Demographic and clinicopathologic characteristics

Demographic data were collected through a standardized self-report questionnaire, including age at the first diagnosis of CRC, gender, marital status, education level, geographic region, and occupation. Clinicopathological characteristics included the site of cancer occurrence (colon or rectum), pathological TNM stage at first diagnosis, metastatic status at the survey, the reason for the initial hospital visit, and the number of hospitals visited.

Awareness of CRC risk factors, screening, and treatment

Patients’ awareness regarding high-risk factors for colorectal cancer, CRC screening procedures, and treatment options before their diagnosis was gathered through a semi-structured questionnaire (SSQ). The SSQ was developed following the Chinese guidelines [12, 13]. It comprises three multiple-choice questions, and detailed information on the questions is presented in Table S2. Further details about the SSQ can be found in a previously published study [11].

Patients’ experiences with CRC screening, diagnosis, and treatment

Another SSQ was employed to gather information concerning CRC screening, diagnosis, and treatment. Patient screening history data were collected, including whether the patient had undergone screening, and information about barriers to not having a colonoscopy was collected based on patient self-reports. These barriers included lack of awareness, insufficient time for a colonoscopy, concerns about the discomfort associated with the procedure, cost-related challenges, waiting time for colonoscopy appointments, and issues with insurance coverage. With regards to CRC diagnosis and treatment, the following information was collected based on patients’ self-reports: (1) the utilization of gene testing, any barriers encountered, and the results of gene testing. (2) the adoption of currently available treatment modalities, such as targeted therapy, surgery, radiotherapy, chemotherapy, endoscopic treatment, and immunotherapy.

Medical expenditure

Medical expenses data was collected either from the medical records, or through patients’ self-reports. The gathered information will encompass patients’ out-of-pocket expenditures related to CRC diagnosis and treatment, reimbursement rates for all medical costs, annual household income, the perspective of patients on the cost of colorectal cancer treatment, and the type of health insurance.

Health-related quality of life

Health-related quality of life (HRQOL) was assessed based on two questionnaires: the Chinese Functional Assessment of Cancer Therapy-Colorectal (FACT-C) V.4 and the Chinese version of the European Organization for Research and Treatment of Cancer (EORTC) QLQ-C30 V.3. The FACT-C V.4 comprises 36 items distributed across five function subscales: physical, social/family, emotional, functional, and a colorectal cancer subscale [11, 14, 15]. Meanwhile, the traditional Chinese version of EORTC QLQ-C30 V.3 includes 30 items grouped into five function subscales (physical, role, emotional, cognitive, and social), nine symptom subscales (fatigue, nausea/vomiting, pain, dyspnea, insomnia, appetite loss, constipation, diarrhoea, and financial difficulties), and a global health/QOL subscale [16, 17]. In this study, a scale named FACT-C-plus-QLQ-C9 was created, consisting of 45 items selected based on expert opinions. This scale includes all FACT-C items along with nine items from QLQ-C30, as outlined in Table S3. The self-developed scale covers six functioning subscales (physical, social/family, emotional, functional, colorectal cancer subscale, and cognitive), two symptom subscales (fatigue and sleep disturbance), and one item related to financial difficulties. Higher scores on the functioning subscales and lower scores on the symptom subscales indicate a better quality of life. The Chinese versions of FACT-C and EORTC QLQ-C30 have been validated in prior studies [14,15,16,17].

Patients’ quality of life was assessed after CRC treatment. The summary score of HRQOL for each patient will be calculated across all items, including functioning scales and symptom scales (with inverted scores), resulting in a range from 0 to 180. A higher score indicates a better HRQOL. The Cronbach’s α coefficient of HRQOL in our questionnaire was 0.80.

Statistical analysis

Categorical data were described by frequencies and percentages, and continuous data using standard deviations (SD). The t-test, chi-square test, and Mann–Whitney U test were used to compare the characteristics of the EOCRC and LOCRC groups. Multivariate regressions were conducted to evaluate the associations between early diagnosis and HRQOL. The following variables were adjusted in multivariate regressions: HRQOL before treatment, sex, cancer location, education level, and TNM stage at initial diagnosis. Statistical significance was set at a P-value < 0.05. Data analysis was performed using R software (version 4.2.0, R Foundation for Statistical Computing).