A chatbot-powered G8 screening intervention to facilitate referrals to a comprehensive geriatric assessment among older adults with cancer

大多数癌症诊断和发病发生在 65 岁及以上的患者中。这个年龄段有 发病率增加 多种共病、老年综合征、工具依赖 日常生活活动和活动，以及反映衰老异质性并影响衰老的多重用药 老年人的治疗结果。临床指南建议常规使用综合治疗 针对老年癌症患者的老年评估 (CGA)，具有识别和优化非肿瘤学潜力的能力 问题并改变化疗决策并提高治疗耐受性。然而，CGA 是耗时的 由于缺乏足够的资源，不可能让每一位患有癌症的老年人都去看老年科医生。 老年病学家。 G8 是一份包含八项内容的问卷，已被推荐作为简短的筛选工具， 通常由具有良好敏感性的医疗保健提供者进行管理，以识别最有可能的患者 受益于 CGA。然而，G8 在常规肿瘤学临床实践中的实施情况参差不齐。 由于临床和系统障碍，结果吸收不理想。为了填补这一空白，我们建议使用患者- 中介实施方法，特别是采用聊天机器人技术来提供结构化集 G8 筛选问题可以模拟现实生活中对话所经历的内容，例如与 医疗保健提供者。我们提议的聊天机器人应用程序 JeffSeniorChat 将促进 G8 筛选 用于解释 G8 工具的目的和优点的支持性和教育性内容，每个问题都包含 帮助回忆的示例和探究，以及 G8 评估完成后的后续步骤以及如何 结果可能会影响他们的癌症治疗。以健康信念模型为指导，健康传播最佳 JeffSeniorChat 将通过实践和我们的初步形成工作进行迭代开发和完善 使用患者利益相关者的输入和通俗易懂的基于证据的内容进行快速原型设计过程 可用性测试以最终确定干预措施。我们将研究该方案的可行性和可接受性 JeffSeniorChat 针对 150 名老年人的 G8 自我评估完成情况和 CGA 出席情况进行干预 患有癌症。 G8 总分将由 JeffSeniorChat 后端算法自动计算，并 G8 评分 ≤ 14 的患者将被提示在杰斐逊高级成人肿瘤中心安排 CGA。 可行性将根据预先指定的入学率和 G8 完成率来确定。可接受性将是 通过参与者报告的干预可信度和满意度来衡量。 G8 评分 ≤ 14 的患者 提到 CGA，探索性结果将检查 CGA 出勤率、预测值 JeffSeniorChat 收集了 G8 分数，并由 CGA 确定了最终的衰弱指定，并认为 通过后续访谈了解 CGA 不依从患者的障碍。将 G8 分类流程简化为 CGA 通过支持聊天机器人的患者自我报告的评估和干预应该可以改善临床 效率并提高 CGA 就诊率，最终优化治疗选择和结果。

The majority of cancer diagnoses and morbidity occurs in patients age 65 and over. This age group has increased incidence of by multiple co-morbid conditions, geriatric syndromes, dependence in instrumental activities and activities of daily living, and polypharmacy reflecting the heterogeneity of aging and influence the treatment outcomes for older adults. Clinical guidelines recommend the routine use of a Comprehensive Geriatric Assessment (CGA) for older cancer patients with the potentials to identify and optimize nononcologic issues and alter chemotherapy decision and improve treatment tolerance. However, a CGA is time intensive and it is not feasible for every older adult with cancer to be seen by a geriatrician given the shortage of geriatricians. G8, an eight-item questionnaire, has been recommended as a brief screening tool which is usually administered by a health care provider with good sensitivity to identify patients who are most likely to be benefit from a CGA. However, the implementation of the G8 in routine oncology clinical practice has mixed results with sub optimal uptake due to clinical and system barriers. To fill the void, we propose using a patient- mediated implementation approach, specifically adopting the Chatbot technology to delivering a structured set of G8 screening questions that can simulate the content experienced by real-life conversation such as with a health care provider. Our proposed Chatbot application, JeffSeniorChat, will facilitate G8 screening with supportive and educational content to explain the purpose and benefit of the G8 tool, each question with examples and probes to help with recall, and next steps after the G8 assessment is completed and how the results might impact their cancer treatments. Guided by the Health Belief Model, health communication best practices and our preliminary formative work, JeffSeniorChat will be iteratively developed and refined through a rapid prototyping process using patient-stakeholder inputs and plain-language evidence-based content and usability testing to finalize the intervention. We will examine the feasibility and acceptability of the JeffSeniorChat intervention targeting G8 self-assessment completion and CGA attendance in 150 older adult with cancer. Overall G8 score will be automatically calculated by the JeffSeniorChat backend algorithm and patients with G8 scores ≤ 14 will be prompted to schedule a CGA at Jefferson’s Senior Adult Oncology Center. Feasibility will be determined by pre-specified enrollment and G8 completion rate. Acceptability will be measured by participant reported intervention credibility and satisfaction. Among patients with G8 scores ≤ 14 referred to a CGA, exploratory outcomes will examine CGA attendance compliance, predictive value of the JeffSeniorChat collected G8 score with the final frailty designation determined by the CGA, and perceived barriers among CGA non-adherent patients via follow-up interviews. Streamlining G8-triage process to a CGA through the Chatbot-enabled patient self-reported assessment and intervention should improve clinical efficiency and improve CGA attendance rate, ultimately optimizing treatment selection and outcomes.