Introduction
Colorectal cancer is a major global health issue, with over 1.9 million new cases in 2020, ranking third in men and second in women worldwide (Sung et al. Reference Sung, Ferlay and Siegel2021). The highest incidences were observed in Europe, Australia, New Zealand, and North America, with projections estimating 3.2 million new cases and 1.6 million deaths by 2040 (Morgan et al. Reference Morgan, Arnold and Gini2023). In 2021, Serbia reported 5,174 new cases, representing 12.4% of all cancers and making it the second most common cancer in the country. It affects more men than women, with age-adjusted rates of 44.3 and 24.7 per 100,000, respectively, and is most common in those over 75 years old (Cancer Registry Reference Cancer Registry2023). The incidence of colorectal cancer has been increasing over the last 30 years in Serbia and other Balkan countries (Todorovic et al. Reference Todorovic, Stamenkovic and Stevanovic2023).
Assessing quality of life (QoL) is essential in cancer management, offering insights into the patient’s experience of disease and treatment effects, covering physical, emotional, cognitive, and social aspects (Aaronson et al. Reference Aaronson, Ahmedzai and Bergman1993; Qedair et al. Reference Qedair, Al Qurashi and Alamoudi2022; Parsons et al. Reference Parsons, Kruijt and Fox2019). The EORTC QLQ-CR29, a module for colorectal cancer derived from the QLQ-CR38, is designed to measure health-related QoL, addressing specific symptoms, treatment side effects, body image, sexuality, and future perspectives (Ganesh et al. Reference Ganesh, Agarwal and Popovic2016; Gujral et al. Reference Gujral, Conroy and Fleissner2007; Sprangers et al. Reference Sprangers, te Velde and Aaronson1999; Whistance et al. Reference Whistance, Conroy and Chie2009). The validation of this instrument among different populations is crucial for its application, ensuring the relevance and sensitivity to the specific quality of life impacts of colorectal cancer and its treatment (Bachri et al. Reference Bachri, Essangri and El Bahaoui2023; Ihn et al. Reference Ihn, Lee and Son2015; Kishore et al. Reference Kishore, Jaswal and Kulkarni2021; Lin et al. Reference Lin, Zhang and Wu2017; Sanna et al. Reference Sanna, Bereza and Paradowska2017). The Serbian adaptation of the FACT-C questionnaire highlights the importance of cultural and linguistic adjustments for such tools (Ilić-Živojinović et al. Reference Ilić-Živojinović, Krdžić and Jovanović2022).
The aim of our study was to translate, culturally adapt, and psychometrically validate the EORTC QLQ-CR29 questionnaire for Serbian colorectal cancer patients.
Methods
The prospective cohort study was conducted at the Clinic for Digestive Surgery, University Clinical Center of Serbia. It is the largest university center in our country, covering around 1.6 million residents of the capital city of Belgrade, and part of Serbia. The study population consisted of patients admitted to the clinic due to colorectal carcinoma between May 2022 and February 2023.
The study inclusion criteria were: patients aged 18 and older who were diagnosed with colorectal adenocarcinoma and were candidates for colorectal surgery. All patients were Serbian native speakers. Patients with a stoma due to the presence of obstructive colorectal cancer were also included in the study.
The exclusion criteria for the study were: patients younger than 18 years, presence of inflammatory bowel disease, colorectal surgeries for diverticulitis and benign polyps, ischemic colitis, hereditary colorectal neoplasia, recurrent adenocarcinoma, patients with present metastases, patients who were unable or refused to give informed consent, as well as patients with whom it was not possible to communicate due to cognitive impairment.
Questionnaires
EORTC QLQ-C30 Version 3.0: A 30-item questionnaire developed to assess the quality of life of cancer patients. It includes 5 functional scales (physical, role, emotional, cognitive, and social functioning), 3 symptom scales (fatigue, nausea/vomiting, and pain), a global health/QoL scale, and 6 single items (dyspnea, insomnia, appetite loss, constipation, diarrhea, and financial difficulties). Items 29 and 30 have 7 possible responses, while the rest have 4 (ranging from 1 [not at all] to 4 [very much]). For all scales, firstly, the raw score was calculated as the mean of the component items. The score for the functional scales score was calculated according to the formula: score = (1 − (RS − 1)/range) × 100 and for the symptom scales or items and global health status/QoL according to the formula: score = ([RS − 1]/range) × 100. All of the scales and single items range in score 0–100 (Aaronson et al. Reference Aaronson, Ahmedzai and Bergman1993; Fayers et al. Reference Fayers, Aaronson and Bjordal2001; Velikova et al. Reference Velikova, Coens and Efficace2012).
