Reliability and validity of the modified chinese version of new Cleveland clinic colorectal cancer quality of life questionnaire

Abstract

Purpose

The New Cleveland Clinic Colorectal Cancer Quality of Life Questionnaire (CCF-CaQL) was developed to assess postoperative quality of life (QoL) in patients with colorectal cancer (CRC). Since its introduction, it has been validated in several countries. This study aimed to evaluate the reliability and validity of the Modified Chinese version of the CCF-CaQL and its applicability in the Chinese population.

Methods

Data analyses were performed using R (version 4.3.3). Questionnaire reliability was assessed using Cronbach’s alpha and composite reliability, with additional reliability evaluated using Spearman’s correlation coefficient and the split-half method. Construct validity was examined using confirmatory factor analysis (CFA), with convergent validity assessed based on factor loadings and discriminant validity evaluated using the heterotrait–monotrait (HTMT) ratio.

Results

A total of 215 patients undergoing colorectal cancer surgery were included in this study. The revised questionnaire demonstrated high internal consistency, with a Cronbach’s alpha coefficient of 0.882 and composite reliability (CR) values exceeding 0.70 for all factors. Confirmatory factor analysis showed acceptable convergent validity, with all factors yielding average variance extracted (AVE) values greater than 0.40 and CR values ranging from 0.769 to 0.890. The HTMT coefficients indicated adequate discriminant validity, with most values below the recommended threshold of 0.85 and only one slightly above (0.887). All factors were significantly correlated with the total questionnaire score, with correlation coefficients ranging from 0.508 to 0.855. Furthermore, the questionnaire showed strong correlations with the Chinese versions of the EORTC QLQ-C30 and QLQ-CR29.

Conclusion

The modified Chinese version of the CCF-CaQL exhibited strong psychometric properties, confirming its validity and reliability as a disease-specific tool for evaluating quality of life in colorectal cancer patients.

What does this paper add to the literature

This study is the first to translate and validate the CCF-CaQL in Chinese, filling a gap in validated tools for assessing postoperative quality of life in patients with colorectal cancer. The availability of this instrument enables reliable clinical and research applications in Chinese-speaking populations.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1007/s00384-026-05215-6.

Introduction

Colorectal cancer (CRC) is among the most common malignant tumors in China, ranking second in incidence and fourth in mortality for all malignancies [1]. With the advancement of surgical techniques, the survival rate of patients with CRC has significantly improved, and increasing attention has been paid to patients’ postoperative quality of life (QoL), a predictor of survival [2]. To quantify postoperative QoL, several disease-specific instruments, such as the European Organisation for Research and Treatment of Cancer Colorectal Cancer 29-item Questionnaire (EORTC QLQ-CR29) and Functional Assessment of Cancer Therapy-Colorectal (FACT-C), have been developed for patients with CRC [3, 4]. These questionnaires have been widely used in clinical research to evaluate the impact of different surgical approaches on patient QoL [5]. However, existing instruments may not accurately or comprehensively capture the QoL of patients with CRC [6, 7]. To address this issue, Zutshi et al. [8]. developed the New Cleveland Clinic Colorectal Cancer Quality of Life Questionnaire (CCF-CaQL) in 2020, which incorporates previously overlooked domains such as neoadjuvant therapy and offers a novel, rapid, and comprehensive tool for assessing QoL in patients with CRC. To address the urgent need for reliable, language-specific QoL instruments for patients with CRC in Turkey, Ibrahim et al. [9]. Conducted the first validation of the CCF-CaQL in a non-English-speaking context, successfully confirming its reliability and validity in 2023.

Furthermore, currently available Chinese-language instruments for assessing CRC QoL, such as the QLQ-C30 and QLQ-CR29, although validated, contain too many items to be practical for routine clinical use. In contrast, the CCF-CaQL contains only 24 items, making it more concise and efficient, while maintaining strong disease specificity—offering a promising new assessment option for clinical practice. Therefore, this study aimed to translate the CCF-CaQL into Chinese, evaluate its reliability and validity in a Chinese population, and explore its applicability across different cultural contexts.

