal pressure probe provides the essential tool in the verification of respiratory effort-related arousals (RERA) [710].

Initially, single flexible pressure sensors were pulled through the pharynx in order to identify a pharyngeal collapse site [304]. As this method turned out to be too time-consuming, catheters with up to six pressure sensors were developed. With these sensors, it is also possible to measure nasal as well as oral airflow and to identify several collapse sites during a whole night recording. The multichannel pressure measurements using catheters of a diameter of 2 mm are well tolerated [521], affecting neither sleep structure [684] nor breathing during sleep [751]. Moreover, it could be demonstrated that the placement of the oropharyngeal sensor at the free margin of the soft palate under visual control was sufficient to differentiate between “upper” and “lower” obstructions. The distribution of the collapse sites during sleep could be reproduced in 90% of the individuals during a second night if there were less than 40% or more than 60% palatal obstructions [621]. However, “lower” obstructions cannot be differentiated into tongue base and epiglottic obstructions. Furthermore, when comparing the levels of obstruction as defined during simultaneous drug-induced sleep endoscopy, there seem to be relevant differences [679].

Skatvedt recommends pressure probes in the selection of patients for laser-assisted uvulopalatopharyngoplasty (UPPP) even though there was only minimal impact on the reduction of upper hypopneas [683]. In a very small sample of 14 patients, Osnes found a significantly better outcome after UPPP in the subgroup of 11 patients with mainly “upper” obstructions [533]. Tvinnereim attests pressure recordings before Coblation assisted upper airway procedures a better outcome even though there is no control group [753]. Other study groups are more cautious in regard to the predictive value, because postoperative shifts of the collapse site into a different pharynx level, both toward cranial and caudal, have been observed [303, 461, 663, 686]. In summary, pressure recordings are reliable, but evidence concerning their impact on surgical outcome is limited.

Several reasons have been discussed for the conflicting results of topodiagnosis with pressure measurements in regard to the selection of patients for palatal surgery. On the one hand, pressure probes are incapable of recognizing segments that are already severely constricted but not yet totally collapsed [831]. Suratt et al. [731] observed a shift of the obstruction site during a single apnea phase toward caudal. Furthermore, other authors have registered such a high number of retrolingual obstructions that they in principle recommend the inclusion of the tongue base into the surgery concept along the lines of the so-called multilevel surgery [685, 830].

A fiberoptic endoscopy of the upper airway can be administered without difficulty on the awake patient in sitting and supine positions. But it must be said that the results do not correlate with the results gained during sleep [296]. Sleep videoendoscopy is a very sophisticated procedure and has to be restricted to specific indications [61, 567, 620]. Disadvantages of the method are, among other aspects, the reduction of the cross-section of the airway by the endoscope, arousal reactions due to the mechanic stimulus, visual obstruction by the phlegm, the simultaneous assessment of only one level of the airway, and the personnel-intense aspects of the procedure [296]. The Müller maneuver consists in the endoscopic observation of the upper airway during intensified inspiratory respiration with closed nose and closed mouth. While older studies considered the Müller maneuver as an identification method for the velar collapse type, and therefore recommended it as selection criterium for a successful UPPP [666], nowadays, this investigation is no longer regarded to have that value [60, 613, 825].

The flexible endoscopic assessment of the upper airway during pharmacologically induced sleep was first suggested for children [131], and later also for adults [130]. In addition to the disadvantages of an endoscopy during sleep already mentioned above, the employment of this procedure is further restricted by the fact that pharmacologically induced sleep cannot simply be equated with natural sleep [133, 435, 586]. Differences in collapse sites according to different sleep phases have already been mentioned. However, there is some evidence that at least breathing during pharmacologically induced sleep is not relevantly altered compared to natural sleep [636]. Furthermore, sleep endoscopy is able to convincingly display different patterns and sites of obstruction to surgeon and patient. Recent data suggest that the treatment plan may change in 40-75% of the cases if drug-induced sleep endoscopy is performed before treatment [280, 446]. In fact, it continues to be recommended for preoperative topodiagnosis [278]. Chisholm and Kotecha achieved a high success rate (90%) of LAUP eventually with tonsillectomy in 20 patients with moderate to severe OSA who were selected only if they demonstrated palatal or oropharyngeal obstruction during drug-induced sleep nasendoscopy [110]. Sleep endoscopy with gentle mandibular advancement was shown to be a good selection tool before mandibular advancement device therapy [332]. Yet we do not know of any study which was able to convincingly demonstrate that the surgical success can actually be improved by virtue of such a preoperative diagnosic procedure.