EORTC QLQ-CR29: A disease-specific module for colorectal cancer patients addressing symptoms, side effects, body image, sexual functioning, and future perspective. It consists of 29 questions covering 5 functional domains (Appearance, Anxiety, Weight, and Sexual Interest in Men and Women) and 18 symptom-related domains. It has 4 multi-item scales (urinary frequency, blood and mucus in stool, stool frequency, and body image) and 19 single-items. Separate questions are provided for patients with and without a stoma, and questions for men and women regarding sexuality. Most questions pertain to the period of the last 7 days, except for sexuality, which covers 4 weeks. For all scales, firstly, the raw score was calculated as the mean of the component items. For single-items raw score is the value of that item. The score for the functional scales score was calculated according to the formula: score = (1 − (RS − 1)/range) × 100 and for the symptom scales or items and global health status/QoL according to the formula: score = ([RS − 1]/range) × 100. All of the scales and single items range in score 0–100. Higher functional scores indicate better functioning and higher symptom scores indicate more problems (Whistance et al. Reference Whistance, Conroy and Chie2009).
Translation and cultural adaptation of the EORTC QLQ-CR29
The translation process was carried out according to the recommended steps (EORTC Quality of Life Group 2017), starting with 2 forward translations by Serbian native speakers fluent in English. These versions were then reconciled by a third party, combining the best elements of both. Subsequently, this reconciled version underwent back-translation into English by 2 translators, highly proficient in English. Then back translation report was compiled detailing the translation steps, choices made, and comments, which was reviewed by the EORTC translation unit (TU). Upon addressing any feedback, an external proofreader, selected by the TU, reviewed the preliminary translation. The feedback received was thoroughly discussed until a consensus was reached. Pilot-testing with 15 patients confirmed the translation’s comprehensibility, with no significant issues reported. After thorough review and consensus, the TU approved the final translation, ensuring it met EORTC standards and the validation study’s objectives.
Statistical analysis
Data are presented as mean ± SD for continuous variables and number (percentage) for categorical variables. For the comparison of CR29 scores and items between groups independent T test or Mann-Whitney test were used where appropriate based on the normality of distribution. The p-values less than 0.05 were considered statistically significant. At scale level, floor and ceiling effects were considered to be present if more than 15% of respondents achieved, respectively, the lowest or the highest possible score (Terwee et al. Reference Terwee, Bot and de Boer2007). The statistical analyses were performed using SPSS version 23.0 software (SPSS Inc., Chicago, IL, USA) and R 4.3.2. (R Core Team (2023). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. URL https://www.R-project.org/.).
Construct validity
For construct validity Cronbach’s alpha coefficients were calculated to assess internal consistency. A value above 0.70 was considered acceptable (Cronbach Reference Cronbach1951). Construct validity involved assessing convergent and discriminant validity using the multitrait-multimethod (MTMM) analysis (Lowe and Ryan-Wenger Reference Lowe and Ryan-Wenger1992). Convergent Validity was determined by examining the correlations between items and scales within the EORTC QLQ-CR29 that should be related. Pearson’s correlation coefficients were used to assess these relationships, with coefficients of 0.40 or higher indicating good convergent validity. Discriminant validity was assessed by investigating the extent to which scales and items intended to measure different constructs demonstrated low correlations with each other. Low correlations (r < 0.30) between constructs that are theoretically distinct, demonstrating the questionnaire’s ability to differentiate various aspects of QOL.
Reliability
To estimate the reliability of the CR29 questionnaire, a split-half reliability analysis was conducted (de Vet et al. Reference de Vet, Mokkink and Mosmuller2017). This method evaluates the consistency of responses across 2 equivalent halves of the test. The correlation between the 2 halves’ scores was determined using the Pearson correlation coefficient as a preliminary step in the reliability estimation. This correlation served as the basis for the split-half reliability calculation, which was then adjusted for the full test length using the Spearman-Brown prophecy formula (de Vet et al. Reference de Vet, Mokkink and Mosmuller2017).
Concurrent validity
The concurrent validity of the EORTC QLQ-CR29 questionnaire was evaluated by comparing it with the EORTC QLQ-C30. The analysis focused on calculating Pearson’s correlation coefficients between corresponding scales and items of the QLQ-CR29 and QLQ-C30 questionnaires (Lin and Yao Reference Lin, Yao and A.C2014).