Materials and methods

Translational and content validity

Following the translation methodology recommended by the World Health Organization for cross-cultural QoL instruments, the Modified Chinese version of the CCF-CaQL was initially translated independently by two native Chinese-speaking professional translators proficient in English to ensure both conceptual and linguistic accuracy [10]. The translators reviewed and reconciled the two versions to resolve discrepancies, resulting in a second draft. Subsequently, a highly qualified English language expert with no prior knowledge of the original questionnaire back-translated the second draft into English. Before the translation, the questionnaire's purpose and structure were explained to the back-translator to minimize bias. The back-translated version was then compared with the original to identify inconsistencies, and the necessary modifications were made to finalize the Chinese version.

Five clinical experts specializing in colorectal surgery were invited to evaluate each item to assess the validity of the final version [11]. A 5-point Likert questionnaire was used to rate the clarity and relevance of each item, with response options of “strongly disagree,” “disagree,” “uncertain,” “agree,” and “strongly agree” scored as 1–5, respectively. The average score for each item ranged from 4 to 5, indicating good content validity. To ensure a robust psychometric evaluation of the instrument’s reliability and validity, the recommended sample size was 5–10 times the number of items in the questionnaire. Because the original CCF-CaQL contains 24 items, the minimum required sample size was 120.

Patients and data collection

The study participants were patients who underwent CRC surgery at our hospital between September 2023 and June 2024. Each patient completed the CCF-CaQL, EORTC QLQ-C30, and QLQ-CR29 questionnaires within 3 months postoperatively via outpatient visits or telephone follow-ups. Informed consent was obtained from all participants, and their demographic, surgical, and pathological data were collected.

A total of 245 eligible patients were enrolled (Fig. 1). Among them, 11 with recurrent or metastatic disease, eight with psychiatric disorders impairing their decision-making capacity or who were aged > 80 years, and four who refused to complete the questionnaires were excluded. Seven participants did not complete the questionnaire for various reasons. Trained physicians provided onsite guidance during outpatient visits to ensure that all items were completed. For the 17 patients who were unable to complete the questionnaire in person, telephone follow-ups were conducted. Ultimately, 215 patients completed all of the required questionnaires, and their data were collected and organized by the research team.

Fig. 1.

Flowchart of patient enrollment, inclusion, and exclusion

Ethics statement

The study protocol was approved by the Ethics Committee of Changhai Hospital (Approval No: CHEC2021-06). All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional research committee and with the 1964 Helsinki declaration and its later amendments. Written informed consent was obtained from all individual participants included in the study. Permission for the modification has been requested from Dr. Massarat Zutshi, the corresponding author of the original CCF-CaQL. Any necessary adjustments will be made upon receipt of their response.

Instruments

CCF-CaQL: The Modified Chinese version of the CCF-CaQL includes minor adjustments to the order of the items (Appendix 2). The CCF-CaQL [8] consists of two main domains: physical and psychological. Most items are scored on a 0–4 Likert scale, with 4 representing the highest score. One item is rated on a 0–10 scale, with 10 indicating the highest score. One item contains six checkboxes, and its score is calculated by subtracting the number of checked boxes from 6. Two items include a "Not Applicable" option; if selected, the maximum score of 4 is assigned to that item. The questionnaire is divided into four subscales: physical activity, physical health, psychological well-being, and social functioning. The total score is the sum of all item scores, with higher scores indicating better QoL.

EORTC QLQ-C30: The EORTC QLQ-C30 [12–14] is a widely used general cancer QoL questionnaire consisting of 30 items. It includes five functional domains (physical, role, emotional, cognitive, and social), nine symptom domains (fatigue, nausea/vomiting, pain, dyspnea, insomnia, appetite loss, constipation, diarrhea, and financial difficulties), and one global health status/QoL domain. Each item is rated on a questionnaire of 1–4, and the scores are linearly transformed to 0–100. Higher scores on functional domains indicate better functioning, whereas higher scores on symptom domains indicate greater symptom burden.