According to our own experience, videoendoscopy under sedation is helpful in assessing laryngeal obstructions. There is no other tool that can clearly show impaired breathing due to supraglottic or glottic collapse.

Another development is the digital analysis of fixated endoscopic images of the soft palate with the help of appropriate software. In an initial study, morphological differences were described in a pool of 121 primary snorers, 79 patients after LAUP, and 51 healthy control subjects [592]. It remains to be seen whether this concept will foster a viable method for the clinical routine.

In principle, a differentiation between the retropalatal and retrolingual collapse site can be established with the help of a recording of the respiratory sounds during sleep using Fast Fourrier Transfer (FFT) analyses [646]. It was possible to raise the on-target rate of the UPPP from 52.6% [665] to approximately 75% [646]. However, results regarding other interventions such as LAUP are not unanimous [68, 431]. In addition, snoring sound analysis could only identify pure tongue base snoring and not distinguish between different levels of obstruction as seen during drug-induced sleep endoscopy [643].

In the U.S. such a diagnostic option for the clinical routine exists in the form of the so-called SNAP procedure [790]. The investigator has the possibility to send in an audio cassette with snoring sounds for both an FFT analysis and an acoustic evaluation by an experienced listener. The analytic criteria of the system are not open and the surgeon has to blindly trust the company. In Europe the working group of Osman et al. developed the so-called Glan Clwyd Snoring Box, an instrument to differentiate between palatal from nonpalatal snoring [529, 532]. There are some other systems on the market now, but none of them has proven its superiority over simple clinical examination as far as surgical outcome is concerned. Nevertheless, continuous research is performed in order to clarify the role of sound analysis as a preoperative diagnostic procedure.

Many studies have attempted to establish the collapse site with the help of somnofluorscopy, radiocephalometry, computer tomography, and magnetic resonance imaging. Overall, the mentioned imaging procedures are only of limited use in predicting the surgical success of a UPPP [39]. Body mass and AHI continue to be decisive parameters.

Radiocephalometry is more successful in determining a retrolingual than a retropalatal collapse site [545, 601]. Accordingly, in the case of therapy failures after UPPP, it is possible to determine a more constricted retrolingual airway and a hyoid bone situated lower in relation to the mandible [602]. Both parameters also influence the success rate of nasal surgery in regard to the AHI in patients with mild OSA [659]. While radiocephalometry cannot generally be recommended as a routine form of diagnosis, it is certainly of use in patients with malocclusion or suspected retrolingual collapse site, and in patients who need to undergo surgery. An absolute indication is given before a planned maxillomandibular osteotomy [39, 284].

Fluoroscopy, rapid computer tomography, and functional magnetic resonance tomography have not been able to become part of the clinical routine, as they are too cost-intensive and cover a too short period of sleep.

Another instrument in this context is acoustic reflectometry. A probe generates a noise signal and measures the reflecting sound using a microphone [181]. As the probe is flexible, it already has proven feasibility in sleeping sleep apneics [183]. Unfortunately, the system is not commercially available any longer. These procedures remain reserved for specific lines of research [567].

Table 3.1 summarizes once more the advantages and disadvantages of the diagnostic procedures described above. Even though all these procedures have relevantly increased our understanding of and insight into the pathophysiology of OSA, their significance for daily practice is limited. We have essentially made the decision that in addition to the otorhinolaryngological evaluation, we only perform an endoscopy during wakefulness and sometimes a radiocephalogram. All of the other procedures outlined above are available in our center, but are in need of a specific indication. More research is strongly needed in order to specify their potential benefits.