Known-groups validity
Clinical validity was assessed using the known-group method (Davidson Reference Davidson and A.C2014). Groups for comparison were based on the tumor location (colon and rectum), presence of a stoma and neoadjuvant therapy.
The study was conducted in accordance with the principles of the Declaration of Helsinki (2000 revision of Edinburgh). Our study was approved by the Ethics Committee of the University Clinical Center of Serbia (Number 808/5). All patients signed the informed consent to participate in the study before filling out the questionnaires.
Results
A total of 153 patients were approached for participation in the study: Of this, 2 declined, and communication was not possible with 1, resulting in 150 patients being included. The average age of included patients was 64.7 years, with a slight male predominance (53.3%). Sociodemographic and clinical characteristics of patients are presented in Table 1.
Sociodemographic and clinical characteristics of patients
SD – standard deviation, BMI – body mass index, CCI – Charlson Comorbidity Index, ASA – American Society of Anesthesiologists
Among the symptom scales, urinary frequency, blood and mucus in stool and stool frequency scales demonstrated good internal consistency with Cronbach’s alpha values of 0.769, 0.802 and 0.810, respectively. Furthermore, regarding functional scales, the body image scale exhibited a Cronbach’s alpha value of 0.855, indicating excellent internal consistency. The CR-29 questionnaire structure with internal consistency is presented in Table 2.
EORTC QLQ-CR29 questionnaire structure and internal consistency
SD – standard deviation, % floor – proportion of patients who scored at the lowest possible score, % ceiling – proportion of patients who scored at the highest possible score, α – Cronbach’s alpha.
The analysis revealed a split-half reliability coefficient of 0.872, signifying an excellent level of reliability of EORTC QLQ-CR29 questionnaire.
Overall, the EORTC QLQ-CR29 questionnaire shows good convergent validity for all examined scales. Divergent validity is also generally supported, with most scales showing low correlations with unrelated constructs (Table 3).
Convergent and divergent validity of EORTC QLQ-CR29 questionnaire
a Pearson’s correlation coefficient; bSpearman’s correlation coefficient; α – Cronbach’s alpha.
The correlations presented in Table 4 between CR-29 and QLQ-C30 scales/items demonstrate the concurrent validity of the CR-29 questionnaire. Body image positively correlated with most functional scales. Conversely, scales such as urinary frequency, blood and mucus in stool, and stool frequency showed negative correlations with functional scales but positive correlations with symptom scales like fatigue.
Correlation between EORTC QLQ-CR29 and C30 questionnaires
** Correlation is significant at the 0.01 level, *Correlation is significant at the 0.05 level, QL – quality of life, PF – physical functioning, RF – role functioning, EF – emotional functioning, CF – cognitive functioning, SF – social functioning, FA – fatigue, NV – nausea and vomiting, PA – pain, DY – dyspnea, SL – insomnia, AP – appetite loss, CO – constipation, DI – diarrhea, FI – financial difficulties.
Known-groups validity
The scales and single-items of CR-29 questionnaire based on tumor location, stoma presence, and neoadjuvant therapy are shown in Table 5. Regarding tumor location (Colon vs. Rectum) there were significant differences in a few areas: Blood/mucus in stool and stool frequency were more problematic in rectal cancer patients, with p-values of 0.046 and 0.029, respectively. Abdominal pain and dry mouth were significantly higher in colon cancer patients (p = 0.016, and p = 0.013). Dyspareunia showed a significant difference (p = 0.006), with higher scores in rectal cancer patients. Urinary incontinence and embarrassment were significantly more common in patients with a stoma (p = 0.040, p < 0.001). Hair loss and embarrassment were significantly more common in patients who received neoadjuvant therapy, with p-values of 0.026 and 0.048, respectively while dry mouth was more common in patients who did not receive neoadjuvant therapy (p = 0.015).
Known-groups validity of EORTC QLQ-CR29 questionnaire
Discussion
Considering that QoL assessments have become integral to the comprehensive management of cancer patients, offering indispensable insights into the disease and treatment impacts from the patient’s perspective. Our study aimed to translate, culturally adapt, and psychometrically validate the EORTC QLQ-CR29 questionnaire for Serbian colorectal cancer patients. This effort is particularly relevant given the rising incidence of colorectal cancer (Cancer Registry Reference Cancer Registry2023; Morgan et al. Reference Morgan, Arnold and Gini2023).
