研究目的

本研究旨在评估结直肠癌（colorectal cancer, CRC）患者在抗肿瘤治疗期间使用电子化患者报告结局工具（electronic patient-reported outcomes, ePRO）的依从性及其影响因素，基于实施科学理论和德尔菲专家咨询法，构建促进ePRO在CRC患者症状管理中应用的实施策略。

研究方法

本研究基于一项多中心临床研究，回顾性分析2023年1月至11月期间全国36家医院招募的326例CRC患者的ePRO使用数据，计算患者对ePRO的总体应答率（overall response rate, ORR）、平均应答率（average response rate, ARR）、单次应答率（single response rate, SRR）及患者水平应答率（patient response rate, PRR），并采用广义线性混合模型分析应答率的影响因素。

随机选取患者高依从性（CRC患者平均ORR高于60%）和患者低依从性（CRC患者平均ORR低于30%）的研究中心各1家，开展半结构化访谈，访谈对象包括17名CRC患者、10名CRC患者家属、17名肿瘤科医生和3名营养师。由两名研究人员基于实施性研究综合框架（consolidated framework for implementation research, CFIR），使用NVivo 12软件对访谈资料进行编码分析，识别ePRO在CRC患者症状管理中应用的障碍和促进因素，将“被提及频率>50%”的障碍因素界定为关键障碍。借助CFIR-ERIC实施策略匹配工具，初步匹配应对关键障碍的理论指导的实施策略，并通过小组讨论形成可操作化的实施策略。通过两轮德尔菲专家咨询评估实施策略的重要性和可行性，并对策略内容进行优化，引入动态可持续性框架（dynamic sustainability framework, DSF）对实施策略进行层级归类。

研究结果

CRC患者在ePRO各模块的ORR为31.98%-60.36%，ARR为31.49%-62.09%，不同治疗周期总数患者的PRR均值为27.33%-51.15%。影响因素分析结果显示，年龄<40岁、肿瘤分期为II/III期、居住于乡村地区及接受免疫治疗的患者，其ePRO应答率相对更高。

基于CFIR框架，从访谈内容中识别出32项促进因素与26项障碍因素，其中15项为关键障碍，涉及ePRO个性化不足、与临床流程不兼容及患者认知局限等方面。结合CFIR-ERIC实施策略匹配工具和小组讨论，初步形成9项可操作化的实施策略，并经两轮德尔菲专家咨询进行优化，共邀请18名来自临床流行病学、实施科学等领域专家参与，两轮问卷回收率为100%，专家权威系数分别为0.84和0.86，重要性协调系数W为0.12和0.32，可行性协调系数W为0.11和0.34，专家意见权威性和协调性良好。最终确定的9项实施策略包括：促进适应性、获取和利用患者及家属反馈、制定正式的实施蓝图、开展教育会议、制定激励和奖励机制、识别并培养支持者、进行教育性外展访问、建立和维持联盟、开发学术合作伙伴关系。基于DSF框架，该9项实施策略被归类至干预措施、实践环境、生态系统三个维度。

研究结论

CRC患者使用ePRO进行症状管理的依从性整体偏低，受多种人口学和疾病相关因素影响。本研究基于实施科学理论构建的9项实施策略，能够有针对性地应对当前面临的关键障碍问题，为提升CRC患者ePRO使用依从性和症状管理效率等提供理论依据与实施路径。

Objective 

This study aimed to evaluate the adherence to electronic patient-reported outcomes (ePRO) tools among patients with colorectal cancer (CRC) during antitumor treatment and to identify its influencing factors. Based on implementation science theories and the Delphi expert consultation method, we developed implementation strategies to promote the application of ePRO in symptom management for CRC patients.

Methods 

This study was based on a multicenter clinical study. A retrospective analysis was conducted on ePRO usage data from 326 CRC patients recruited between January and November 2023 across 36 hospitals in China. We calculated the overall response rate (ORR), average response rate (ARR), single response rate (SRR), and patient response rate (PRR), and used a generalized linear mixed model to analyze factors influencing adherence.
    Two study centers were randomly selected: one with high adherence (average ORR >60%) and one with low adherence (average ORR <30%). Semi-structured interviews were conducted with 17 CRC patients, 10 family caregivers, 17 oncologists, and 3 nutritionists. Two researchers coded and analyzed the interview data using NVivo 12 software based on the Consolidated Framework for Implementation Research (CFIR) to identify barriers and facilitators of ePRO use in symptom management. Barriers mentioned by more than 50% of participants were identified as key barriers. We then used the CFIR-ERIC Matching Tool to initially map theoretically guided implementation strategies addressing the key barriers. These strategies were refined through group discussions. A two-round Delphi expert consultation was conducted to evaluate the importance and feasibility of the strategies and further optimize their content. Finally, the Dynamic Sustainability Framework (DSF) was used to categorize the finalized strategies into implementation domains.

Results 

The ORR of ePRO modules ranged from 31.98% to 60.36%, while the ARR ranged from 31.49% to 62.09%. The mean PRR among patients with different total numbers of treatment cycles ranged from 27.33% to 51.15%. Influencing factor analysis showed that patients aged <40 years, those with stage II/III cancer, those living in rural areas, and those receiving immunotherapy exhibited relatively higher ePRO adherence.
    Based on the CFIR framework, 32 facilitators and 26 barriers were identified from the interviews, of which 15 were defined as key barriers. These included lack of personalization in ePRO tools, poor compatibility with clinical workflows, and limited patient awareness. Using the CFIR-ERIC Matching Tool and group discussions, 9 operational implementation strategies were developed. These were optimized through a two-round Delphi consultation involving 18 experts in clinical epidemiology and implementation science, with response rates of 100% for both rounds. The authority coefficients were 0.84 and 0.86, while the coordination coefficients (Kendall’s W) for importance were 0.12 and 0.32, and for feasibility were 0.11 and 0.34, indicating acceptable authority and consensus among experts. The final 9 strategies included: promoting adaptability, obtaining and using patient and caregiver feedback, developing a formal implementation blueprint, conducting educational meetings, establishing incentive and reward mechanisms, identifying and preparing champions, conducting educational outreach, building and maintaining coalitions, and developing academic partnerships. According to the DSF framework, these strategies were categorized into three dimensions: intervention, practice setting, and ecological system.

Conclusion 

Adherence to ePRO tools for symptom management among CRC patients is generally low and is influenced by various demographic and clinical factors. The nine implementation strategies developed based on implementation science theory effectively address key barriers and provide both a theoretical foundation and practical guidance for improving ePRO adherence and enhancing symptom management in CRC care.