EORTC QLQ-CR29: The EORTC QLQ-CR29 [14, 15], a CRC-specific QoL module developed to supplement the QLQ-C30, includes five functional domains and 18 symptom domains. The module contains four multi-item scales (urinary frequency, blood and mucus in stool, stool frequency, and body image) and 19 single items covering urinary incontinence, dysuria, abdominal pain, buttock pain, bloating, dry mouth, hair loss, altered taste, anxiety, weight loss, flatulence, fecal incontinence, skin soreness, embarrassment, stoma-related problems, interest in sex (male), erectile dysfunction, interest in sex (female), and pain. All items are scored and linearly transformed to a 0–100 scale. Higher scores on the functional scales indicate better function, whereas higher scores on the symptom scales indicate more severe symptoms.

Cultural adaptation and item modification

During the cultural adaptation process, all items of the original questionnaire were carefully reviewed. Most items underwent only minor linguistic adjustments to align with Chinese expression patterns, with no substantial modifications to their content. However, two sexuality-related items (Q16 and Q17) exhibited response quality issues during the pilot testing. After consulting with colorectal surgery experts, we excluded these two items from the psychometric validation, but retained them as "optional/experimental supplemental items" in the Chinese version, with asterisks (*) marked in the questionnaire. To clearly indicate this adjustment, we refer to this version as the "modified Chinese version of the CCF-CaQL."

Statistical analysis

All statistical analyses were conducted in the R programming environment (R version 4.3.3). The main R packages used included psych (Revelle, 2025) for descriptive statistics, reliability analysis (including Cronbach’s alpha), and correlation analysis; lavaan (Yves Rosseel, 2012) for structural equation modelling (SEM) and confirmatory factor analysis (CFA); semTools (Jorgensen, 2025) for SEM-related auxiliary functions such as calculation of convergent validity (AVE) and discriminant validity (heterotrait–monotrait ratio [HTMT]); and ggplot2 (Wickham, 2016) for the visualization of correlation heatmaps. Statistical significance was set at p < 0.05.

Descriptive statistics are presented as mean ± standard deviation for continuous variables and as frequency and percentage for categorical variables. Internal consistency was assessed using Cronbach's alpha [16], and test–retest reliability was evaluated using Pearson's correlation coefficients. A Confirmatory Factor Analysis (CFA) was performed to confirm the questionnaire’s construct validity.

The suitability of the data for factor analysis was assessed using the Kaiser–Meyer–Olkin (KMO) measure and Bartlett’s test of sphericity. Standardized factor loadings were used to evaluate the contribution of each item to its latent construct; items with factor loadings ≥ 0.50 and p < 0.05 were considered to exhibit good convergent validity [17].

Discriminant validity was assessed based on the HTMT as recommended by Henseler et al. [18] with a cutoff value of 0.85 to determine the degree of differentiation between latent constructs. Additionally, the maximum shared variance (MSV) and average shared variance (ASV) were calculated as supplementary indicators of discriminant validity [17]. A latent variable was considered to have good discriminant validity if its AVE exceeded its MSV and ASV values.

The Chinese versions of the EORTC QLQ-C30 and QLQ-CR29 have demonstrated acceptable reliability and validity for evaluating postoperative QoL in patients with CRC [13, 14, 19–21]. Therefore, the criterion-related validity of the CCF-CaQL was assessed by analyzing the correlations between its domain scores and those of the EORTC QLQ-C30 and QLQ-CR29.

Results

Patient characteristics

This study included 215 patients in the final data analysis. All 132 patients with rectal cancer underwent Dixon surgery. Among the 83 patients with colon cancer, 26 underwent sigmoidectomy, 41 underwent right hemicolectomy, and 16 underwent left hemicolectomy. The baseline characteristics of the study participants are presented in Table 1. The mean age of the participants (143 men, 72 women) was 62 years (range, 32–78 years). Patients with rectal cancer accounted for 61.4%, those with colon cancer accounted for 38.6%, and 91 (42.3%) had stomas.

Table 1.