The absence of missing data underscores a high level of participant compliance, suggesting that the questionnaire was neither difficult, confusing, nor burdensome for the patients. This implies a strong acceptance of the questionnaire among Serbian patients. In terms of construct validity, our findings revealed that the Cronbach’s alpha coefficients for urinary frequency, blood and mucus in stool, and stool frequency scales were 0.769, 0.802, and 0.810, respectively which indicate good internal consistency. This shows that the items within these scales are cohesively measuring the intended constructs among patients. Furthermore, the body image scale, with a Cronbach’s alpha of 0.855, exhibits excellent internal consistency. Compared to the original study, our Cronbach’s alpha values were higher for all scales (Whistance et al. Reference Whistance, Conroy and Chie2009). Moreover, when compared to the Spanish, Dutch, Chinese, Malaysian, Moroccan, Ethiopian, Mexican and Polish versions, our study reported higher Cronbach’s alpha coefficients across all scales (Abebe et al. Reference Abebe, Wondimagegnehu and Woldemariam2021; Arrarás et al. Reference Arrarás, Suárez and Arias de la Vega2011; Hernández-Marín et al. Reference Hernández-Marín, Galindo-Vázquez and Calderillo-Ruíz2024; Lin et al. Reference Lin, Zhang and Wu2017; Magaji et al. Reference Magaji, Moy and Roslani2015; Sanna et al. Reference Sanna, Bereza and Paradowska2017; Stiggelbout et al. Reference Stiggelbout, Kunneman and Baas-Thijssen2016; Yacir et al. Reference Yacir, Hadj and Hafid2022). Studies by Ihn et al. (Reference Ihn, Lee and Son2015) and Shen et al. (Reference Shen, Chen and Ho2018) presented Cronbach’s alpha values similar to ours, indicating a consistent measure of internal consistency across different populations and settings.
For patients without a stoma, the internal consistency across all scales being above the 0.7 threshold (range, 0.764–0.871) highlights the questionnaire’s robustness in this subgroup. However, in patients with a stoma, while the urinary frequency and body image scales show strong internal consistency (alpha coefficients greater than 0.8), the blood/mucus in stool scale shows a lower reliability (alpha = 0.654), and the stool frequency scale exhibits poor reliability (alpha = 0.225). Other studies also showed better internal consistency of these scales among patients without stoma (Lin et al. Reference Lin, Zhang and Wu2017; Magaji et al. Reference Magaji, Moy and Roslani2015; Yacir et al. Reference Yacir, Hadj and Hafid2022). In comparison with the original study, we found that the internal consistency of CR29 scales was better in patients without stoma. However, in patients with stoma, Cronbach’s alpha coefficients were slightly lower for scales blood/mucus in stool and especially for stool frequency (Whistance et al. Reference Whistance, Conroy and Chie2009). Lin et al. (Reference Lin, Zhang and Wu2017) and Magaji et al. (Reference Magaji, Moy and Roslani2015) also revealed the lowest Cronbach’s alpha coefficient value in stool frequency among patients with stoma. These findings suggest the need for cautious interpretation of scores from these 2 scales in patients with a stoma.
The split-half reliability in our study was 0.872. Our study demonstrated that the EORTC QLQ-CR29 is a reliable tool for assessing quality of life in patients with colorectal cancer, providing stable and consistent results across its scales. This outcome aligns with data from other studies that have also confirmed the CR29 questionnaire’s reliability (Arrarás et al. Reference Arrarás, Suárez and Arias de la Vega2011; Ihn et al. Reference Ihn, Lee and Son2015; Lin et al. Reference Lin, Zhang and Wu2017; Shen et al. Reference Shen, Chen and Ho2018; Whistance et al. Reference Whistance, Conroy and Chie2009; Wickramasinghe et al. Reference Wickramasinghe, Dayasena and Seneviratne2020). In research contexts, the high reliability of the EORTC QLQ-CR29 supports its use in longitudinal studies to assess changes over time, as well as in intervention studies to evaluate treatment outcomes. This reinforces the questionnaire’s value in both clinical practice and research, providing reliable insights into patients’ quality of life.
The Serbian CR29 questionnaire demonstrated strong convergent validity, indicating that the scales are effective in measuring related constructs, aligning with the original study (Whistance et al. Reference Whistance, Conroy and Chie2009). Divergent validity was mostly satisfactory, with items showing low or no correlation with unrelated scales, ensuring a comprehensive and multi-dimensional QoL assessment without overlap. Exceptions were noted in stool frequency and body image scales among stoma patients, which was similar to divergent validity challenges in Polish, Korean, Moroccan, and Chinese versions (Ihn et al. Reference Ihn, Lee and Son2015; Lin et al. Reference Lin, Zhang and Wu2017; Sanna et al. Reference Sanna, Bereza and Paradowska2017; Yacir et al. Reference Yacir, Hadj and Hafid2022). Overall, the Serbian CR29’s construct validity is very good.
