Data collection and participants
We retrospectively selected participating patients from the database of the National Oncology Institute of Rabat and the Private Oncology Center, on the period extending from January 2012 to March 2019. We did not include patients who underwent an abdominoperineal amputation, pseudocontinent perineal colostomy and those who had a stoma. The inclusion criteria were patients aged older than 18 years, diagnosed with rectal adenocarcinoma and who received low anterior resection with anastomosis creation or after a minimum 6 months interval after stoma reversal by October first 2019. Our exclusion criteria included cognitive dysfunction, the inability to speak moroccan arabic dialect and history of inflammatory bowel disease or any disease with bowel function impairment namely, Crohn's disease, irritable bowel syndrome, ulcerative colitis or others.
Selected patients were approached during their check ups in the day clinic and those who could not read received the help of an interviewer whose mission was to solemnly read the questions and answers. We also gave the questionnaire to patients over telephone interviews. Patients received the LARS questionnaire and the EORTC QLQ-C30 to prove convergent validity. Patients also received the WEXNER score for the purpose of another study.
Written, informed consent was obtained from all patients who participated in the study at the time of their initial visit so that their clinical data could be used for clinical studies so long as their privacy was not jeopardized.
Description of the instruments
The LARS was first developed on a nationwide cohort study of 961 Danish patients, then translated to English in 2012. (5,14) The LARS questionnaire consists of five items: “incontinence for flatus,” “incontinence for liquid stool,” “frequency of bowel movements,” “clustering of stools” and “urgency”. Each item has three to four response choices that are assigned with different score values. The third item has four choices, including “>7 times per day,” “4 to 7 times per day,” “1 to 3 times per day,” and “less than once per day,” assigned with values of 4, 2, 0, and 5 respectively. All the other four items have three choices, including “no, never,” “yes, less than once per week,” and “yes, at least once per week,” and are assigned with the values of 0, 4, and 7 for the first item; 0, 3, and 3 for the second item; 0, 9, and 11 for the fourth item; and 0, 11, and 16 for the fifth item, respectively.
The total score ranges from 0 to 42 and is categorized into three different groups: 0 to 20 points - no LARS, 21 to 29 points - minor LARS and 30 to 42 points - Major LARS.(5,14)
The original authors of the LARS did not respond to our emails, therefore we could not receive permission for translation. The translation to Moroccan Arabic was performed by two independent professional translators who discussed the translations until a provisional consensual version was reached. Thereafter a third translator who wasn’t familiar with the English version, back translated the agreed Arabic version to English. The three translators compared the English back-translated version against the original and a final Moroccan Arabic version was formed (Appendix 1). We conducted a preliminary test on a sample group from the target population to ensure understanding and absence of difficulties answering the questions.
The translations aimed at achieving conceptual equivalence rather than a word-for-word translation and the process followed the recommendations of the WHO and the European Organisation for Research and Treatment of Cancer (EORTC) (15) (16)
Reliability of the instrument was evaluated through Internal consistency and reproducibility which we investigated by the means of cronbach’s alpha coefficient calculation and test retest reliability respectively. A high positive value for Cronbach’s alpha superior to 0.70 suggests that the LARS score measures consistently.
We randomly selected a group of patients who responded to the questionnaire a second time after an interval of 2 to 4 weeks. Patients were asked if they had experienced any significant change in bowel function between the first and the second test. Those confirming a change in bowel function were excluded from the test– retest analysis. The correlation between the numerical value of the LARS score at the first and second test was assessed by the means of a Bland–Altman plot with 95% limits of agreement, as well as through measuring the intraclass correlation coefficient, which is considered to be adequate when superior to 0.80. Furthermore, for each of the 5 individual questions of the score, the agreement between the first and second response was explored by means of computing the percentage of perfect, moderate, and no agreement. A perfect agreement was assigned when participants ticked off exactly the same category at the first and second test, moderate agreement was assigned when responses differed by only 1 category, and no agreement was assigned when responses differed by 2 or 3 categories at the 2 tests.
For the purpose of convergent validity testing, the EORTC QLQ-C30 questionnaire was used. This quality of life assessment tool includes five functional subscales (i.e., physical functioning, role functioning, emotional functioning, cognitive functioning, and social functioning), three symptom subscales (i.e., fatigue, nausea and vomiting, and pain), a global QoL subscale, and six single symptom items (i.e., dyspnea, insomnia, appetite loss, constipation, diarrhea, and financial difficulties). We specifically analysed the correlation between the LARS score results and both the functional scales and the diarrhea symptom scale. (17)
With regards to the scoring instructions for this instrument, a high score represented a high QoL or a high level of functioning for the global QoL subscale and functional subscale. Opposingly, for a symptom subscale/item the higher the score, the more severe the symptoms are. (18)
For discriminant validity testing, we primarily hypothesized that the LARS score would differentiate between the bowel functions of patients with different demographic or clinical features such as sex, age, length of postoperative period (time since stoma-free rectal resection surgery or reversal surgery of temporary stoma), distance of the tumor from the anal verge, radiation therapy, extent of mesorectal excision, prior temporary stoma and post operative complications.
Demographic and clinical variables were analyzed by using descriptive statistics. When assessing the test-retest reliability, the Spearman correlation coefficient was used because both scores at the first and second surveys are non-normally distributed. The correlations between the LARS score and the subscales of the EORTC QLQ were evaluated by using Spearman's correlations as well. All p values <0.05 were considered statistically significant. All statistical analyses were performed using IBM SPSS Statistics (version 22).
Literature review :
We conducted a comprehensive search of prior translations and validations of the LARS score in MEDLINE (PUBMED) on a search period extending from January 1st, 2012 to October 1st, 2019. The MeSH and main keywords were the following: “Rectal neoplasms”, “Postoperative complications”, “fecal incontinence”, “Patient Outcome Assessment”, “ Surveys and Questionnaires”, “Translations”, “ Reproducibility of Results”, and “Psychometrics”.
The study received the approval of the biomedical ethics committee of the faculty of medicine in Rabat (Number: 99/19) and was approved and published in the international clinical trials database under the title: Validation of the Moroccan Arabic Version of the Low Anterior Resection Syndrome (LARS) and Wexner Score of Continence Among Rectal Cancer Patients (MA_LARSWEX) and the number: NCT04128657. (19)