Patient characteristics

Characteristic (n = 215)
Gender
Male (%)	143 (66.5)
Female (%)	72 (33.5)
Age,median [Min,Max]	62.0 [32.0,78.0]
BMI,median [Min,Max]	23.4 [15.0,30.9]
Disease classification
Rectal cancer (%)	132 (61.4)
Colon cancer (%)	83 (38.6)
Ostomy
Yes (%)	91 (42.3)
No (%)	124 (57.7)
TNM Stage
Ⅰ	49 (22.8)
Ⅱ	72 (33.5)
Ⅲ	85 (39.5)
Ⅳ	9 (4.2)

Among the 215 participants, 198 completed the questionnaire in person, while 17 were interviewed via telephone. To assess potential bias related to the mode of administration, we compared baseline characteristics between the two groups. No significant differences were observed in age (t = 0.88, p = 0.382), BMI (t = 1.59, p = 0.114), total scale score (t = 0.38, p = 0.705), colorectal cancer type (χ² = 0.237, p = 0.627), or ostomy status (χ² = 0.127, p = 0.722), indicating that the two groups were comparable at baseline and that the mode of administration did not significantly bias the data.

Cultural adaptation outcomes

To accommodate cultural differences, the number and order of specific items in the Chinese version of the CCF-CaQL were modified. The original questionnaire consisted of 24 items. During clinical follow-up, two sexuality-related items—“How much have bowel problems affected your sex life?” and “How much has your illness or its treatment affected your sex life?”—demonstrated limited response variability. Statistical analysis revealed that the factor loadings for these two items (Q16 and Q17) were 0.198 and 0.152, both substantially below the recommended threshold of 0.30.

Further analysis suggested that, due to traditional Chinese cultural norms, patients tend to be relatively conservative when discussing topics related to sexual life and often provide socially desirable responses. This likely contributed to the poor data quality and low factor loadings observed for these items. After removing these two items, model fit improved considerably (CFI increased from 0.791 to 0.841; TLI from 0.777 to 0.829). The total Cronbach's α remained essentially stable (from 0.880 to 0.882), whereas the Physical Health subscale α increased from 0.820 to 0.859. In addition, the factor loadings of the remaining items were minimally affected by the deletion (see Table 9), suggesting that these two items were not essential to the overall factor structure.

Based on statistical evidence, cultural considerations, and consultation with clinical experts, we decided to exclude these two items from subsequent psychometric analyses of the scale. The impact of this modification on the overall questionnaire is presented in Table 2 and Table 9. Additionally, during the translation process, field experts recommended moving item 24 of the original questionnaire (“Compared to before your diagnosis or treatment, how much do you enjoy life?”) for the second item of the Chinese version to enhance questionnaire coherence. This suggestion was incorporated accordingly.

Table 2.

Confirmatory factor analysis fit indices

Fit Index	Value (Original model)	Value (Modified model)
x2/df	3.029	2.643
CFI	0.791	0.841
TLI	0.777	0.829
RMSEA	0.097	0.087
SRMR	0.074	0.066
Total CA	0.880	0.882
CA(Physical health measure)	0.820	0.859

CFI: Comparative fit index

TLI: Tucker-Lewis index

RMSEA: Root mean square error of approximation

SRMR: Standardized Root Mean Square Residual

Confirmatory factor analysis

As shown in Table 2, the four-factor model showed adequate fit on some indices (χ²/df, SRMR), while the global fit indices (CFI/TLI) fell below the ideal 0.90 threshold, suggesting that cultural differences may have altered the structural dynamics of the original framework. Notably, model fit improved after the removal of the two sexuality-related items.

Reliability

Reliability analysis was conducted (Table 3). The overall Cronbach's alpha coefficient for the Modified Chinese version of the CCF-CaQL was 0.882. The Cronbach's alpha and composite reliability (CR) values for all four dimensions exceeded 0.7, indicating good internal consistency reliability [16]. To assess the instrument's stability, test–retest reliability—reflecting the consistency of measurements under the same conditions across different time points—was evaluated. According to Nunnally and Bernstein, a coefficient above 0.70 is acceptable for test–retest reliability. For this purpose, 30 participants were randomly selected to complete the questionnaire via telephone 2 weeks after the initial assessment. The Pearson correlation coefficient between the two administrations was 0.904, demonstrating high consistency over time and confirming the questionnaire's excellent test–retest reliability.

Table 3.