The significant correlations, both positive and negative, across various scales and items between the CR29 and C30 questionnaires indicate that the CR29 is effectively capturing aspects of health status and quality of life that align with the broader and well-established QLQ-C30. Scales and items with closely related content exhibit higher correlation coefficients, aligning with findings from previous research (Abebe et al. Reference Abebe, Wondimagegnehu and Woldemariam2021; Ihn et al. Reference Ihn, Lee and Son2015; Sanna et al. Reference Sanna, Bereza and Paradowska2017). Additionally, the presence of scales and items with low correlation coefficients suggests that the CR29 and C30 questionnaires are measuring distinct concepts, highlighting the complementary nature of the 2 instruments (Abebe et al. Reference Abebe, Wondimagegnehu and Woldemariam2021; Lin et al. Reference Lin, Zhang and Wu2017; Sanna et al. Reference Sanna, Bereza and Paradowska2017). Consequently, to ensure a comprehensive assessment of quality of life in patients with colorectal cancer, the CR29 module should be used together with the core C30 questionnaire.
The CR29 questionnaire effectively differentiates QoL issues in colorectal cancer patients based on tumor location, stoma presence, and neoadjuvant therapy. Rectal cancer patients reported more issues with blood/mucus in stool and stool frequency, whereas colon cancer patients experienced more abdominal pain and dry mouth, linked to colon blockage and stress-related symptoms (Acevedo-Ibarra et al. Reference Acevedo-Ibarra, Juárez-García and Espinoza-Velazco2021; Antoniadis et al. Reference Antoniadis, Giakoustidis and Papadopoulos2024). Additionally, the significantly higher rates of dyspareunia in rectal cancer patients could be attributed to the proximity to sexual organs, impacting sexual function (Ihn et al. Reference Ihn, Lee and Son2015; Kowal et al. Reference Kowal, Douglas and Jayne2022; Reese et al. Reference Reese, Handorf and Haythornthwaite2018). Stoma-associated embarrassment and psychological burden are significant, with rectal cancer patients often experiencing worse body image and urinary incontinence issues, although body image differences were not statistically significant (Abebe et al. Reference Abebe, Wondimagegnehu and Woldemariam2021; Bachri et al. Reference Bachri, Essangri and El Bahaoui2023; Ihn et al. Reference Ihn, Lee and Son2015; Lin et al. Reference Lin, Zhang and Wu2017; Sanna et al. Reference Sanna, Bereza and Paradowska2017; Whistance et al. Reference Whistance, Conroy and Chie2009). Chemotherapy, a common component of neoadjuvant therapy for colorectal cancer, can cause hair loss (Saraswat et al. Reference Saraswat, Chopra and Sood2019). This side effect can be particularly distressing for patients, as hair is often associated with personal identity and social perceptions of health (Özüsağlam and Can Reference Özüsağlam and Can2021; Mao et al. Reference Mao, Chen and Wu2019). Studies showed that hair loss is more common in patients who received neoadjuvant therapy (Ihn et al. Reference Ihn, Lee and Son2015; Shen et al. Reference Shen, Chen and Ho2018).
This study has several strengths, including a thorough validation process and high participant compliance. The study’s limitations are primarily due to its single-center design, which might restrict the generalizability of the findings. However, it’s important to note that the Clinic for Digestive Surgery at the University Clinical Center of Serbia, is the leading and the largest center for colorectal cancer treatment in the country. This center draws patients from diverse regions and socio-economic backgrounds across Serbia, somewhat mitigating concerns about representativeness. Other limitations are the relatively small sample size for specific subgroups, such as patients with a stoma which might have affected results in this subgroup.
The QLQ-CR29 questionnaire has been skillfully translated, adapted to the cultural context and rigorously validated though psychometric methods for Serbian patients with colorectal cancer. The findings underscore the questionnaire’s reliability and validity across various areas. The high level of participant compliance suggests the questionnaire is well-accepted among Serbian patients, indicating its appropriateness and non-burdensome nature. This validated tool enables healthcare professionals to comprehensively assess and monitor the quality of life impacts specific to colorectal cancer.
Funding
This research received no specific grant from any funding agency, commercial, or not-for-profit sectors.
Competing interests
The authors declare none.
Open access
