With an increasing prevalence of OSA and rising numbers of patients receiving treatment, the need for alternative therapies is also growing, with more patients terminating PAP therapy. Though highly efficacious and cost effective, PAP therapy as the first-line treatment is often not tolerated in chronic use, which leaves many patients with a need for alternatives to control their OSA [17, 18, 22]. This is at least partially accelerated by the lack of structured and adequately funded follow-up pathways for patients using PAP therapy, which contributes to the high rate of PAP termination. In addition, as reported recently by Woehrle et al., large numbers of patients diagnosed OSA do not receive PAP therapy [23]. Though the reasons were not identified in this study, it highlights the need for alternative treatments. Since PAP therapy would have been theoretically available for all patients, they potentially opted against it and could benefit from another form of treatment.

As experts in the field and important influencers of patient decision-making, sleep physicians play a relevant role in the allocation of alternative OSA treatments, and their knowledge and perceptions are crucial for patient access. To the best of our knowledge, this is the first study to evaluate knowledge of, perceived barriers to, and indication of alternative treatments for OSA from the perspective of sleep physicians in Germany. Our study shows a moderate average degree of self-reported knowledge among German sleep physicians from both the inpatient and the outpatient sector. Except for positional therapy, only a few physicians reported expert-level knowledge of alternative OSA treatments. In light of other findings from this study reported earlier, where a majority of participants reported good knowledge of recent clinical guidelines, there might be a different appraisal of high-level knowledge of guidelines and knowledge of individual treatment modalities.

While in general, knowledge of more common alternatives like MAD and positional therapy is higher, understanding of specialized surgical interventions such as HNS, MMA, and UA surgery is lower, which translates into lower self-reported indications for these treatments. On the other hand, participants reported relatively high numbers of indications for alternative OSA treatments, which can be interpreted as these being well accepted among physicians. This is supported by the fact that only a small minority of 1.1% reported not indicating any treatments beside PAP therapy.

Our study underscores one important feature of sleep medicine in Germany, which is the need for multidisciplinary provision of care. This relies on local or regional networks of different medical specialists that collaborate in the provision of services. As such, most alternative treatments are provided after referral to other physicians of a different discipline. The only exception is positional therapy, which is mostly offered within the own clinic or practice. The results presented here are supported by earlier reports from this project that show a high degree of cooperation in sleep medicine, especially among the two disciplines that dominate the provider landscape, namely otorhinolaryngology and respiratory medicine [21]. Since this multidisciplinary cooperation requires significant interaction and communication between providers, ongoing reimbursement challenges in sleep medical services—present in both the outpatient and the inpatient sector—could create disincentives for collaboration and threaten patient access. A potential solution to further increase collaboration and interdisciplinary exchange could be the implementation of local OSA boards, comparable to the “heart team” approach in cardiac care, in which members of the team discuss patient cases and optimal treatment allocation on a regular basis.

Although theoretically, all treatments except MAD, for which coverage was only introduced in 2021, are part of the benefit scheme of statutory insurance, reimbursement by third-party payors is perceived as the most relevant barrier to indication across all alternative treatments assessed. Interestingly, physicians from the inpatient sector perceive reimbursement limitations significantly more often for the treatment options UA surgery and HNS, which are mainly provided in hospitals. A common practice in Germany is ex-post denial of hospital claims by the statutory insurance’s medical service, which is reported to occur frequently and which could influence the perception of sleep physicians for these treatments [24]. This is especially important as HNS therapy was funded under special agreements and only introduced into the regular DRG catalogue in 2021.

Given the invasive nature of surgical treatments, it is not surprising that no demand from patients is reported as a barrier to these alternatives, since surgical treatment in general is often not preferred by patients in comparison to nonsurgical options [25, 26]. Although UA surgery, HNS, and MMA are recommended in OSA practice guidelines published by the German Sleep Society, which require a rigorous assessment of clinical evidence, and randomized clinical trials have been published for these interventions, a fairly large number of participants report insufficient evidence as a barrier to utilizing these treatments [20, 27]. Factors that lead to this perception could not be established from this study but should be a further part of future studies given the importance of this aspect in the provision of alternative treatments and the central role of the physician in the decision-making process.

With a large sample size representative of the German care structure in sleep medicine, we believe that this study provides strong evidence to support the findings presented. Nevertheless, the study is subject to some limitations, which should be considered when interpreting the results. Though we were able to recruit a large sample, the overall response rate was only 9%, which appears to be low in comparison to other studies [28]. This response rate is related to the fact that a large database was used to distribute the survey to ensure that all potential physicians were reached. Also, the questionnaire was quite comprehensive and required substantial time to complete, which could have deterred some physicians. However, the geographical distribution of participants and the proportion of participants with regards to the medical disciplines corresponded to a recent analysis of sleep medical care in Germany, as discussed in the initial analysis of the survey [21]. Furthermore, the study was based on a simple survey design that relied on self-reported answers, which creates several issues. Beside overconfidence in reporting knowledge, there is a risk of socially desirable answering behavior which might influence the responses. However, with a relatively large and balanced sample, we believe that the risk of bias is limited. It is also important to have in mind that the survey did not test the knowledge, which might have led to more precise estimation of knowledge. The time burden associated with a more sophisticated survey design was considered too high, and it was decided to use this approach to lower the risk of dropouts. Another limitation arises from the fact that knowledge and perceptions are fluid and subject to constant external influences. Ongoing continuous medical education, changing medical practice guidelines, and reimbursement decisions will impact upon how available treatments are perceived and indicated in routine care. Recently, MAD treatment was added to the benefit catalogue of the statutory health insurance, which will reduce reimbursement limitations present before this decision [29]. Finally, knowledge and perceived barriers are only parts of the actual decision-making process, and other factors, e.g., immediate availability in the local healthcare ecosystem, personal preferences, and administrative aspects, will influence adoption and utilization of treatments. To further assess the factors driving these circumstances, additional research is warranted.