Oral Cancer , Oral Squamous Cell Carcinoma, Oral Cavity Cancer

To evaluate the impact of AI-powered chatbot interactions versus traditional educational handouts on increasing participants' knowledge of oral cancer and its prevention

Delays between surgery and the initiation of post-operative radiotherapy (S-PORT) in head and neck cancer (HNC) patients are prevalent and associated with worse oncologic outcomes. This pilot study evaluates whether a proactive, automated care coordination system (PRONTO-HN) can improve adherence to recommended S-PORT intervals of ≤42 days at the Centre hospitalier de l'Université de Montréal (CHUM). The study is a single-center, quasi-experimental, interrupted time-series design with two phases. Group A (control) includes patients treated before the intervention (using a prospectively maintained database from August 2024 to August 2025). Group B (intervention) includes patients treated with the new coordination system from September 2025 to September 2026. The intervention includes automated alerts, multidisciplinary task coordination, and risk stratification based on a predictive model developed and published by our team. This model uses only pre-operative data to estimate the likelihood that a patient will require adjuvant therapy after surgery, stratifying patients into high- and low-risk categories. High-risk patients receive intensified coordination protocols, including early oncology and dental consultations and shorter target times for pathology results. Primary objective: Reduce the proportion of patients with S-PORT \> 42 days. Secondary/tertiary objectives: Reduce mean S-PORT time. Evaluate impacts on overall, locoregional, and disease-free survival in a 2- and 3-year follow up study. Patients are identified at the time the operating room request is submitted. Demographic, clinical, and oncologic data are collected and stored securely in REDCap. As the intervention is administrative in nature and does not modify patient care, consent is waived. Statistical analysis will evaluate the intervention's effect and identify predictors of delays. A sample of 38 patients per group provides adequate power to detect a drop in S-PORT \> 42 days from 80% to 50%.