Association between AI-driven conversational agents and physician-patient interaction quality during outpatient consultations: A propensity score matching study in China.

Sleep disturbances are increasingly being recognized as a problem in patients treated for head and neck cancers (HNC) , . This is particularly relevant in cancers involving upper-airway structures, including oropharyngeal cancer (OPC), where both the tumor itself and its treatment may affect airway anatomy and function. However, the existing literature remains scarce and fragmented, in part because studies often combine different tumor sites and treatment modalities under the broad HNC category , . This heterogeneity, spanning surgical, chemoradiation, and radiotherapy approaches, complicates the interpretation of sleep-related symptoms and limits more precise clinical characterization. In tumors affecting the oropharyngeal and related upper-airway regions, radiotherapy may further contribute to changes in airway patency and surrounding tissues, potentially increasing the risk of obstructive sleep apnea (OSA) and other sleep-related complaints . Furthermore, differences in anatomical involvement and treatment exposure, including the use of chemotherapy, may contribute to variability in the type and severity of sleep-related symptoms across HNC survivors , . Indeed, although radiotherapy is more directly implicated in structural upper-airway changes, chemotherapy may also influence sleep-related symptom burden through broader effects on respiratory function and muscle strength, and general treatment toxicity , . Accordingly, a more focused characterization of sleep symptoms in radiotherapy-treated HNC survivors, particularly those with tumors involving airway-relevant regions across clinically relevant disease- and treatment-related subgroups, may help clarify the nature and magnitude of this burden and better inform screening and survivorship care. Importantly, beyond prevalence estimates, quantifying how these symptoms compare to normative population values obtained from larger studies provides a clearer perspective on the excess burden attributable to cancer and its treatment . The objective of this study was to characterize sleep-related complaints, including snoring, OSA risk, excessive daytime sleepiness, insomnia symptoms, sleep quality, and fatigue, in a cohort of head and neck cancer survivors with tumors of airway relevance in a tertiary oncology center. Additionally, we explored whether these complaints differed according to clinically relevant disease- and treatment-related characteristics, particularly anatomical involvement and chemotherapy exposure. This work represents the first phase of a larger clinical research initiative aimed at improving the recognition and management of sleep disturbances in this population (). Recruitment was conducted through printed flyers and direct invitation. In total, 200 flyers were distributed across the stomatology, radio-oncology, and pulmonology departments at the Centre hospitalier de l'Université de Montréal (CHUM). The survey platform logged 91 entries, of which 15 were duplicates and 16 were incomplete (i.e., only partial questionnaires). For duplicate entries (i.e., multiple submissions from the same individual at different times), the first entry of each participant was retained. Moreover, two additional participants were excluded because: a) their radiotherapy treatment was on hold and had not yet resumed at the time of study participation; b) although affecting the maxilla, the cancer was an oligodendroglioma located in the brain (inferior parietal region). This left 58 unique participants with adequately completed questionnaires (≈30 min) for inclusion in the analyses. All participants had previously undergone curative-intent radiotherapy for HNC at CHUM. This cross-sectional study received institutional ethics approval (protocol #21–209), and participants provided online informed consent. Eligibility criteria included age ≥ 18 years, prior completion of curative-intent radiotherapy for HNC, and the ability to complete online questionnaires in French or English. Participants completed web-based questionnaires collecting sociodemographic and clinical information (e.g., age, sex, body mass index [BMI], education, employment, marital status, and income). Additionally, the Hospital Anxiety and Depression Scale (HADS, range 0–21) and the 12-item Short Form Health Survey (SF-12, norm-based score) were completed for quality of life . For sleep and fatigue variables, the STOP-Bang questionnaire (range 0–8) , Epworth Sleepiness Scale (ESS, range 0–24) , Insomnia Severity Index (ISI, range 0–28) , Pittsburgh Sleep Quality Index (PSQI, range 0–21) , and Chalder Fatigue Scale (CFS, range 0–33) , were used. Although snoring is included as one item within the STOP-Bang questionnaire, it was also described separately because STOP-Bang was used as a composite OSA-risk screening tool, whereas snoring was considered a symptom-level sleep complaint. This distinction was considered clinically relevant because STOP-Bang integrates symptom items with demographic and an