Indicators for reliability and validity testing

	CA	CR	AVE	MSV	ASV
Physical health measure	0.859	0.863	0.440	0.418	0.296
Physical activity measure	0.885	0.889	0.453	0.203	0.142
Emotional measure	0.883	0.890	0.444	0.351	0.397
Social measure	0.773	0.769	0.412	0.249	0.406

CA: Cronbach's Alpha

CR: Composite Reliability

AVE: Average Variance Extracted

MSV: Maximum Shared Variance

ASV: Average Shared Variance

Validity

The questionnaire demonstrated a KMO value of 0.86, while Bartlett’s test of sphericity yielded p < 0.001, indicating that the sample data were suitable for factor analysis [22, 23]. A CFA was conducted to assess the convergent and discriminant validity of the questionnaire, followed by the construction of an SEM to evaluate the structural relationships among the latent variables. The factor loadings of most items exceeded 0.50 and were statistically significant (p < 0.05), with the exception of three items (Q5, Q20a, and Q20f) that fell below this threshold. Based on the expert panel's recommendations, these three items were retained for their content validity and clinical relevance. Although the AVE values for the four factors were below 0.5, they were all above 0.4. The CR values exceeded 0.6, indicating an acceptable level of convergent validity (Fornell & Larcker, 1981). Detailed results are summarized in Tables 3 and 4.

Table 4.

Factor loadings for convergent validity

Physical health measure	Std.all	P value	Std.Err
Q7d	0.774	0.000	0.028
Q8	0.663	0.000	0.032
Q9	0.508	0.000	0.026
Q15d	0.695	0.000	0.031
Q15e	0.730	0.000	0.030
Q20d	0.545	0.000	0.032
Q20e	0.643	0.000	0.031
Q21	0.703	0.000	0.039
Physical activity measure
Q1	0.695	0.000	0.047
Q2	0.798	0.000	0.033
Q3	0.729	0.000	0.030
Q4	0.725	0.000	0.033
Q5	0.493	0.000	0.040
Q6	0.648	0.000	0.035
Q7a	0.524	0.000	0.035
Q7b	0.701	0.000	0.034
Q7c	0.648	0.000	0.036
Q18	0.708	0.000	0.032
Emotional measure
Q15a	0.816	0.000	0.043
Q15b	0.797	0.000	0.040
Q15c	0.779	0.000	0.041
Q19	0.805	0.000	0.026
Q20a	0.326	0.000	0.047
Q20b	0.666	0.000	0.041
Q20c	0.556	0.000	0.040
Q20f	0.438	0.000	0.043
Q22	0.748	0.000	0.030
Q23	0.535	0.000	0.031
Q24	0.658	0.000	0.028
Social measure
Q10	0.635	0.000	0.043
Q11	0.608	0.000	0.044
Q12	0.606	0.000	0.042
Q13	0.612	0.000	0.045
Q14	0.739	0.000	0.032

The HTMT is considered more sensitive than the Fornell-Larcker criterion and cross-loading methods in detecting discriminant validity problems [18]. Tables 3, 4 and 5 present the discriminant validity of the questionnaire, showing that the majority of the HTMT coefficients were below 0.85 and that the MSV and ASV were lower than the AVE values. Only one HTMT coefficient was slightly above the 0.85 threshold (0.887), but still within an acceptable range. These findings indicate good discriminant validity among the factors. Table 6 shows the correlations between the scores of the four factors and the total score of the Modified Chinese version of the CCF-CaQL. All Pearson's correlation coefficients exceeded 0.5 and were statistically significant.

Table 5.

HTMT coefficients for discriminant validity among factors

Heterotrait-Monotrait Ratio of Correlations(HTMT)
	Physical health measure	Physical activity measure	Emotional measure	Social measure
Physical health measure	1	0.618	0.812	0.501
Physical activity measure	0.618	1	0.690	0.617
Emotional measure	0.812	0.690	1	0.887
Social measure	0.501	0.617	0.887	1

Table 6.

Internal correlation of the CCF-CaQL

Pearson correlation
	Physical health measure	Physical activity measure	Emotional measure	Social measure
Physical health measure	-	0.224*	0.700**	0.403**
Physical activity measure	0.224*	-	0.162*	0.243*
Emotional measure	0.700**	0.162*	-	0.732**
Social measure	0.403**	0.243*	0.732**	-
Total score	0.792**	0.508**	0.855**	0.697**

*: p < 0.05; **: p < 0.01

Table 7 evaluates the known-group validity of the CCF-CaQL by comparing scores across different patient groups based on tumor location and stoma status. Specifically, the total score differed significantly between patients with tumors in the colon versus rectum (125.6 ± 5.7 vs. 122.4 ± 7.4; p = 0.028), and between those with and without a stoma (119.7 ± 5.3 vs. 124.1 ± 7.5; p = 0.037). The physical domains showed significant differences between patients with versus without a stoma as well as between those with tumors located in the colon versus rectum. This finding aligns with the results of previous studies, [24] indicating that the CCF-CaQL effectively differentiates between various QoL groups among distinct patient populations.

