Background
Inflammatory bowel diseases (IBD) are a group of chronic diseases characterised by an early onset and by a relapse and remission pattern. Prevalence has increased over time, mostly due to the low mortality rates associated with these conditions [
1]. Prevalence estimates in Portugal are, currently, among the highest in Europe [
2,
3]. The unpredictable onset of disease flares, the associated symptoms and the effects of treatment regimens strongly impact patients’ health-related quality of life (HRQoL) and, therefore, this is currently acknowledged as an important patient-reported outcome in IBD [
4].
Chen et al
. recently conducted a critical review of HRQoL instruments specifically designed for IBD patients [
5]. Among the many assessment tools, the IBD Questionnaire, developed by Guyatt et al
., was considered to be among the most suitable, valid and reliable. Irvine and colleagues validated a short, self-administered version of the IBD Questionnaire, which was also found to be valid and reliable in assessing the HRQoL of Crohn’s disease (CD) and ulcerative colitis (UC) patients’ [5–7]. This Short Inflammatory Bowel Disease Questionnaire (SIBDQ) version became widely known and is currently used worldwide both in clinical practice and clinical research. The SIBDQ comprises a total of 10 questions grouped into four different dimensions: social, bowel, emotional, and systemic [
6,
7]. Each question is scored by a 7-point Likert scale, ranging from 1 (a severe problem) to 7 (not a problem at all), giving an absolute SIBDQ score ranging from 10 (poor HRQoL) to 70 (optimal HRQoL). A SIBDQ score below 50 was considered as poor QoL. There are no validated cut-offs for the different dimensions’ scores. Hence, higher scores indicate a better HRQoL concerning that specific domain.
Although patient-reported outcomes are highly valuable for better patient care, patient responses to HRQoL instruments can be impacted by underlying cultural trends [
8]. Therefore, the direct translation of HRQoL questionnaires does not guarantee their adequacy and utility in countries others than those in which the questionnaire was designed and initially validated. As stressed, the SIBDQ is among the best-characterised tools to access IBD HRQoL. Since its development, it has been increasingly used in observational studies, clinical trial and clinical practice settings. However, a Portuguese translation and validation of the SIBDQ was yet to be performed.
This article describes the adaptation and validation of the Portuguese version of the SIBDQ (SIBDQ-PT). Our aim was to adapt an international instrument so that it could effectively and adequately evaluate HRQoL among IBD patients in Portugal.
Discussion
The importance of patient-reported outcomes, namely in what concerns HRQoL, is increasingly acknowledged as a key aspect in the management of IBD patients. HRQoL is a multidimensional concept that can be viewed as a latent construct encompassing physical, social and psychological aspects of well-being and role functioning [
21]. But despite the far-reaching nature of this definition, the importance of using culturally-validated instruments in local language is often ignored. Consequently, the quality of psychometric data falls behind that of the somatic data. With this study we aimed to translate and validate the widely used SIBDQ so that it could be adequately used in Portugal.
Eight translations of the original IBDQ have been validated in Europe, as well as two translations of the SIBDQ [
22‐
31]. No IBDQ nor SIBDQ translations have, to the best of our knowledge, been validated in Portugal. Some reviews do refer to a Portuguese validation of the IBDQ, but this validation was made in the Brazilian context [
32].
After independent forward and backward translations, context-validation by an IBD cohort, and an expert panel meeting, we have translated the SIBDQ to Portuguese (SIBDQ-PT). Also, as a result of this study the SIBDQ-PT was found to be structurally valid, reliable, convergent with sexual satisfaction and depression, to have an appropriate score distribution and to be responsive to change. In order to assess the psychometric data of the SIBDQ-PT, we have applied the questionnaire at three different timepoints separated by a two-week interval. The baseline questionnaire was self-administered to a panel of 92 IBD outpatients accounting for 92% of invited individuals. This panel size was considered to be appropriate, having into account that the original SIBDQ comprises a total of 10 items and the minimum patient-to-item ratio recommended to run a factor analysis is 7:1 [
19]. Regarding the second and third timepoints, and despite the fact that all 92 patients were invited to participate, only 33 were available to do so. Still, the sample size was appropriate to check the parameters that depended on the repetition of the questionnaire (re-test and responsiveness analysis).
