Background The ever-increasing number of patients diagnosed with obstructive sleep apnoea (OSA) and treated by long-term continuous positive airway pressure (CPAP) overstretches conventional follow-up pathways. New approaches to the management of CPAP-treated patient follow-up are needed to strike a balance between remote monitoring through digital technologies and in-person patient-healthcare-professional contacts. Focusing on the reshaping of the management of care pathways of CPAP-treated patients, with a specific focus on telemonitoring platforms, we aimed to review the evidence on how digital medicine and artificial intelligence (AI) tools are facilitating patient phenotyping and triage, risk stratification and the allocation of resources between the various healthcare professionals for an optimal follow-up of CPAP-treated patients. Phenotyping OSA is a heterogeneous condition with diverse phenotypes differing in symptoms, comorbidities, demographics, lifestyle and socioeconomic context. Different phenotypes are associated with different CPAP adherence patterns and differing long-term prognosis. This diversity demands greater plurality in management pathways with different types and levels of support to ensure treatment adherence and risk reduction for patients while easing the burden on health services. In multidimensional phenotyping, we discuss alternatives to the apnoea hypopnoea index (AHI) as a measure of OSA severity. Then we consider risk stratification taking advantage of the wealth of CPAP monitoring data already available in databases that can now be exploited using AI and machine learning to direct (stratify) patients into appropriate follow-up management pathways. Integrated care clinics for high-risk patients We look at the role of integrated OSA care clinics particularly for the management of high-risk patients with low adherence and progression of comorbidities. Here, multidisciplinary teams might propose comorbidity management, and the use of connected wearable devices for long-term monitoring of physical activity, along with remote CPAP monitoring. Remote management pathways The pros and cons of remote management pathways to replace in-person follow-up visits are considered, including the need to re-evaluate CPAP-device reimbursement policies in some countries. While remote CPAP monitoring has become the cornerstone of follow-up providing information on adherence and efficacy, the processing of alerts needs to be improved, particularly regarding mask changes and early detection of CPAP failures. Challenges The implementation of CPAP monitoring alone, as well as its extension to multimodal monitoring, can present challenges that remain to be addressed (eg, access to digital care). The extent and components of remote follow-up must be adapted to each specific OSA phenotype. Finally, we give examples of certain patient phenotypes (eg, comorbid insomnia with OSA) with specific follow-up requirements, for which remote (even multimodal) monitoring alone has limitations and the intervention of both sleep specialists and/or their colleagues from other disciplines is needed. Conclusion Appropriately tailored combined digital and in-person CPAP follow-up pathways would present advantages both for patients with OSA and healthcare services.

中文翻译：

---

CPAP 治疗睡眠呼吸暂停患者随访的新管理途径，包括数字医学和多模式远程监测


背景 越来越多的患者被诊断患有阻塞性睡眠呼吸暂停 （OSA） 并接受长期持续气道正压通气 （CPAP） 治疗，使常规随访途径超负荷。需要新的方法来管理接受 CPAP 治疗的患者随访，以便在通过数字技术进行远程监测和面对面的患者-医疗保健-专业人员联系之间取得平衡。专注于重塑 CPAP 治疗患者的护理路径管理，特别关注远程监测平台，旨在回顾有关数字医学和人工智能 （AI） 工具如何促进患者表型和分诊、风险分层和各种医疗保健专业人员之间的资源分配的证据，以实现 CPAP 治疗患者的最佳随访。表型 OSA 是一种异质性疾病，具有不同的表型，在症状、合并症、人口统计学、生活方式和社会经济背景方面有所不同。不同的表型与不同的 CPAP 依从性模式和不同的长期预后相关。这种多样性要求具有不同类型和水平支持的管理途径具有更大的多元化，以确保患者的治疗依从性和降低风险，同时减轻卫生服务的负担。在多维表型分析中，我们讨论了呼吸暂停低通气指数 （AHI） 的替代方案，作为 OSA 严重程度的衡量标准。然后，我们考虑风险分层，利用数据库中已有的大量 CPAP 监测数据，现在可以使用 AI 和机器学习来利用这些数据来指导（分层）患者进入适当的随访管理途径。 高危患者的综合护理诊所 我们研究了 OSA 综合护理诊所的作用，特别是对于管理依从性和合并症进展程度低的高危患者。在这里，多学科团队可能会提出合并症管理，并使用互联可穿戴设备长期监测身体活动，以及远程 CPAP 监测。远程管理途径 考虑了替代面对面随访的远程管理途径的利弊，包括需要重新评估一些国家/地区的 CPAP 设备报销政策。虽然远程 CPAP 监测已成为随访的基石，提供有关依从性和有效性的信息，但警报的处理需要改进，特别是在面罩更换和 CPAP 故障的早期检测方面。挑战 单独实施 CPAP 监测及其扩展到多模式监测可能会带来仍有待解决的挑战（例如，获得数字护理）。远程随访的范围和组成部分必须适应每种特定的 OSA 表型。最后，我们给出了某些患者表型（例如，与 OSA 共病失眠）的示例，这些表型具有特定的随访要求，对于这些表型，单独的远程（甚至多模式）监测有局限性，需要睡眠专家和/或其他学科的同事的干预。结论 适当定制的数字和面对面 CPAP 联合随访路径将为 OSA 患者和医疗保健服务带来优势。