Table 7.

Known groups validity of CCF-CaQL questionnaire

	Tumor Location			Stoma
	Colon (n = 83)	Rectum (n = 132)	P value	Yes (n = 91)	No (n = 124)	P value
Physical element	64.5 ± 2.2	62.1 ± 3.4	0.035*	60.3 ± 2.0	62.8 ± 3.9	0.012*
Mental element	61.1 ± 3.6	60.3 ± 4.9	0.087	59.4 ± 3.1	61.3 ± 4.0	0.055
Total score	125.6 ± 5.7	122.4 ± 7.4	0.028*	119.7 ± 5.3	124.1 ± 7.5	0.037*

In summary, the Modified Chinese version of the CCF-CaQL demonstrates strong psychometric properties, supporting its use as a valid and reliable tool for evaluating postoperative QoL in Chinese patients with CRC.

Criteria-related validity

Multiple studies have confirmed that the Chinese versions of the QLQ-C30 and QLQ-CR29 accurately reflect postoperative QoL of patients with CRC and have been widely applied in oncological clinical research [13, 14]. Therefore, we adopted the validated Chinese versions of the QLQ-C30 and QLQ-CR29 as the criteria for the QoL assessment. As shown in Table 8, the Modified Chinese version of the CCF-CaQL demonstrated strong correlations with the QLQ-C30 and QLQ-CR29, with total score correlations of 0.82 and 0.88, respectively (p < 0.05). Moreover, the correlations between conceptually similar domains were higher than those between dissimilar domains. For example, the correlation between the symptom domain and physical activity measures was 0.81, whereas the correlation with the emotional measure was only 0.35. This consistent and logical pattern supported the questionnaire's construct validity.

Table 8.

Criteria-related validity of the CCF-CaQL

	Symptom domain	Function domain	Total score (CR29)	Functional domains	Symptom domains	Global Health Status	Total score (C30)
Physical health measure	0.77**	0.73**	0.65**	0.76**	0.70**	0.82**	0.61**
Physical activity measure	0.81**	0.65**	0.72**	0.73**	0.71**	0.79**	0.67**
Emotional measure	0.35	0.78**	0.46*	0.34	0.38	0.75**	0.34*
Social measure	0.33	0.52*	0.35*	0.42*	0.21	0.56*	0.43*
Physical element	0.74**	0.66**	0.66*	0.70**	0.79**	0.81**	0.60*
Mental element	0.39	0.58*	0.41*	0.45*	0.55*	0.67*	0.52*
Total score (CCF-CaQL)	0.78**	0.67**	0.88**	0.71**	0.73**	0.89**	0.82**

*p < 0.05; **: p < 0.01

Significant statistical correlations were observed between the subscales and total scores of both instruments, confirming the validity of the CCF-CaQL and elucidating the relationships among the different QoL dimensions. Although the combined use of the QLQ-C30 and QLQ-CR29 provides a comprehensive assessment of QoL, practical challenges arose during administration. Patients exhibited lower compliance when required to complete both questionnaires consecutively, which may have compromised data quality. In contrast, the use of a single questionnaire is more convenient and may improve patient compliance.

The CCF-CaQL is specifically tailored to address the challenges faced by patients with CRC, including fecal incontinence and other bowel disorders, offering a unique and comprehensive perspective that allows for a more detailed evaluation of QoL related to bowel conditions.

Discussion

This study represents the first validation of the Chinese Revised version of the CCF-CaQL; to date, only the Turkish and Serbian version have been validated [9, 24]. Our results closely match the Turkish and Serbian validation study, indicating that the CCF-CaQL has good generalizability. In terms of reliability, the questionnaire demonstrated good internal consistency, with Cronbach’s alpha coefficients exceeding 0.8 for the total questionnaire and its four subquestionnaires. Before large-questionnaire testing, the content's validity was evaluated through expert consultations and a pilot study.