The SIBDQ-PT construct validity was demonstrated through the confirmation of the a priori formulated hypotheses: an overall SIBDQ-PT score was positively correlated with sexual QoL (SQoL) and negatively correlated with depression symptoms (PHQ-9). The study conducted in the context of the British translation [
28] unveiled the presence of a strong correlation between the HRQoL, measured by the SIBDQ, and physician-evaluated disease activity indexes namely the Simple Clinical Colitis Activity index (SCCAI) and the Seo index: r = 0.83 (
p < 0.01) for the SCCAI and r = 0.61 (
p < 0.01) for the Seo index. However, the original American SIBDQ [
6,
7] acknowledges the existence of psychosocial dimensions other than the “bowel dimension”. Moreover, IBD-focused studies have shown that the patient self-perception of the illness is a better predictor of QoL, than IBD activity [
33‐
35]. Thus, this study aimed to be fully based on patient-reports
. We have, therefore, investigated the presence of correlations between HRQoL and other patient-reported concerns/symptoms like sexual QoL and depression, and identified a fair and a strong correlation, respectively. Sexual QoL may be a very different construct which is related to IBD-specific HRQoL but only distally, accounting for the fair correlation. Depression on the other hand, seems to be a measure better fitting this “patient perception of the illness” umbrella to test convergent validity of the SIBDQ. Construct validity was validated by the hypothesis testing method [
19] using only patients self-rated measures.
The results of the EFA analysis confirmed that the SIBDQ-PT measures different components of QoL. Similarly to what is observed in the original version [
6,
7], four dimensions—bowel (expressed by items 4, 6, 9 and 10), social (expressed by items 2, 3 and 5), emotional (expressed by items 1 and 8) and systemic (expressed by item 7)—were perceived in the Portuguese version of the SIBDQ. However, if we compare the original version and the Portuguese version (see Additional file
1), we notice that these dimensions were not expressed exactly by the same items. Item 10, for example, refers to the time during which the patient “felt angry as a result of his/her bowel problem”. In the original American SIBDQ this item is included in the emotional dimension, whether in the Portuguese version it is included in the bowel dimension. This question refers to “anger” related to the “bowel problem”, thus it adequately expresses both dimensions depending on the tone we put on the emotion and on the bowel condition. This is highly influenced by patients’ interpretations in different contexts, according to their language and population structure, and further reinforces the importance of using language-validated instruments. Items one and 5, referring to “exhaustion” and “demotivation”, also express different dimensions comparing both SIBDQ (systemic and emotional) and SIBDQ-PT (emotional and social), and a similar explanation applies. As so, we believe our four-dimension solution is not only supported by the total variance explained (over 70%), but also by its interpretability. As for the dimensions’ individual scores, alike the original SIBDQ there are no validated cut-offs. Hence, higher scores indicate a better HRQoL concerning that specific domain.
SIBDQ-PT was shown to have high internal consistency and test–retest reliability. The assessment of the score distribution included the evaluation of SIBDQ-PT total scores and the assessment of floor/ceiling effects, which were absent both at baseline and on the re-test. Sensitivity to change was stressed by the responsiveness analysis for patients who self-reported an overall worsening in their condition.
This study has several strengths that should be highlighted. Since the only criteria for participation in this study was age 18–65 years and a diagnosis for at least two years, the study population was community-based and, therefore, likely representative of all stages of the disease. As people tend to differ in their tendency to engage in socially desirable responding (SDR), which is a concern when analysing self-administered questionnaires, a SDR questionnaire—the SDS-FS [
10]—was administered in parallel with the SIBDQ-PT. Additionally, and despite the absence of heterogeneity in terms of age and gender distribution, the sample included patients with different educational levels and working/relationship status, and was therefore considered to be adequate for a cultural adaptation and validation [
5]. To our knowledge, this is the first study evaluating the score distribution and measurement bias of SIBDQ—neither the original version developers nor the authors of the British and German translations assessed this aspect [
28,
29]. Another important strength is the fact that the questionnaire was self-administered, ensuring patients were given the same instructions and the instrument was fulfilled under the same conditions. Data collection in other translation validations is performed over the phone and with different interviewers, which may influence the results. Nevertheless, this study also has limitations that should be noted. The study was conducted at a single centre, the sample is modest and recruited by convenience. Still this sample is assumed to be representative of the Portuguese population for the purposes of this translation, allowing adequate and complete psychometric assessment.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.