During the cultural adaptation process, most items were retained with only minor linguistic adjustments. However, two sexuality-related items (Q16 and Q17) exhibited poor response quality in the Chinese population, with factor loadings of 0.198 and 0.152, respectively—well below the recommended threshold of 0.30. Qualitative observations suggested that Chinese patients tend to be conservative when discussing sexual topics and often provide socially desirable responses, likely compromising data quality.

After consulting with clinical experts and considering both statistical and cultural factors, we decided to exclude these two items from the validated version of the scale. As shown in Table 2 and Appendix 1, their removal improved model fit (χ²/df: 3.029 → 2.643; CFI: 0.791 → 0.841; TLI: 0.777 → 0.829; RMSEA: 0.097 → 0.087; SRMR: 0.074 → 0.066), while Cronbach's α remained essentially stable (0.880 → 0.882). The factor loadings of the remaining items were minimally affected, indicating that the overall structure remained largely intact. However, the instrument lacks assessment of sexual life, which may reduce direct comparability with international cohorts. The translated items have been retained in Appendix 2 as optional supplements, marked with asterisks (*) to indicate that they are not scored and do not contribute to the total scale score, for potential future evaluation.

CFA was conducted to evaluate the fit of the original four-factor structure in the Chinese sample. As shown in Table 2, the revised 22-item four-factor model demonstrated an acceptable fit to the data: χ²/df = 2.643 (below the recommended threshold of 3), SRMR = 0.066 (below 0.08), and RMSEA = 0.087, which falls within the acceptable range of 0.05–0.10. The CFI (0.841) and TLI (0.829) were slightly below the conventional cutoff of 0.90.

We carefully re-examined the CFA model specifications and confirmed that the domain structure strictly follows the original CCF-CaQL framework, with no post-hoc modifications or correlated error terms applied. We acknowledge that the CFI and TLI values are slightly below the conventional cutoff of 0.9 [25]. Regarding RMSEA, different cutoff values have been proposed in the literature: values below 0.05 [26] or 0.06 [27] are typically interpreted as indicating "close fit"; values between 0.05 and 0.10 are considered to reflect "acceptable fit" [26, 28]; and values above 0.10 suggest "poor fit". In our model, the RMSEA value of 0.087 falls within the "acceptable fit" range.

As discussed by Lai and Green [29] it is not uncommon for fit indices to yield seemingly inconsistent conclusions, particularly in complex measurement models. Given the complexity of the measurement model in the context of cross-cultural adaptation—including the challenges of translating sensitive topics—some degree of inconsistency among fit indices is not unexpected. Importantly, all other psychometric properties of the revised scale—including internal consistency (Cronbach's α = 0.882), composite reliability (all factors > 0.70), and factor loadings of retained items (all exceeding 0.40)—met or exceeded recommended standards. These findings collectively support the overall adequacy and construct validity of the Modified Chinese version of the CCF-CaQL.

Regarding convergent validity, although the AVE values for all factors fell below the conventional threshold of 0.50, the CR for each factor exceeded 0.70 (ranging from 0.769 to 0.890). As noted by Fornell and Larcker, convergent validity may still be considered adequate when CR is high, even if AVE is slightly below 0.50, as CR is less affected by random measurement error. Therefore, despite the modest AVE values, the satisfactory CR indicates that the items reliably capture the underlying constructs.

Regarding discriminant validity, the HTMT ratios for most factor pairs were below the recommended cutoff of 0.85, with only one value marginally exceeding this threshold (0.887). Given that the deviation from the threshold was minimal, discriminant validity is considered generally acceptable. The observed overlap between the Social and Emotional domains may reflect the high integration of emotional and social well-being within the Chinese cultural context, where these dimensions are often intertwined.

Limitations

First, this was a single-center, cross-sectional study at a tertiary hospital in China, which may limit generalizability. Additionally, the lack of longitudinal data precludes assessment of the instrument's responsiveness to clinically meaningful changes. Future multicenter longitudinal studies are needed.

Second, QoL was assessed at only one postoperative time point, preventing evaluation of the instrument's sensitivity to changes over time or treatment-related improvements. Multiple assessment points are required to capture the trajectory of QoL in CRC patients.

Third, although the model fit was acceptable on some indices (RMSEA = 0.087, SRMR = 0.066), the CFI (0.841) and TLI (0.829) fell slightly below 0.90. Additionally, three items had factor loadings below 0.50, all four AVE values were below 0.50, and the HTMT ratio between the Social and Emotional domains (0.887) slightly exceeded the 0.85 threshold, with a narrow AVE-ASV margin (0.412 vs. 0.406) for the Social subscale. These issues may be attributable to the complexities of cross-cultural adaptation and the conceptual overlap between emotional and social well-being in the Chinese population.

Fourth, the exclusion of two sexuality-related items (Q16 and Q17) due to cultural sensitivity may affect content validity and limit comparability with international cohorts.

Fifth, EFA was not performed prior to CFA, as our primary aim was to validate the pre-specified four-factor structure rather than to explore a new one. Future studies with larger and more diverse samples could conduct EFA to further examine the factor structure in other Chinese populations.

Despite these limitations, the modified Chinese version of the CCF-CaQL demonstrated satisfactory psychometric properties overall and shows promise as a reliable and valid instrument for assessing QoL in Chinese CRC patients.

Conclusion

The current study validated the reliability and validity of the Modified Chinese version of the CCF-CaQL, demonstrating its robustness and effectiveness in assessing the QoL of Chinese patients with CRC. Our findings indicate that this instrument can be confidently applied in the clinical research setting. This questionnaire further enhances the postoperative QoL assessment framework for patients with CRC. Its use may improve the monitoring of treatment outcomes and the evaluation of intervention strategies within the Chinese population of patients with CRC, thereby contributing to better patient care and prognosis.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Appendix 1

Table 9

Table 9.

Factor loadings of all items before and after removal of the two sexuality-related items

Physical health measure	Std.all (Original model)	Std.all (Modified model)
Q7d	0.772	0.774
Q8	0.661	0.663
Q9	0.503	0.508
Q15d	0.694	0.695
Q15e	0.733	0.730
Q20d	0.546	0.545
Q20e	0.644	0.643
Q21	0.703	0.703
Q16	0.198	NA
Q17	0.152	NA
Physical activity measure
Q1	0.695	0.695
Q2	0.798	0.798
Q3	0.729	0.729
Q4	0.725	0.725
Q5	0.493	0.493
Q6	0.648	0.648
Q7a	0.523	0.524
Q7b	0.701	0.701
Q7c	0.648	0.648
Q18	0.708	0.708
Emotional measure
Q15a	0.817	0.816
Q15b	0.795	0.797
Q15c	0.777	0.779
Q19	0.805	0.805
Q20a	0.326	0.326
Q20b	0.668	0.666
Q20c	0.557	0.556
Q20f	0.438	0.438
Q22	0.749	0.748
Q23	0.535	0.535
Q24	0.658	0.658
Social measure
Q10	0.635	0.635
Q11	0.608	0.608
Q12	0.606	0.606
Q13	0.612	0.612
Q14	0.739	0.739

Author contributions

All authors meet the criteria for authorship as defined by the International Committee of Medical Journal Editors (ICMJE). Specifically, each author made substantial contributions to: (a) the conception and design of the study, or the acquisition, analysis, or interpretation of data; (b) drafting the article or revising it critically for important intellectual content; and (c) final approval of the version to be published.

Funding

This study was financially supported by grants from the National Natural Science Foundation of China (82272705, 82203137, 82503486); Shanghai "Rising Stars of Medical Talents" Youth Development Program of Shanghai Municipal Health Commission (2024–70), Naval Medical University "Deep Blue" Project Talent Program, the First Affiliated Hospital of Naval Medical University Special Project for Basic Medical Research (2023QD003, 2023PY04); the First Affiliated Hospital of Naval Medical University Special Project for Clinical Research (2024LYA02); Health Care Research Project 2024 (24BJZ10); Commission Health Industry Clinical Research Project (20224Y0348).

Data availability

The data supporting the findings of this study are available within the article and its supplementary materials.

Declarations

Competing of interests

We declare no competing interests.